RU2792840C1 - Information search device - Google Patents

Abstract

FIELD: computing technology.

SUBSTANCE: information search device contains N≥2 mask storage 1₁–1 _N , N selection blocks 2₁-2 _N , frequency divider 3, time interval shaper 4, search strategy register 5, transition mask address generation block 6, indication block 7, N selective search time controllers 8_1-8 _N , main search time controller 9, clock pulse generator 10, fuzzy search scenario analysis unit 11, fuzzy search scenario transformation unit 12, catastrophe analysis unit 13.

EFFECT: increasing the reliability and stability of the implementation of search queries in the analysis of large arrays of heterogeneous data.

1 cl, 14 dwg

Description

The invention relates to the field of telecommunications and computer technology and is intended for searching and prompt identification of information in packet-switched data networks, in information and reference (search) systems and in systems for intelligent analytical processing of large arrays of heterogeneous data on cybersecurity events in the interests of assessing the state , decision support and investigation of computer incidents in critical infrastructures.

An information retrieval device is known, including N≥2 mask storage blocks, N selection blocks, a frequency divider, a time interval generator, a search strategy register, a transition mask address generation unit, an indication unit, N selection search time controllers, a main search time controller and a clock generator. pulses (see RF patent No. 2553093 "Information retrieval device" IPC G06F 9/46, published 06/10/2015, Bull. 16). The device makes it possible to increase the probability of timely information retrieval under conditions of continuous dynamics of changes in the state of search requests and taking into account influencing factors, due to decryption, dynamic correction and synchronization of the values (boundaries) of the maximum search time for each specific scenario (search request).

The disadvantage of this device is a relatively low scope, which does not allow to determine the type of expected blocks of binary information (BDI), in conditions where the parameters of the search script (values of identification signs (bit masks) and their order) may be unreliable.

An information search device is known that contains N≥2 mask storage blocks, N selection blocks, a frequency divider, a time interval generator, a search strategy register, a transition mask address generation unit, an indication unit, N selection search time controllers, a main search time controller, a generator clock pulses, a block for checking the search script and a block for correcting the search script (see patent No. 2656736 "Information search device" IPC G06F 9/46, published 06.06.2018, Bull. 18). This device allows you to increase the reliability of the implementation of search queries through preliminary analysis, identification (selection) of ambiguously (inaccurately, incompletely) identifiable values of the BDI type code and mathematically correct verification of these data, by converting identifiable (defined, recognized) inaccurately (incompletely, inconsistently) values code, to a form suitable for unambiguously making a reliable decision about the type of expected BDI.

However, this device has a drawback - a relatively low reliability of the implementation of search queries in conditions typical for real modern information and reference (search) systems - when the search scenario is implemented in conditions of fuzzy data on the true values of the parameters of this scenario.

The closest in technical essence to the claimed device (prototype) is an information retrieval device (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published 06/25/2020, 2020, Bull. 18), containing N ≥2 mask storage units, N selection units, frequency divider, time interval shaper, search strategy register, transition mask address generation unit, indication unit, N search time selection controllers, main search time controller, clock generator, fuzzy search scenario analysis unit and a fuzzy search script transformation unit. In this case, the clock input of the frequency divider is the first clock input of the device, and the output of the frequency divider is connected to the clock input of the time interval generator, the write permission inputs of N mask storage blocks are combined and are the device write permission input, L-bit information inputs, where L≥2, N selection blocks are combined and are an L-bit information input of the device. Moreover, the first L-bit inputs "Mask 1" and "Mask 2" of N mask storage units are the first L-bit inputs, respectively, "Mask 1" and "Mask 2" of the device. The second L-bit outputs "Mask 1" and "Mask 2" of N mask storage units are connected to the second L-bit inputs "Mask 1" and "Mask 2" of the corresponding selection units. The "Initial Reset" input of the time interval shaper is connected to the "Initial Reset" input of the transition mask address generation unit and is the "Initial Reset" input of the device. At the same time, the M-bit input "Waiting time code" of the time interval shaper, where M≥2 is the bit length of the waiting time code, is the M-bit input "Waiting time code" of the device, and the output of the time interval shaper is connected to the "Reset" input of the block formation of the address of the mask of transitions, the signal output of the register of the search strategy is connected to the signal inputs of the shaper of time intervals and the block for generating the address of the transition mask. Moreover, the K-bit address input, where K=(log ₂ N)+1, the control input, the N-bit information input and the enable input of the search strategy register are, respectively, the K-bit address input, the control input, the N-bit information input and the enable device input, the "Crystal Select" and "Read/Write" inputs of the search strategy register are respectively the "Crystal Select" and "Read/Write" inputs of the device, the N-bit "Search Completion Rule" input and the "Search Result" output of the display unit are respectively, the N-bit input "Search Completion Rule" and the output "Search Result" of the device. In this case, the output "Comparison result" of the n-th selection block, where n=1,2,...,N, is connected to the input "Comparison result" of the n-th selection search time controller, the output "Comparison result" of which is connected to the n-th with the "Comparison Result" input of the search strategy register and with the nth "Comparison Result" input of the transition mask address generation unit. The output of the clock pulse generator is connected to the clock input of the n-th selection block, the "Zeroing" input of which is the n-th "Zeroing" input of the device. At the same time, the S-bit corrective input, where S≥2 is the width of the corrective code of the search time, of the n-th selection block is combined with the S-bit test input of the n-th selection search time controller and is connected to the n-th S-bit output of the main controller search time, the N S-bit inputs of which are the corresponding N S-bit inputs of the “Maximum Search Time Correction” of the device. Moreover, the K-bit output "Event Code" of the transition mask address generation unit is connected to the K-bit information input of the fuzzy search scenario analysis unit, the K test outputs of which are connected to the corresponding K test inputs of the fuzzy search scenario conversion unit, the K-bit output of which and K -bit transit output of the fuzzy search scenario analysis block are interconnected and are K-bit inputs of the “Event Code” of the search strategy register and the indication block, respectively.

The prototype implements the possibility of increasing the reliability of the implementation of search queries through preliminary analysis, identification (selection) of fuzzy (observable) code values of the BDI type and mathematically correct verification of these data, by converting fuzzy (observable) code values, to a form suitable for unambiguous making a reliable decision about the type of BDI expected.

However, the prototype has a drawback - a relatively low reliability and stability of operation under conditions inherent in the real process of receipt (circulation) of large arrays (streams) of heterogeneous data in data transmission networks (DTN), in information reference and retrieval systems (ISPS) and in intelligent systems. Analytical Data Processing (SIAOD) for cybersecurity analysis of critical infrastructures (CVI). The prototype device is not capable of performing search queries in the dynamics of real SPD, ISPS and SIAOD, when the intensity of incoming data, the intensity of the BDI can smoothly change under the influence of control actions (search time for current search queries) or external factors, creating a potential threat of blocking (collapse ) the process of implementing search queries (the process of information retrieval) for systems of this class.

In other words, in the prototype there is no way to identify and verify the boundary and emergency (catastrophic) states characteristic of the emergency, critical position of the parameters of reliability and stability of the process of implementing search queries, when there are smooth and slight variations in external conditions and control actions (such as the intensity of incoming BDI or search time) occurring in real SPD, ISPS and SIAOD. This is due to the fact that the prototype device does not allow identification, verification and a priori warning about emergency (catastrophic) conditions for the implementation of the search process in large amounts of information, the implementation of search queries in large heterogeneous data streams. This excludes the use of the prototype for a reliable, stable, timely and dynamic information retrieval in real conditions, when there is a threat of reaching a critical (on the border of the possibility of implementing search queries) intensity of BDI arrival with smooth changes in the parameters of external conditions and control actions.

Such a situation may not manifest itself immediately during the next stage of the implementation of search queries, however, a gradual increase in the intensity of the receipt of BDI, in terms of maintaining the reliability and stability of the information search process (with smooth and slight variations in the parameters of the information flow), is very dangerous. The issues of identification and verification of possible catastrophic states of a controlled object are dealt with by a section of mathematical theory called the theory of catastrophes [1-5]. This theory is devoted to abrupt changes in the states of a controlled and analyzed object that occur as a sudden response of the system (object) to a smooth change in parameters caused by control or external influences.

Disasters on radio engineering, information and telecommunication, information retrieval and computing systems (for example, such as SPD, ISPS and SIAOD) lead to blocking, collapse of the process of their functioning, can act as unexpected "avalanche" failures, overloads of switching devices, abrupt fluctuations in channel capacity, abrupt changes in cybersecurity parameters, parameters of the signal propagation environment, etc. For example, in order to implement the process of intellectual analytical processing of large arrays of heterogeneous data on cybersecurity events in the interests of assessing the state, supporting decision-making and investigating computer incidents in critical infrastructures, the operator (administrator) generates control actions calculated for a certain throughput of the search system, taking into account a certain intensity of intake of the analyzed BDI. However, during the operation of the SPD, ISPS or SIAOD, a smooth drift of the parameters of the analyzed information flow occurs (for example, a smooth, increasing change in the intensity of receipt of blocks of binary information), which at an unforeseen point in time can lead to an abrupt change in the state of the quality indicators of the system as a whole - to a loss of reliability and sustainability of its operation. An adequate implementation of search queries should be focused on the mandatory identification and verification of the states of the boundary and emergency (catastrophic) intensity of the analyzed BDI in these streams, should predict a possible catastrophic state of the information retrieval system itself, thereby giving the administrator (user, operator) the opportunity to avoid states, characteristics of an emergency, critical state of the parameters of reliability and stability of the information retrieval process. Not taking into account the smooth change in the parameters of the analyzed information flows facilitates the task of implementing search queries, but sharply reduces the level of reliability of the information search results.

The “state of the boundary and emergency (catastrophic) intensity of incoming BDIs (λ=w/t _nab ) is understood as the number w of incoming data to control and identify elements of the incoming data stream, i.e., the number of BDIs to be analyzed coming to the L-bit (where L≥2) information input for the time interval t _set , specified by the control procedures and determined by the number of clock pulses, for example, from 10 to 1000. In this case, the number w of incoming elements of the incoming data stream (the number of incoming BDIs) is capable of smoothly increasing to a critical figure , at an unforeseen point in time, lead to an abrupt (in the overwhelming majority of cases - negative) change in the state of the quality indicators of the system, to the loss of reliability and stability of its functioning as a whole, can cause an avalanche-like change in the quality of the device, up to its collapse (blocking).

By "search" is meant a parallel analysis of the values of the attributes of identification of the BDI and control of the order of their sequence for compliance with the given rules for determining the code of the type of the BDI.

The “search script” is understood as the classical method of formal graphics, which determines the given rules for following the BDI.

The "identification flag" refers to the values of the bits in the corresponding positions of the BDI.

By "BDI type code" is meant the code corresponding to the search scenario and, from the point of view of the physics of the process, the address at which the transition mask is stored in the random access memory, which determines this particular type of expected BDI.

By "parameters of the search script" is meant a set of parameters that describe the rules for the sequence of blocks of binary information - a set (a set of code bits) of values of the identification signs (bit masks) of these blocks and their order.

The "implementation of search queries" is understood as a set of actions of the information and reference (search) system, including the selection of a request for information search from the queue, the allocation of a resource to this request and the actual search for information in accordance with a given search scenario, as well as performing final operations. Search request - sending a signal to search, initiating a response.

The purpose of the claimed technical solution is to create a controlled information retrieval device capable of increasing the reliability and stability of operation under conditions inherent in the real process of receipt (circulation) of large arrays (streams) of heterogeneous data in SPD, ISPS and in SIAOD, capable of both performing search queries and and to identify and verify the boundary and emergency (catastrophic) values of the intensity of the analyzed information flows under smooth changes in external conditions and control actions (intensity of incoming BDI). With this in mind, it is required to create an information retrieval device capable of timely notifying (warning) the user (operator, system administrator) of such a retrieval system about its possible emergency state, based on the received identification and verification data.

This goal is achieved by the fact that in a well-known information retrieval device containing N≥2 mask storage blocks, N selection blocks, a frequency divider, a time interval generator, a search strategy register, a transition mask address generation unit, an indication unit, N search time selection controllers, the main search time controller, a clock pulse generator, a fuzzy search scenario analysis unit and a fuzzy search scenario conversion unit, an additional catastrophe analysis unit is included, designed to carry out procedures for identifying and verifying the states of the boundary, emergency (catastrophic) intensity of incoming blocks of binary information with smooth changes in parameters external conditions and control actions, as well as for generating logical zero or logical one signals (prediction and warning signals), characterizing, respectively, the absence or presence of a possible catastrophic state of the system. In this case, the clock input of the frequency divider is the first clock input of the device, and the output of the frequency divider is connected to the clock input of the time interval shaper. The write enable inputs of the N mask storage units are combined and are the write enable input of the device. The first L-bit inputs “Mask 1” and “Mask 2” of the n-th mask storage block, where n=1,2,…,N, are the n-th first L-bit inputs, respectively, “Mask 1” and “Mask 2 » devices. The "Initial Reset" input of the time interval shaper is connected to the "Initial Reset" input of the transition mask address generation unit and is the "Initial Reset" input of the device. In this case, the M-bit input "Waiting time code" of the time interval shaper, where M≥2 is the width of the waiting time code, is the M-bit input "Waiting time code" of the device, and the output of the time interval shaper is connected to the "Reset" input of the formation unit jump mask addresses. The signal output of the search strategy register is connected to the signal inputs of the time slot generator and the transition mask address generation unit, respectively. K-bit, where K=(log ₂ N)+1, address input, control input, N-bit information input, and enable input of the search strategy register are respectively K-bit address input, control input, N-bit information input, and enable device input. The Chip Select and Read/Write inputs of the search strategy register are respectively the Chip Select and Read/Write inputs of the device. The N-bit input "Search Completion Rule" and the output "Search Result" of the display unit are, respectively, the N-bit input "Search Completion Rule" and the output "Search Result" of the device. The "Comparison Result" outputs of N selection blocks are connected to the corresponding "Comparison Result" inputs of N corresponding search time selection controllers, the "Comparison Result" outputs of which are connected to the corresponding N "Comparison Result" inputs of the search strategy register and to the corresponding N "Comparison Result" inputs transition mask address generation block. The output of the clock pulse generator is connected to the clock inputs of each of the N selection blocks, the "Reset" inputs of which are the corresponding "Reset" inputs of the device. Moreover, S-bit, where S≥2 is the bit length of the corrective search time code, the corrective input of the n-th selection block is connected to the S-bit test input of the n-th selection search time controller and is connected to the n-th S-bit output of the main time controller search, N S-bit inputs of which are the corresponding N S-bit inputs "Maximum search time correction" of the device. At the same time, the K-bit output "Event Code" of the transition mask address generation unit is connected to the K-bit information input of the fuzzy search scenario analysis unit, the K test outputs of which are connected to the corresponding K test inputs of the fuzzy search scenario transformation unit, the K-bit output of which and The K-bit transit output of the fuzzy search scenario analysis block are interconnected and are the K-bit inputs of the “Event Code” of the search strategy register and the indication block, respectively. L-bit information inputs, where L≥2, N selection blocks are combined and are an L-bit information output of the catastrophe analysis block, the L-bit information input of which is an L-bit information input of the device. The second L-bit outputs "Mask 1" and "Mask 2" of the n-th mask storage unit are connected to the corresponding n-th second L-bit inputs "Mask 1" and "Mask 2" of the disaster analysis unit, the n-th second L- the bit outputs "Mask 1" and "Mask 2" of which are connected to the corresponding second L-bit inputs "Mask 1" and "Mask 2" of the n-th selection block. The output of the clock pulse generator is connected to the clock input of the catastrophe analysis unit, the control input of which is the “Input of intensity threshold values” input of the device, the warning output of the catastrophe analysis unit is the device’s “Threat of a catastrophe” output.

The disaster analysis unit consists of an intensity calculator, G executable random access memory (RAM), where G≥1 is the number of time intervals t _{nab of} observation, a read only memory (ROM), an iterative comparison element, a comparison element, an intermediate RAM, an intermediate AND element, element AND, additional RAM, test RAM and counter. In this case, the clock input of the intensity calculator is connected to the clock inputs of the ROM and the counter and is the clock input of the disaster analysis unit, the L-bit information input of the intensity calculator is the L-bit information input of the disaster analysis unit and the L-bit information input of the device, the L-bit information output of the intensity calculator is an L-bit information output of the catastrophe analysis unit, the g-th executive output of the intensity calculator, where g=1,2,…,G, is connected to the input of the g-th execution RAM, the inputs of G of the execution RAM are combined and connected to the first input the iterative comparison element and the second input of the comparison element, the outputs G of the execution RAM are combined and connected to the second input of the iterative comparison element. The output of the ROM is connected to the first input of the comparison element, the output of the iterative comparison element is connected to the input of the intermediate RAM and the second input of the intermediate AND element, the first input of which is connected to the output of the intermediate RAM, the output of the intermediate AND element is connected to the output of the comparison element and is connected to the first input of the AND element , the second input of which is connected to the test input of the ROM and is the test output of the test RAM. The output of the AND element is connected to the reading inputs of the intensity calculator and the additional RAM and is a warning output of the catastrophe analysis unit and the device's Threat of a catastrophe output. Moreover, N L-bit inputs "Mask 1" and N L-bit inputs "Mask 2" of the additional RAM are the corresponding N second L-bit inputs "Mask 1" and the corresponding N second L-bit inputs "Mask 2" of the disaster analysis unit, N L-bit outputs "Mask 1" and N L-bit outputs "Mask 2" of the additional RAM are the corresponding N second L-bit outputs "Mask 1" and the corresponding N second L-bit outputs "Mask 2" of the disaster analysis unit. The clock output of the counter is connected to the clock input of the test RAM, the release output of which is connected to the release input of the counter, the control input of the test RAM is the control input of the catastrophe analysis unit and the input "Input intensity thresholds" of the device.

The number "N, (n=1,2,…,N; N≥2)" (blocks, bits, inputs, outputs, etc.) is determined in accordance with the possible number of BDI types (they determine the total number of transition masks, characterizing the composition of search scenarios) and, as a rule, ranges from 2 (two) to 500 (five hundred).

The number "K, (where K=(log ₂ N)+1)" characterizes the bit depth of the transition mask address code, the address where the transition mask is stored in the random access memory, which determines the type of BDI expected according to the search scenario. In other words, this is the number of bits sufficient to address N jump masks and the search script start mask, typically between 2 (two) and 20 (twenty).

The number "M, (M≥2)" characterizes the length of the waiting time code - the code of the allowable time interval during which the next BDI is expected, specified by the search scenario and, as a rule, ranges from 2 (two) to 10 (ten).

The number "L, (L≥2)" characterizes the maximum possible number of digits in the BDI used in the search scenario and ranges from 2 (two) to 10 (ten).

The number "S, (S≥2)" characterizes the word length of the search time correction code, the word length of the code that determines the initial or corrected (new) value of the maximum search time for each specific scenario (search query) and, as a rule, ranges from 2 (two ) up to 10 (ten).

The number "λ, (λ∈Λ; λ≥1; λ=w/t _nab )" characterizes the possible intensity of the elements of the incoming data stream arriving at the input of the device (for control and identification), has the physical meaning of the number of BDI w to be analyzed, arriving for unit of time t _nab , λ is determined by the system's search capabilities and performance boundary values (in terms of operational information identification). This number varies within λ=1,2,…,Λ and, as a rule, ranges from 1 (one) to 1000 (thousands).

The number "g, (g∈G; G≥1; g=1,2,…,G)" characterizes the possible number of independent, autonomous, successive time intervals t _{nab of} observation in the implementation of search queries, is determined by the system's capabilities in terms of general the duration of information retrieval, performance limits, and, as a rule, ranges from 2 (two) to 100 (one hundred).

The number "w, (w∈W; w≥1)" characterizes the possible number of elements of the incoming data stream (the number of BDIs to be analyzed) coming as part of the search for information at the input of the device, is used to calculate the intensity of incoming BDIs when implementing search queries and, as usually ranges from 1 (one) to 10,000 (ten thousand).

The principle of creating the proposed controlled information retrieval device is based on the well-known results of research in the field of catastrophe theory, presented in [1-6]. The analysis of these works makes it possible to form a mathematically correct algorithm for identifying and verifying boundary and emergency (catastrophic) states, characteristic of an emergency, critical state of the reliability and stability parameters of the system, for a situation that can manifest itself with smooth changes in the parameters of external conditions and control actions (such as the intensity of the flow BDI) taking place in real SPD, ISPS and SIAOD.

Thus, within the framework of the controlled process of searching for information in the incoming flows of the BDI, the problem of a priori analysis (estimation) and comparison of the intensity values of these flows is solved. From the point of view of physical interpretation, this is the process of a priori statistical analysis of smooth and insignificant changes in external conditions and control actions on the system with the possibility of alerting (warning) the user (operator, system administrator) about potential catastrophic consequences in its behavior that are not visible at first glance and almost never taken into account when implementing algorithms for implementing search queries.

With this approach to information retrieval, it is possible to represent the dynamics of changes in the state of this system with smooth and slight variations in external conditions and control actions, in the form of dynamics of changes in the intensity (λ=w/t _nab ) of incoming BDI to be analyzed - changes in the number w of incoming elements of the incoming data flow, i.e., the number of BDIs to be analyzed, arriving at the L-bit (where L≥2) information input of the device for the time interval t _nab specified by the control procedures and determined by the number of clock pulses (for example, from 10 to 1000). Analysis of the results of works [1-6] makes it possible to provide in the information retrieval device the possibility of identifying and verifying the boundary and emergency (catastrophic) states of a system of this class with smooth changes in the parameters of external conditions and control actions.

Mathematical formalization of the physical parameters of external conditions and control actions that affect the behavior of the information retrieval device can be represented by statistically determining the corresponding intensity values λ as the number of BDI w arriving per unit of time t _nab , where λ=1,2,…,Λ (can be, for example, from 1 (one) to 1000 (thousands)). The number g, where g=1,2,…,G is the number of successive time intervals t _of observations (can be, for example, from 1 (one) to 100 (one hundred)).

The total number of such BDI intensity values for the entire observation time in the interests of implementing search queries is Λ, and these values represent a set:

where each g-th element of the set, except for Λ _G (t _nab + (G-1)), is a subset of Λ _G and has the physical meaning of exceeding the threshold of the system's capabilities to search for information coming with a given intensity, and, as a result, a high probability the transition of the process of implementing search queries into an emergency (catastrophic) state at the next time interval (t _nab +1) of monitoring and operation of the information retrieval device.

Obviously, in order to solve the problem of a priori calculation and comparison of the intensity values of incoming BDI with smooth changes in the parameters of external conditions and control actions, it is necessary to carry out current step-by-step monitoring, identify and verify the boundary and emergency (catastrophic) states of the system.

Identification of the boundary and emergency (catastrophic) states of the system is carried out by step-by-step (at each step, i.e., each g-th time interval (t _set + g) observation in the interests of implementing search queries) a priori estimation and comparison of the values of each λ _g g -th element of the set Λ _g in order to determine the presence or absence of a possible excess of these values of the permissible threshold of the intensity of the intake of BDI, determined by the expression:

where Λ _G is the allowable value of the intensity of the arrival of BDI at the g-th time interval (t _nab + g) of observation in the interests of implementing search queries, if it is exceeded, the system will most likely go into an emergency (catastrophic) state from any other state, taking into account the fact that, in general, the operation of an information retrieval device can be considered as a process of transition of this device from state to state.

For the first step of observation in the interests of implementing search queries, the initial (first, starting) time interval of the information search process, this expression has the form:

Exceeding, on one of the subsequent (( _tset +1), ( _tset +2), etc.) time intervals, by any value of the intensity λg of this threshold, characterizes the beginning of a smooth change in the parameters of external conditions and control actions.

Verification of boundary and emergency (catastrophic) states is an identification-independent process that characterizes the excess of any value of the intensity of BDI λg in the previous time interval t _nab (parameter reading (λ ₁ )) over the value from each subsequent reading (λ2), in the next time interval interval (t _nab +1) and is determined, for example, for the first and second observation time interval in the interests of implementing search queries, in accordance with the expression

The physical meaning of the verification procedure is to identify the trend of changing the parameters of external conditions and control actions (intensity of incoming BDI) towards the boundary and emergency (catastrophic) state of the system. In both cases of a priori calculation and comparison of the intensity values of incoming BDI, as in the implementation of the procedure for identifying the boundary and emergency (catastrophic) states of the system, when an event is identified

and during the verification procedure, when the trend of changing the parameters of external conditions and control actions (intensity of incoming BDI) is confirmed towards the boundary and emergency (catastrophic) state of the information retrieval system

the user (operator, system administrator) who manages the process of implementing search queries should be notified (warned) about a possible emergency state of the information retrieval system.

If the user (operator, system administrator) is not able to influence an undesirable change in the parameters of external conditions and control actions (intensity of incoming BDI) or needs, for example, to check the operation of the search system in critical conditions, in obtaining precisely its boundary and emergency (catastrophic) states , the process of implementing search queries will be carried out without adjusting the intensity of incoming BDI or threshold values of this intensity.

If the user (operator, system administrator) is able to influence an undesirable change in the parameters of external conditions and control actions (the intensity of incoming BDI) and the process of implementing search queries is carried out within the framework when emergency conditions are unacceptable, based on the received identification and verification data, an external correction of the intensity of incoming BDI or their threshold values in order to prevent an emergency (catastrophic) abrupt change in the states of the system under small disturbances [2].

Examples illustrating similar, from the point of view of catastrophe theory, operations to prevent loss of stability and reliability of complex controlled systems under smooth changes in external conditions are given in [1] and [2], here are algorithms for analyzing the structural stability of objects and estimating critical points (Morse points ) in the process of the system functioning, characterizing the local maxima and minima of the stable (non-catastrophic) behavior of the object under smooth changes in external conditions.

The analysis of expressions (1) (6) allows us to conclude that it is technically possible to implement search queries and procedures for identifying and verifying the boundary and emergency (catastrophic) states of the information retrieval system with smooth changes in the parameters of external conditions and control actions (smooth changes in the intensity of incoming elements of the information flow) .

Thanks to a new set of essential features, due to the introduction of a catastrophe analysis unit designed to carry out procedures for identifying and verifying the states of the boundary, emergency (catastrophic) intensity of incoming blocks of binary information with smooth changes in the parameters of external conditions and control actions, as well as for generating logical zero or logical unit (prediction and warning signal), characterizing, respectively, the absence or presence of a possible catastrophic state of the system, in the claimed controlled device, it is possible to increase the reliability and stability of its operation under conditions inherent in the real process of searching in large amounts of information, the real process of implementing search queries in streams of heterogeneous data of large volume. This provides a reliable, stable, timely and dynamic search and operational identification of information in real conditions, when there is a threat of reaching a critical (at the border of the possibilities of implementing search queries) intensity of the arrival of BDI with smooth changes in the parameters of external conditions and control actions, as well as the ability to timely notify ( warn) the user (operator, system administrator) of the system about its possible emergency state, based on the received identification and verification data.

The claimed device is illustrated by drawings, which show:

in fig. 1 is a block diagram of an information retrieval device;

in fig. 2 block diagram of the disaster analysis block;

блока хранения маски;in fig. 3 - block diagram of the n-th

mask storage unit;

блока селекции;in fig. 4 - block diagram of the n-th

selection block;

in fig. 5 - block diagram of the shaper of time intervals;

in fig. 6 is a block diagram of a search strategy register;

in fig. 7 - block diagram of the block for generating the address of the transition mask;

in fig. 8 - block diagram of the display unit;

селекционного контроллера времени поиска;in fig. 9 block diagram of the n-th

selection search time controller;

in fig. 10 block diagram of the main controller of the search time;

in fig. 11 block diagram of the fuzzy search scenario analysis unit;

in fig. 12 is a block diagram of the fuzzy search script conversion unit;

in fig. 13 is an example of a search script;

in fig. 14 is an example of filling in the transition masks, the mask of the beginning of the scenario and the mask of the end of the scenario.

The device (see Fig. 1) consists of N, where N≥2, mask storage blocks 1 ₁ - 1 _N , N selection blocks 2 ₁ - 2 _N , frequency divider 3, time interval shaper 4, search strategy register 5, block formation of the address of the transition mask 6, the display unit 7, N selection search time controllers 8 ₁ - 8 _N , the main search time controller 9, the clock pulse generator 10, the fuzzy search scenario analysis unit 11, the fuzzy search scenario conversion unit 12 and the catastrophe analysis unit 13 .

The elements are interconnected as follows (see Fig. 1). The clock input 38 of the frequency divider 3 is the first clock input 08 of _the device, _and the _output 39 of the frequency divider 3 is connected to the clock input ₄₁ of the time interval generator 4. device write permissions 01. The first L-bit inputs “Mask 1” 17 _n and “Mask 2” 18 _n of the n-th mask storage block 1 _n , where n=1,2,…,N, are the n-th first L-bit inputs, respectively, “Mask 1" 02 _n and "Mask 2" 03 _n devices. The input "Initial reset" 42 shaper time intervals 4 is connected to the input "Initial reset" 63 of the formation address mask transitions 6 and is the input "Initial reset" 05 of the device. At the same time, the M-bit input "Wait time code" 43 of the time interval generator 4, where M≥2 is the width of the wait time code, is the M-bit input "Wait time code" 06 of the device, and the output 44 of the time interval generator 4 is connected to the "Reset" input 64 of the transition mask address generating block 6. The signal output 59 of the search strategy register 5 is connected to the signal inputs 45 and 65 of the time interval generator 4 and the transition mask address generating block 6, respectively. K-bit, where K=(log ₂ N)+1, address input 53, control input 54, N-bit information input 55 and enable input 58 of search strategy register 5 are respectively K-bit address input 09, control input 010, N-bit information input 011 and enabling input 014 of the device. The Chip Select 56 and Read/Write 57 search strategy register 5 inputs are Chip Select 012 and Read/Write 013 inputs of the device, respectively. The N-bit input "Search completion rule" 72 and the "Search result" output 73 of the display unit 7 are, respectively, the N-bit input "Search completion rule" 07 and the output "Search result" 015 of the device. The outputs "Result of comparison" 26 ₁ - 26 _N selection blocks 2 ₁ - 2 _N are connected to the corresponding inputs "Result of comparison" 81 ₁ - 81 _N of the corresponding selection controllers of the search time 8 ₁ - 8 _N , outputs "Result of comparison" 83 ₁ - 83 _N of which are connected to the corresponding inputs "Comparison result" 51 ₁ - 51 _N of the search strategy register 5 and to the corresponding inputs "Comparison result" 61 ₁ - 61N of the transition mask address generation unit 6. The output 101 of the clock generator 10 is connected to the clock inputs 25 ₁ - 25 _N of each of the N selection blocks 2 ₁ - 2 _N , the "Zero" inputs 24 ₁ - 24 _N of which are the corresponding "Zero" inputs 017 ₁ - 017 _N of the device.

Moreover, S-bit, where S≥2 is the bit length of the corrective search time code, the corrective input 27 _n of the n-th selection block 2 _n is connected to the S-bit test input 82 _n of the n-th selection controller of the search time 8 _n and is connected to n- th S-bit output 92 _n of the main search time controller 9, N S-bit inputs 91 ₁ - 91 _N of which are the corresponding N S-bit inputs "Maximum search time correction" 016 device. At the same time, the K-bit output "Event Code" 62 of the block for generating the address of the transition mask 6 is connected to the K-bit information input 111 of the block for analyzing the fuzzy search scenario 11, K test outputs 112 ₁ - 112 _K of which are connected to the corresponding K test inputs 121 ₁ - 121 _K of the conversion block of the fuzzy search script 12, the K-bit output 122 of which and the K-bit transit output 113 of the analysis block of the fuzzy search script 11 are combined and are K-bit inputs "Event Code" 52 and 71 of the search strategy register 5 and block indication 7 respectively. L-bit information inputs 21 ₁ - 21 _N , where L≥2, N selection blocks 2 ₁ - 2 _N are combined and are L-bit information output 131 of the disaster analysis block 13, L-bit information input 130 of which is L-bit information input 04 of the device. The second L-bit outputs "Mask 1" 14 _n and "Mask 2" 15 _n of the n-th mask storage unit 1 _n are connected to the corresponding n-yn second L-bit inputs "Mask 1" 132 _n and "Mask 2" 133 _n catastrophe analysis block 13, the n-th second L-bit outputs "Mask 1" 134 _n and "Mask 2" 135 _n of which are connected to the corresponding second L-bit inputs "Mask 1" 22 _n and "Mask 2" 23 _n n- th selection block 2 _n . The output 101 of the clock generator 10 is connected to the clock input 138 of the catastrophe analysis unit 13, the control input 136 of which is the input "Input intensity thresholds" 018 of the device, the warning output 137 of the catastrophe analysis unit 13 is the output "Catastrophe threat" 019 of the device.

The catastrophe analysis unit 13 shown in FIG. 2 is designed to carry out procedures for identifying and verifying the states of the boundary, emergency (catastrophic) intensity of incoming blocks of binary information with smooth changes in the parameters of external conditions and control actions, as well as for generating signals of a logical zero or a logical unit (a prediction and warning signal), characterizing, respectively, the absence or presence of a possible catastrophic state of the system.

The disaster analysis unit 13 (Fig. 2) consists of an intensity calculator 13.0, G executive random access memory (RAM) 13.1 ₁ , 13.1 ₂ , ..., 13.1 _G , read only memory (ROM) 13.2, iterative comparison element 13.3, comparison element 13.4 , intermediate RAM 13.5, intermediate element And 13.6, element And 13.7, additional RAM 13.8, test RAM 13.9 and counter 13.10.

The clock input 13.0-3 of the intensity calculator 13.0 is connected to the clock inputs 13.2-1 and 13.10-3 of the ROM 13.2 and the counter 13.10, respectively, and is the clock input 138 of the catastrophe analysis unit 13, the L-bit information input 13.0-1 of the intensity calculator 13.0 is L-bit information input 130 of the disaster analysis unit 13 and the L-bit information input 04 of the device, the L-bit information output 13.0-2 of the intensity calculator 13.0 is the L-bit information output 131 of the disaster analysis unit 13, the g-th executive output 13.0-4 _g of the intensity calculator 13.0, where g=1,2,...,G, is connected to the input of the g-th executive RAM 13.1 _g , the inputs G of the executive RAM 13.1 ₁ - 13.1 _G are combined and connected to the first input 13.3-1 element of the iterative comparison 13.3 and the second input 13.4 -2 comparison element 13.4, the outputs G of the execution RAM 13.1 ₁ - 13.1 _G are combined and connected to the second input 13.3-2 of the iterative comparison element 13.3. The output of the ROM 13.2 is connected to the first input 13.4-1 of the comparison element 13.4, the output of the iterative comparison element 13.3 is connected to the input of the intermediate RAM 13.5 and the second input 13.6-2 of the intermediate element And 13.6, the first input 13.6-1 of which is connected to the output of the intermediate RAM 13.5, the output intermediate element And 13.6 connected to the output of the comparison element 13.4 and connected to the first input 13.7-1 element And 13.7, the second input 13.7-2 which is connected to the test input 13.2-2 ROM 13.2 and is a test output 13.9-1 test RAM 13.9. The output of the AND element 13.7 is connected to the reading inputs 13.0-5 and 13.8-1 of the intensity calculator 13.0 and the additional RAM 13.8, respectively, and is a warning output 137 of the catastrophe analysis block 13 and the output "Threat of a catastrophe" 019 of the device. In this case, N L-bit inputs "Mask 1" 13.8-2 ₁ - 13.8-2 _N and N L-bit inputs "Mask 2" 13.8-3 ₁ - 13.8-3 _N additional RAM 13.8 are the corresponding N second L-bit inputs "Mask 1" 132 ₁ - 132 _N and the corresponding N second L-bit inputs "Mask 2" 133 ₁ - 133 _N of the disaster analysis unit 13, L-bit outputs "Mask 1" 13.8-4 ₁ - 13.8-4 _N and N L-bit outputs "Mask 2" 13.8-5 ₁ - 13.8-5 _N additional RAM 13.8 are the corresponding N second L-bit outputs "Mask 1" 134 ₁ - 134 _N and the corresponding N second L-bit outputs "Mask 2" 135 ₁ - 135 _N of the catastrophe analysis block 13. The clock output 13.10-1 of the counter 13.10 is connected to the clock input 13.9-2 of the test RAM 13.9, the release output 13.9-3 of which is connected to the release input 13.10-2 of the counter 13.10, the control input 13.9-4 of the test RAM 13.9 is the control input 136 of the catastrophe analysis block 13 and the input "Input of intensity threshold values" 018 of the device.

The intensity calculator 13.0 of the disaster analysis block 13 is designed to record, store and read from the executive outputs 13.0-4 ₁ - 13.0-4 _G in binary code of the values of the incoming data stream (incoming BDI), calculating the intensity of incoming BDI for each g-th (g= 1,…,G) successive time interval t _{nab of} observation, as well as reading from the L-bit information output 13.0-2 of the identified and verified values of the BDI binary digits in the parallel code to the L-bit information inputs 21 ₁ - 21 _N blocks selection 2 ₁ - 2 _N . The construction scheme of the intensity calculator 13.0 is known, it can be technically implemented on the basis of a commercially available programmable summing counter with a RAM function and controlled reset, the scheme of which is known and described, for example, in [Ugryumov E.P. Digital circuitry: textbook. allowance for universities. - 3rd ed., revised. and additional - St. Petersburg: BHV-Petersburg, 2010. - 816 p., S. 246-247, fig. 3.46].

Execution RAM 13.1 ₁ - 13.1 _G of the catastrophe analysis unit 13 are identical and are designed to record, store and read in binary code from the memory cells the intensity values of the incoming BDI for each g-th (g=1, ..., G) successive time interval t _nab observation, i.e., for example, intensity values (λ ₁ ) at t _nab -th (previous) observation time interval, intensity values (λ ₂ ) at (t _nab +1)-th (next), etc. etc., the time interval of observation and operation of the information retrieval device. Executive RAM 13.1 ₁ - 13.1 _G can be technically implemented on the basis of commercially available random access memory, as shown in [Averchenkov O.E. Circuitry: hardware and software. - M.: DMK Press, 2012. - 588 p., S. 246-247].

Read-only memory 13.2 of the disaster analysis unit 13 is intended for pre-recording, registration, storage and reading in binary code from memory cells to the first input of the comparison element 13.4 of a pre-recorded allowable value (threshold) of the intensity Λ _g of incoming BDI for each g-th (g= 1,…,G) of the time interval t _nab of observation, which makes it possible to determine the presence or absence of a possible excess of the actual values of the BDI intensity λ _g of this admissible threshold. The technical implementation of the ROM 13.2 is possible on the basis of a register with the functions of a read-only memory device, a description of the operation and scheme of such registers are known and are given, for example, in the book [Averchenkov O.E. Circuitry: hardware and software. - M.: DMK Press, 2012. - 588 p., S. 443-445].

The iterative comparison element 13.3 of the catastrophe analysis block 13 is designed for sequential step-by-step (cycle-by-cycle, at each step (cycle), i.e., each g-th time interval (t _set +g) of observation and operation of the search device) a priori estimation (calculation) and comparing the values of each λ _g of the g-th element of the set Λ _g with each other in order to determine the presence or absence of a possible excess of these values of the permissible threshold for the intensity of the analyzed BDI in accordance with expression (4). The iterative comparison element 13.3 can be technically implemented on the basis of a commercially available digital comparison unit (digital comparator), as shown in the book [Averchenkov O.E. Circuit engineering: hardware and software. - M.: DMK Press, 2012. - 588 p., S. 428-429].

Comparison element 13.4 of the disaster analysis block 13 is designed for sequential step-by-step (cycle-by-cycle, at each step (tact), i.e., each g-th time interval (t _set +g) of observation and operation of the search device) a priori estimation (calculation) and comparing the pre-recorded allowable value (threshold) of the intensity Λ _g of the incoming BDI with the real values of the real values of the intensity λg in accordance with expression (2). Comparison element 13.4 is a digital comparison node (digital comparator) described in [Averchenkov O.E. Circuitry: hardware and software. - M.: DMK Press, 2012. - 588 p., S. 428-429].

The intermediate RAM 13.5 of the catastrophe analysis block 13 is intended for recording, intermediate storage and reading in binary code of a logical zero or a logical one characterizing the result of identification and verification obtained at the g-th time interval (cycle). Intermediate RAM 13.5 can be implemented on the basis of dynamic RAM, described in the literature [Averchenkov O.E. Circuitry: hardware and software. - M.: DMK Press, 2012. - 588 p., S. 246-247].

Intermediate element And 13.6 block analysis of catastrophes 13 is designed to compare the obtained result of identification and verification (logical zero or logical unit) on the previous time interval (tact) of observation (t _nab ) with the result of identification and verification (logical zero or logical unit) obtained on the next (t _set +1) time interval (tact) of observation and operation of the search device. A special case of the technical implementation of the intermediate element AND 13.6 is described in [Ugryumov E.P. Digital circuitry: textbook. allowance for universities. - 3rd ed., revised. and additional - St. Petersburg: BHV-Petersburg, 2010. - 816 p., S. 79-81].

Element I 13.7 of the catastrophe analysis block 13 is intended to confirm (actually verify) the identified trend in the change in the intensity of incoming BDI towards the boundary and emergency (catastrophic) state, as well as to implement the procedure for notifying the administrator (user, operator) about a possible catastrophic state of the number of BDI w, arriving per unit of time t _nab (catastrophic state of the intensity of incoming BDI). Element AND 13.7 can be technically implemented on the basis of a typical logical element AND, described in [Ugryumov E.P. Digital circuitry: textbook. allowance for universities. - 3rd ed., revised. and additional .- St. Petersburg: BHV-Petersburg, 2010. - 816 p., S. 79-81].

Additional RAM 13.8 of the catastrophe analysis unit 13 is intended for recording, storing and reading the results of identification and verification on command from the element And 13.7 from its N L-bit outputs "Mask 1" 13.8-4 ₁ - 13.8-4 _N and N L-bit outputs "Mask 2" 13.8-5 ₁ - 13.8-5 _N of the corresponding first and second bit masks of the initial search script and transition masks, which, acting respectively on the second L-bit inputs "Mask 1" 22 ₁ - 22 _N and the second L- bit inputs "Mask 2" 23 ₁ - 23 _N selection blocks 2 ₁ - 2 _N determine the type of BDI expected according to the search scenario. The scheme for constructing additional RAM 13.8 is known, it can be technically implemented on the basis of a commercially available programmable dynamic random access memory with multi-bit (in our case, with L-bit) inputs and outputs, in accordance with the description presented in [Fundamentals of Electronics: Textbook for SPO / O.V. Milovzorov, I.G. Pankov. - 5th ed., revised. and additional - M.: Yurayt Publishing House, 2016. - 407 p. pp. 228-230, fig. 4.3.1].

The test RAM 13.9 of the catastrophe analysis unit 13 is designed to record, store and read in a binary code a sequence of permissible (threshold) values of the intensity of incoming BDI, as well as recording a logical zero or a logical one, characterizing, respectively, the prohibition or permission of the administrator (user, operator) to issue a signal alerts about the presence of a possible catastrophic intensity of incoming elements of the incoming data stream. Check RAM 13.9 can be technically implemented on the basis of commercially available programmable dynamic random access memory in accordance with the description presented in [Fundamentals of electronics: a textbook for SPO / O.V. Milovzorov, I.G. Pankov. - 5th ed., revised. and additional - M.: Yurayt Publishing House, 2016. - 407 p. pp. 229-231, fig. 4.3.2].

The counter 13.10 of the catastrophe analysis block 13 is designed to determine the moments of the beginning of reading in the binary code of the newly introduced control actions - new permissible (threshold) values of the intensity of the incoming elements of the incoming data stream (the intensity of the incoming BDI), as well as the moments of the beginning of reading of a logical zero or a logical unit, characterizing respectively, the prohibition or permission of the administrator (user, operator) to issue an alert signal about the presence of a possible catastrophic intensity of the incoming elements of the incoming stream. The counter 13.10 can be technically implemented on the basis of a commercially available counter with controlled reset, a description of the operation and a diagram of such a counter are given, for example, in [Ugryumov E.P. Digital circuitry: textbook. allowance for universities. - 3rd ed., revised. and additional - St. Petersburg: BHV-Petersburg, 2010. - 816 p., S. 247-248].

блока хранения маски, на фиг. 3.Mask storage units 1 ₁ - 1 _N included in the overall block diagram are identical and are designed to store bit masks used to recognize the elements of the incoming data stream. The principle of operation and the structure of the mask storage units are known, described in the prototype (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published on 06/25/2020, 2020, Bull. 18, Fig. 4), structural the scheme is illustrated, using the example of the n-th

mask storage unit, in Fig. 3.

БС, на фиг. 4.The selection blocks 2 ₁ - 2 _N included in the general block diagram are identical and are designed to control and recognize the corresponding elements of the incoming data stream, generate the search result, control the remaining search time, and also generate control signals after the set initial or newly introduced in dynamics search time management for each search query. The structure of selection blocks (BS) and their algorithm of operation are known, described in detail in the prototype (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published 06/25/2020, 2020, Bull. 18, Fig. 5), the block diagram is shown, on the example of the n-th

BS, in Fig. 4.

The frequency divider 3, included in the overall structural diagram, is designed to increase the repetition period of the pulse sequence arriving at its input. The frequency divider implementation scheme is known and described in the prototype (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published on 06/25/2020, 2020, Bull. 18). In particular, the frequency divider 3 can be implemented as a pulse duration and frequency counter based on the KR155IE6 chip, as described in [Averchenkov O.E. Circuit engineering: hardware and software. M.: DMK Press, 2012. - 588 p., S. 468-469, fig. 13.3]. In this case, the input of the divider will be the counting input of the counter, and the output of the divider will be one of the outputs of the counter.

The shaper of time intervals 4, included in the overall block diagram, is designed to control the time interval between the elements of the incoming data stream and the formation of a signal after its expiration. The structure and principle of operation of the time interval generator are known, described in detail in the prototype (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published on 06/25/2020, 2020, Bull. 18, Fig. 6), the diagram is shown in Fig. 5.

The search strategy register 5, included in the general block diagram, is designed to check whether the order of the elements of the incoming data stream (BDI type code) matches the given rules and generate a signal when an element arrives, expected in accordance with the rules. The composition of the elements and the principle of operation of the search strategy register are known, described in detail in the prototype (see RF patent No. 2724788, "Information search device" IPC G06F 9/46, published on 06/25/2020, 2020, Bull. 18, Fig. 7) , the block diagram of the register is shown in Fig. 6.

The transition mask address generating block 6, which is included in the general block diagram, is designed to generate and store a BDI type code corresponding to a mask that defines the next element to be verified in the incoming data stream. The principle of operation, the composition and interconnection of the elements of the transition mask address generation block are known, described in detail in the prototype (see RF patent No. Fig. 8), the block diagram is shown in Fig. 7.

The indication block 7, included in the general block diagram, is designed to detect signs (values of the search script parameters) indicating the completion of the sequence of elements of the incoming data stream specified by the rules (BDI type code) and the formation of the corresponding signal. The structure and operation of the display unit are known, described in the prototype (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published 06/25/2020, 2020, Bull. 18, Fig. 9), block diagram shown in Fig. 8.

СКВП, на фиг. 9.The search time selection controllers 8 ₁ - 8 _N , included in the general block diagram, are identical and are intended for decoding, additional comparison and control of the new S-bit code introduced in the dynamics of managing the process of implementing search queries, which determines the new value (boundaries) of the maximum time search for each specific search query. The structure of the selection search time controllers (SKVP) and the algorithm for their operation are known, described in detail in the prototype (see RF patent No. Fig. 10), the block diagram is shown, using the example of the n-th

SKVP, in Fig. 9.

The main search time controller 9, included in the overall structural diagram, is designed to dynamically correct the values (boundaries) of the maximum search time for each search query from any combination of N transition masks (search scenarios). The principle of operation, composition, purpose and interconnection of the elements of the main search time controller are known, described in detail in the prototype (see RF patent No. 2724788, "Information search device" IPC G06F 9/46, published 06/25/2020, 2020, Bull. 18 , Fig. 11), the block diagram is shown in Fig. 10.

The clock pulse generator 10, included in the overall block diagram, is designed to generate a synchronizing pulse sequence. The technical implementation of the clock generator 10 is possible on the basis of a commercially available clock distributor, described in [Averchenkov O.E. Circuitry: hardware and software. - M.: DMK Press, 2012. - 588 p., S. 472-473].

The analysis unit of the fuzzy search scenario 11, which is included in the general structural diagram, is intended for preliminary sequential comparison (by the number of binary numbers characterizing any of the K bits of the BDI type code) of the source data arriving in the binary code (BDI type code) and making a decision about their mathematical nature - the BDI type code is specified parametrically or using the membership function characteristic of fuzzy sets. The principle of operation, composition, purpose and interconnection of the elements of the analysis block of the fuzzy search scenario 11 are known, described in detail in the prototype (see RF patent No. 18, Fig. 2), the block diagram is shown in Fig. eleven.

The fuzzy search script conversion block 12, which is included in the general structural diagram, is designed to transform the BDI type code specified in a fuzzy form to a form suitable for obtaining clear, unambiguous (reliable) results of a specific search scenario. The algorithm of operation, the composition, purpose and relationship of the elements of the conversion block of the fuzzy search scenario 12 are known, described in detail in the prototype (see RF patent No. 18, Fig. 3), the block diagram is shown in Fig. 12.

The information retrieval device operates as follows.

The L-bit information input 04 of the device receives the elements of the incoming data stream - the analyzed blocks of binary information. At the same time, the intensity of incoming data, the intensity of incoming BDI, can smoothly change under the influence of control actions or external factors, creating a potential threat of blocking (collapse) of the process of implementing search queries (information retrieval process) for systems of this class. It is determined whether the intensity of incoming BDI exceeds (or does not exceed) the boundary (threshold) values, thereby creating the prerequisites (threat) for achieving critical (at the border of the possibilities of implementing search queries) intensity, as well as a tendency for a critical increase in this intensity. In order to determine the possible boundary and emergency (catastrophic) states that are characteristic of the emergency, critical position of the parameters of reliability and stability of the process of implementing search queries in conditions where there are smooth and slight variations in external conditions and control actions (variations in the intensity of the receipt of BDI), it is necessary to mathematically correct , using the mathematical apparatus of the theory of catastrophes, to identify and verify these states, as well as to carry out a priori warning about emergency (catastrophic) conditions for the implementation of the search process. The stages of implementation of the algorithm for identification and verification of possible boundary and emergency (catastrophic) states of complex technical systems are described in detail, algorithmically and analytically in [2, 3] and by expressions (1)-(6) of this description. At the output of the computational algorithm for identification and verification (from the point of view of catastrophe theory) of possible boundary and emergency (catastrophic) states of the system, we have not only the values of binary digits of the BDI, but also the values of the intensity of their arrival, and also, in the case of a threat of such states, signals notification of the administrator (user, operator) about a possible catastrophic state of the number of BDI.

With this in mind, in the claimed device, a reliable, stable, timely and dynamic search for information is carried out in real conditions, when there is a threat of reaching a critical (on the border of the possibility of implementing search queries) intensity of the receipt of BDI with smooth changes in the parameters of external conditions and control actions.

It is obvious that when implementing search queries, when the intensity of incoming BDIs change smoothly under the influence of control actions or external factors, not only the number of incoming BDIs per unit (period) of time objectively change over time, which predetermine a smooth drift of the search system parameters towards the boundary and catastrophic state , but also the current requirements of the administrator (user, operator) to the need for notification (warning) about its possible emergency state. Under these conditions, when a smooth increase in the number of incoming BDIs per unit (period) of time can lead to an abrupt change in the state of the search system, the reliable and stable operation of the device is hindered, there is a potential danger of its blocking, collapse.

An analysis of the works [1-5] devoted to the algorithms and principles of implementing the methods of the theory of catastrophes in the problems of analyzing complex technical processes and systems allows us to conclude that it is possible to increase the level of reliability and stability of the operation of the claimed device under conditions inherent in the real process of receipt (circulation) of BDI flows , based on the technical implementation of the procedures for identifying and verifying the states of the boundary and emergency (catastrophic) intensity of the analyzed flows, when exceeded, the information retrieval system will most likely go into an emergency state from any other state.

Building an information retrieval device based on the proposed principle of operation makes it possible to gain an advantage over the prototype, providing an increase in the reliability and stability of the operation of the claimed device under conditions inherent in the actual process of receipt (circulation) of information flows in SPD, ISPS and SIAOD, when in the dynamics of the operation of such systems the number of incoming elements of the data flow per unit (period) of time can change smoothly under the influence of control actions or external factors, creating a potential threat of their blocking (collapse).

Technical implementation of dynamic, step-by-step (at each step, i.e., at each g-th time interval (t _set + g) of observation in the interests of implementing search queries) control of the values of the boundary (catastrophic) intensity of incoming data flow elements using theory methods catastrophes in the claimed device is carried out by introducing an external dynamic control of the value of the permissible intensity (the permissible number of incoming BDI over a period of time), if the value of which is exceeded, the search system will most likely go into an emergency state from any other state (in the claimed device - input 018 "Enter threshold intensity values" of the device and the control input 136 of the catastrophe analysis block 13), by introducing an external dynamic control of the formation of a signal of a logical zero or a logical one, characterizing, respectively, the prohibition or permission of the administrator (user, operator) to issue a warning signal about the presence of a possible boundary, emergency (catastrophic) intensity of incoming BDI (in the claimed device - input 018 "Input of intensity threshold values" of the device and control input 136 of the disaster analysis block 13) and the introduction of identification and verification of the states of the boundary and emergency (catastrophic) intensity of incoming BDI, as well as the formation of signals of a logical zero or a logical unit (a signal of prediction and warning), characterizing, respectively, the absence or presence of a possible catastrophic state of the number of incoming BDIs over a period of time (in the claimed device, they are implemented within the framework of the disaster analysis block 13).

In other words, the technical implementation of the principle of increasing the reliability and stability of operation in the claimed device was carried out by introducing preliminary identification and verification of boundary and emergency (catastrophic) values of the intensity of the analyzed information flows with smooth changes in external conditions and control actions (intensity of incoming BDI), as well as timely notification (warnings) of the user (operator, system administrator) of such a search system about its possible emergency state, based on the received identification and verification data (in the claimed device, they are implemented within the framework of the disaster analysis block 13).

With this in mind, in the claimed device, search queries are implemented, where, along with the possibility of searching and recognizing information, the features of which can be specified (defined) both quantitatively and qualitatively fuzzy, with the involvement of a linguistic variable, the identification and verification of the states of the boundary and emergency (catastrophic) intensity of incoming elements of the data stream (intensity of incoming BDI), as well as controlled formation of values of permissible intensity, causing an increase in the reliability and stability of the claimed device in real conditions in which it has to function.

At the same time, as in the prototype, the claimed device implements a syntactic approach to working with images [7], based on the recognition (identification) of individual elements of the incoming data stream - BDI, by parallel analysis of the values of data features and control of their order for compliance given rules. The values of the bits in the corresponding positions of the BDI are used as recognition signs. The rules for following the BDI are given by a formal grammar - search script. To explain the parallel analysis of the values of data features and control the order of their sequence, it is necessary to consider the rules for specifying the search script. The SC search scenario can be represented by the following diagram:

Where:

- множество типов БДИ, входящих в состав сценария;Р={ _Pn },

- a set of types of BDI included in the script;

- множество масок переходов;MP={ _MPn },

- many transition masks;

MB - search script start mask;

MF - search script termination mask;

T is the waiting time for the next BDI.

Recognition of the BDI type in the device is carried out by comparing the values of the identification bits of the BDI with their reference values. Identification bits are BDI bits, the values of which make it possible to recognize (identify) the type of BDI. For each type of BDI, the set of identification bits can be individual. In this regard, each type of BDI is assigned two bit masks:

Where:

M1 _n is the first bit mask of the n-th type BDI;

M2 _n is the second bit mask of the n-th type BDI.

The bit masks contain L bits, where L is the maximum possible number of bits in the BDI used in the search script. The first bitmask is intended to indicate the positions of the identification bits of the ODI. The logical one values in the bits of the first bit mask correspond to the positions of the identification bits. All other bits of the bit mask are set to logic zero. The second bit mask is intended to set the reference values to which the values of the identification bits must correspond. In this case, the bits of the second bit mask, which are not identification, can have arbitrary values, since they do not affect the process of recognition (identification) of the BDI.

A set of transition masks is used to set the order of the BDI within the scenario. The set contains N transition masks, each of which contains N bits. Thus, each type of BDI has its own transition mask. In this case, the n-th mask of transitions MP _n contains information about the types of BDI, which, according to the search scenario, are expected after observing the BDI of the n-th type. The specified information is set by setting the value of logical zero in bits of the transition mask, the serial numbers of which correspond to the types of expected BDI. All other bits of the transition mask are set to logic one.

The search script start mask is intended to indicate the types of BDIs that are expected first in the search script - initial BDIs. The MB mask contains N bits. The bits of the mask of the beginning of the search script, the numbers of which correspond to the initial types of the BDI, are set to logical zero values. In all other digits of the mask of the beginning of the search script, the values of the logical unit are set.

The end mask of the search script is intended to indicate the types of BDI, the observation of which indicates the completion of the search script - final BDI. The MF mask contains N bits. The bits of the search script termination mask, the numbers of which correspond to the final types of the BDI, are set to logical zero values. All other bits of the search script termination mask are set to the values of a logical unit.

In contrast to the total search time, the waiting time for the next BDI T sets the maximum allowable time interval during which the next BDI is expected, specified by the type search script. If the expected BDI type is not found within the specified time interval, the search script is interrupted and the transition to waiting for the initial BDI is performed. The waiting time of the next BDI is set in the form of an M-bit code. In this case, the smallest waiting time corresponds to the largest code, which is the complement to the maximum number representable in the M-bit code.

With this in mind, the claimed device provides a controlled service of search queries and the implementation of user needs for reliable identification of the elements of the incoming data stream.

устройства соответствует множеству типов БДИ, входящих в состав сценария. При обнаружении очередного, заданного сценарием типа БДИ, устройство переходит в состояние, номер которого соответствует типу обнаруженного БДИ. Находясь в одном из состояний устройство ожидает появления БДИ, типы которых определяются маской переходов, номер которой соответствует номеру текущего состояния устройства.In general, the operation of an information retrieval device can be considered as a process of transition of this device from state to state. The set of states W={W _n },

device corresponds to the set of BDI types included in the scenario. When the next BDI type specified by the script is detected, the device switches to the state, the number of which corresponds to the detected BDI type. Being in one of the states, the device waits for the appearance of BDI, the types of which are determined by the transition mask, the number of which corresponds to the number of the current state of the device.

In FIG. 13 shows an example of a search script including N=8 BDI types. The filling of the transition masks, the script start mask and the script end mask corresponding to the given search scenario is shown in FIG. 14.

Device initialization includes the following operations:

initial device reset;

setting the first and second bit masks;

setting the script start mask and transition masks;

setting the end mask of the search script;

setting the waiting time for the next BDI;

setting the initial maximum search time;

setting the time interval (t _nab ) observation within the maximum search time in the interests of the implementation of search queries;

setting the initial threshold values for the intensity of the elements of the data flow (BDI intensity).

The initial reset of the device is carried out as follows. At the enabling input 014 of the device, the value of the logical unit is set, which is fed to the corresponding input 58 of the search strategy register 5. The logical unit at the enabling input 58 of the search strategy register 5, entering the third inputs 5.3 ₁ -3 - 5.3 _N -3 of all three-input elements OR- NOT 5.3 ₁ - 5.3 _N (see Fig. 6), ensures the presence of a logical zero at their outputs 5.3 ₁ -4 - 5.3 _N -4, regardless of the logical values \u200b\u200bset on the first 5.3 ₁ -1 - 5.3 _N -1 and the second 5.3 ₁ -2 - 5.3 _N -2 inputs. In this regard, the signal output 59 of the search strategy register 5 will be set to a logic zero value. A logical zero from the signal output 59 of the search strategy register 5 is fed to the corresponding input 45 of the time interval generator 4 (see Fig. 5) and then to the first information input 4.2-1 (input J) of the JK flip-flop 4.2. At the input "Initial reset" 05 of the device set the value of the logical unit, which is fed to the inputs 42 and 63 shaper time intervals 4 and block generating the address of the transition mask 6, respectively. The value of the logical unit from the input "Initial reset" 42 shaper time intervals 4 (see Fig. 5) through the first two-input element OR 4.1 enters the second information input 4.2-2 (input K) JK-flip-flop 4.2, and through the second two-input element OR 4.6 to the reset input (input R) of the counter 4.7. The logical unit at the input 4.2-2 (input K) of the JK flip-flop 4.2 in the presence of a logical zero at its input 4.2-1 (input J) leads to the setting of a logical zero at the output 4.2-3 of the JK flip-flop 4.2, which, acting on the second input 4.3-2 two-input element And 4.3, leads to the setting of the value of the logical zero at its output 4.3-3, regardless of the logical values at its first input 4.3-1. The logical unit at the reset input R of the counter 4.7 provides the setting of a logical zero at its overflow output P. The value of the logical zero from the overflow output R of the counter 4.7 is fed to the second input 4.5-2 of the second two-input element AND 4.5, which leads to an unconditional installation at its output 4.5- 3, and, accordingly, at the output "Reset" 44 shaper time intervals 4, the value of logical zero. The logical unit from the input "Initial reset" 63 of the block for generating the address of the transition mask 6 (see Fig. 7) through the two-input element OR 6.3 enters the reset input 6.4-3 (input R) of the register 6.4, which leads to the setting of a logical zero at its outputs 6.4-4 (Q ₁ -Q _K ), and, accordingly, on all bits of the K-bit output "Event Code" 62 of the block for generating the address of the transition mask 6, on all K test outputs 112 ₁ - 112 _K and on all bits of the K-bit transit 113 output of the fuzzy search script analysis block 11, as well as on all bits of the K-bit output 122 of the fuzzy search script conversion block 12.

Upon completion of the initial reset operation, at the input "Initial reset" 05, the device sets the value of logical zero, which leads to the setting of the value of logical zero at the second information input 4.2-2 (input K) of the JK flip-flop 4.2 and at the reset input R of the counter 4.7 of the time interval shaper 4.

The installation of the first and second bit masks, providing identification of each of the N types of BDI, is carried out in the corresponding storage blocks of the mask 1 ₁ - 1 _N . To do this, on the first L-bit inputs "Mask 1" 02 ₁ - 02 _N of the device and on the corresponding first L-bit inputs "Mask 1" 17 ₁ - 17 _N of each of the N storage blocks of the mask 1 ₁ - 1 _N , the corresponding first bit masks, and on the first L-bit inputs "Mask 2" 03 ₁ - 03 _N of the device and on the corresponding first L-bit inputs "Mask 2" 18 ₁ - 18 _N of each of the N storage blocks of the mask 1 ₁ - 1 _N , the corresponding second bit masks. At the write permission input 01 of the device and at the corresponding write permission inputs 16 ₁ - 16 _N of each of the N mask storage blocks 1 ₁ - 1 _N , the value of the logical unit is set, which is fed to the initialization inputs (inputs C) of the first and second registers 1.1 and 1.2 each of N mask storage blocks (see Fig. 3) and ensures that the first and second bit masks are written to the corresponding registers. At the end of the write operation, at the write enable input 01 of the device and at the corresponding write enable inputs 16 ₁ - 16 _N of each of the N mask storage blocks 1 ₁ - 1 _N , the value of logical zero is set.

должна быть записана по адресу, соответствующему ее порядковому номеру, то есть адресу, значение которого равно n. Для этого на управляющем входе 010 устройства и, соответственно, на управляющем входе 54 регистра стратегии поиска 5 устанавливают значение логической единицы, которая, поступая на управляющий вход 5.1-3 (вход SE) селектора-мультиплексора 5.1 (см. фиг. 6), обеспечивает коммутацию второй группы информационных входов 5.1-2 (B₁ - B_K) селектора-мультиплексора 5.1 на его выходы 5.1-4 (Q₁ - Q_K), где K=(log₂N)+1 количество двоичных разрядов, достаточное для адресации N масок переходов и маски начала сценария поиска. На K-разрядном адресном входе 09 устройства и, соответственно, на K-разрядном адресном входе 53 регистра стратегии поиска 5 устанавливают K-разрядный адрес, по которому в оперативное запоминающее устройство 5.2 должна быть записана маска начала сценария поиска. С выходов 5.1-4 (Q₁-Q_K) селектора-мультиплексора 5.1 K-разрядный адрес поступает на адресные входы 5.2-2 (A₁ - A_K) оперативного запоминающего устройства 5.2. На N-разрядном информационном входе 011 устройства и, соответственно на N-разрядном информационном входе 55 регистра стратегии поиска 5 устанавливают маску начала сценария поиска, которая поступает на информационные входы 5.2-3 (D₁-D_N) оперативного запоминающего устройства 5.2. Запись осуществляется путем установки логического нуля на входах «Выбор кристалла» 012 и «Чтение/запись» 013 устройства, с которых логический ноль через входы «Выбор кристалла» 56 и «Чтение/запись» 57 регистра стратегии поиска 5 поступает на соответствующие входы 5.2-1

и 5.2-4

оперативного запоминающего устройства 5.2. По окончании записи маски на входе «Выбор кристалла» 012 устанавливают значение логической единицы. Затем на K-разрядном адресном входе 09 устройства устанавливают K-разрядный адрес, по которому в оперативное запоминающее устройство 5.2 должна быть записана первая маска переходов МР₁ (значение адреса равно 1), а на N-разрядном информационном входе 011 устройства устанавливают маску переходов МР₁, после чего путем установки значения логического нуля на входе «Выбор кристалла» 012 инициируют операцию записи в оперативное запоминающее устройство 5.2. Аналогичным образом в оперативное запоминающее устройство записывают все N масок переходов. По окончании записи масок переходов в оперативное запоминающее устройство 5.2 на входе «Чтение/запись» 013 устанавливают значение логической единицы, а на управляющем входе 010 устройства устанавливают значение логического нуля, что обеспечивает коммутацию первой группы информационных входов 5.1-1 (A₁ - A_K) селектора-мультиплексора 5.1 на его выходы 5.1-4 (Q₁ - Q_K).The search script start mask and transition masks are set in the random access memory 5.2 of the search strategy register 5. In this case, the search script start mask MB must be written to the random access memory at address zero, and the n-th transition mask MP _n ,

must be written to the address corresponding to its serial number, that is, the address whose value is equal to n. To do this, at the control input 010 of the device and, accordingly, at the control input 54 of the search strategy register 5, the value of the logical unit is set, which, when fed to the control input 5.1-3 (input SE) of the selector-multiplexer 5.1 (see Fig. 6), provides switching the second group of information inputs 5.1-2 (B ₁ - B _K ) selector-multiplexer 5.1 to its outputs 5.1-4 (Q ₁ - Q _K ), where K=(log ₂ N)+1 the number of bits sufficient for addressing N masks of transitions and a mask of the beginning of the search script. At the K-bit address input 09 of the device and, accordingly, at the K-bit address input 53 of the search strategy register 5, a K-bit address is set at which the search script start mask should be written to the random access memory 5.2. From the outputs 5.1-4 (Q ₁ -Q _K ) selector-multiplexer 5.1 K-bit address is supplied to the address inputs 5.2-2 (A ₁ - A _K ) RAM 5.2. At the N-bit information input 011 of the device and, accordingly, at the N-bit information input 55 of the search strategy register 5, a search script start mask is set, which is fed to the information inputs 5.2-3 (D ₁ -D _N ) of the RAM 5.2. Recording is carried out by setting a logical zero at the inputs "Crystal Select" 012 and "Read/Write" 013 of the device, from which the logical zero through the inputs "Crystal Select" 56 and "Read/Write" 57 of the search strategy register 5 enters the corresponding inputs 5.2- 1

and 5.2-4

random access memory 5.2. At the end of the recording of the mask at the input "Crystal Selection" 012 set the value of a logical unit. Then, on the K-bit address input 09 of the device, a K-bit address is set, at which the first transition mask MP ₁ should be written to the random access memory 5.2 (the address value is 1), and on the N-bit information input 011 of the device, the transition mask MP is set ₁ , after which, by setting the value of a logical zero at the input "Crystal selection" 012, the write operation is initiated in the RAM 5.2. Similarly, all N transition masks are stored in random access memory. At the end of the recording of the transition masks in the random access memory 5.2, a logical unit value is set at the Read/Write input 013, and a logical zero value is set at the control input 010 of the device, which ensures switching of the first group of information inputs 5.1-1 (A ₁ - A _K ) selector-multiplexer 5.1 to its outputs 5.1-4 (Q ₁ - Q _K ).

Setting the end mask of the search script MF consists in setting the bits of the N-bit input "Search completion rule" 07 of the device and, consequently, on the bits of the N-bit input "Search completion rule" 72 of the indication block 7 (see Fig. 8) of the corresponding logical values corresponding to the values of the search script end mask bits.

Setting the waiting time for the next BDI consists in setting on the bits of the M-bit input "Waiting time code" 06 of the device and, therefore, on the bits of the M-bit input "Waiting time code" 43 of the time interval generator 4 (see Fig. 5), logical values corresponding to the values of the bits of the timeout code.

поискового запроса - набора конкретных типов БДИ (определяющего соответствующие маски переходов и характеризующего конкретный сценарий поиска).Setting the initial maximum search time consists in setting on the bits of each of the N S-bit inputs "Maximum search time correction" 016 ₁ - 016 _N of the device (see Fig. 10) through the S-bit inputs 27 ₁ - 27 _N selection blocks 2 ₁ - 2 _N to S-bit inputs 2.7 ₁ -1 - 2.7 _N -1 correction registers 2.7 ₁ - 2.7 _N logical values of the code that specifies the initial maximum search time for each n-th

search query - a set of specific types of BDI (which defines the corresponding transition masks and characterizes a specific search scenario).

Setting the observation time interval (t _nab ) within the maximum search time in the interests of implementing search queries consists in setting a certain number of clock pulses, for example, from 10 to 1000, at the clock input 138 of the catastrophe analysis block 13, for example, from 10 to 1000. pulses is the time (t _nab ) observation.

Setting the initial threshold values of the intensity of the elements of the data stream (intensity of the BDI) is to set at the input "Input threshold intensity" 018 of the device and, therefore, at the control input 136 of the disaster analysis block 13 of a certain number (λ=w/t _nab , while λ , as a rule, ranges from 1 to 1000) characterizing the number of BDI to be analyzed, arriving at the L-bit information input 04 of the device and, therefore, at the L-bit information input 130 of the disaster analysis unit 13 for the time interval t _nab .

After completing the above steps, the device is ready for use.

In the initial period, when the BDI to be analyzed is not received at the input of the device, clock pulses are received from an external generator at the input 38 of the frequency divider 3 through the first clock input 08 of the device. From the output 39 of the frequency divider 3, the clock pulses are fed to the second clock input 41 of the time interval generator 4. As a result of the operation of the initial reset of the device, all information outputs 6.4-4 (Q ₁ - Q _K ) of register 6.4, and, accordingly, on all digits K -bit output "Event code" 62 of the transition mask generation unit 6 and, accordingly, on all K test outputs 112 ₁ - 112 _K and on all bits of the K-bit transit output 113 of the fuzzy search scenario analysis unit 11, as well as on all bits K -bit output 122 of the conversion unit fuzzy search script 12, is set to a logic zero. The control input 54 of the search strategy register 5 is set to a logical zero value, which ensures switching of the logical zero values from the bits of the K-bit input "Event Code" 52 of the search strategy register 5 to the corresponding address inputs 5.2-2 (A ₁ - A _K ) of random access memory devices 5.2. Thus, the zero address is set at the address input of the random access memory, indicating the mask of the beginning of the search script.

Logical zero values from the bits of the K-bit output "Event Code" 62 of the block for generating the address of the transition mask 6 are fed to the corresponding bits of the K-bit information input 111 of the block for analyzing the fuzzy search scenario 11 (see Fig. 11). In this case, the code containing the values of logical zero in all bits is fed through the K-bit information input 111 to the K-bit input 11.2-1 of the counter 11.2 and then from the K outputs 11.2-2 ₁ - 11.2-2 _K of the counter 11.2 to the corresponding K inputs 11.1-2 ₁ - 11.1-2 _K of the storage register 11.1 of the analysis unit of the fuzzy search scenario 11, which provides transit transformation (rewriting) of the code values in the counter 11.2 and in the storage register 11.1 and setting the values of logical zero on all K test outputs 11.1-1 ₁ - 11.1-1 _{K of} the storage register 11.1 and on the corresponding K test outputs 112 ₁ - 112 _K of the fuzzy search scenario analysis block 11, as well as on all bits of the K-bit transit output 11.1-3 of the storage register 11.1 and on the corresponding bits of the K-bit transit exit 113 block 11.

Logical zero values with K test outputs 112 ₁ - 112 _K of the fuzzy search scenario analysis block 11 are fed through K test inputs 121 ₁ - 121 _K of block 12 to the corresponding K inputs 12.1-1 ₁ - 12.1-1 _K of register 12.1 (see Fig. 12), which ensures the registration and setting of logical zero values on all bits of the K-bit output 12.1-4 of the register 12.1 and on the corresponding bits of the K-bit output 122 of the fuzzy search script conversion unit 12.

будет установлено значение логической единицы. Значения логической единицы с инверсных выходов 7.1-2₁ - 7.1-2_N

дешифратора 7.1 поступает на первые входы 7.2₁-1 - 7.2_N-1 соответствующих двухвходовых элементов ИЛИ 7.2₁ - 7.2_N. В результате на выходах 7.2₁-3 - 7.2_N-3 всех двухвходовых элементов ИЛИ 7.2₁ - 7.2_N устанавливается значение логической единицы вне зависимости от логических значений на их вторых входах. Это приводит к установке значения логической единицы на выходе 7.3-2 N-входового элемента И 7.3, а соответственно и на выходе «Результат поиска» 015 устройства. Логические значения на выходе «Результат поиска» 015 устройства имеют следующее значение: логический ноль на указанном выходе означает, что во входящем потоке БДИ обнаружен заданный сценарий поиска, а логическая единица - отсутствие заданного сценария поиска. На входе «Чтение/запись» 57 регистра стратегии поиска 5 установлено значение логической единицы, что обеспечивает перевод оперативного запоминающего устройства 5.2 в режим чтения информации. При этом на входе «Выбор кристалла» 56 регистра стратегии поиска 5 также установлено значение логической единицы. На разрешающем входе 014 устройства и, соответственно на разрешающем входе 58 регистра стратегии поиска 5 установлено значение логической единицы, что обеспечивает установку на сигнальном выходе 59 регистра стратегии поиска 5 значения логического нуля. При этом на выходе 4.2-3 JK-триггера 4.2 формирователя временных интервалов 4 (см. фиг. 5), на его первом 4.2-1 (вход J) и втором 4.2-2 (вход K) информационных входах установлено значение логического нуля.Logical zero values from the bits of the K-bit transit output 113 of the fuzzy search scenario analysis unit 11 and from the corresponding bits of the K-bit output 122 of the fuzzy search scenario conversion unit 12 are fed to the corresponding bits of the "Event Code" input 71 of the display unit 7 and then to the inputs 7.1-1 ₁ - 7.1-1 _K (F ₁ - F _K ) decoder 7.1 (see Fig. 8). If there is a logical zero value at all inputs 7.1-1 ₁ - 7.1-1 _K (F ₁ - F _K ) of the decoder 7.1, at all its inverse outputs 7.1-2 ₁ - 7.1-2 _N

will be set to a logical unit. Logic one values from inverted outputs 7.1-2 ₁ - 7.1-2 _N

decoder 7.1 is supplied to the first inputs 7.2 ₁ -1 - 7.2 _N -1 of the corresponding two-input elements OR 7.2 ₁ - 7.2 _N . As a result, the outputs 7.2 ₁ -3 - 7.2 _N -3 of all two-input elements OR 7.2 ₁ - 7.2 _N is set to the value of a logical unit, regardless of the logical values at their second inputs. This leads to setting the value of a logical unit at the output 7.3-2 of the N-input element And 7.3, and, accordingly, at the output of the "Search result" 015 device. The logical values at the output "Search result" 015 of the device have the following meaning: a logical zero at the indicated output means that the specified search script was found in the incoming stream of the BDI, and a logical one means the absence of the specified search script. At the input "Read/Write" 57 of the search strategy register 5 is set to a logical unit, which ensures the transfer of random access memory 5.2 in read mode. At the same time, at the input "Crystal Selection" 56 of the search strategy register 5, the value of a logical unit is also set. At the enabling input 014 of the device and, accordingly, at the enabling input 58 of the search strategy register 5, the value of the logical unit is set, which ensures that the signal output 59 of the search strategy register 5 is set to a logical zero value. At the same time, at the output 4.2-3 of the JK flip-flop 4.2 of the time interval generator 4 (see Fig. 5), on its first 4.2-1 (input J) and second 4.2-2 (input K) information inputs, the value of logical zero is set.

Further stages of the implementation of search queries, including the stage of identification and verification of the states of the boundary and emergency (catastrophic) intensity of incoming BDI, as well as the controlled formation of values of the permissible intensity, are implemented as follows. The investigated input stream, i.e., the stream of BDI to be analyzed, entering the L-bit information input 04 of the device and the L-bit information input 130 of the disaster analysis unit 13, sets logical values corresponding to the values of the binary bits of the BDI. From the L-bit information input 130, the values of the binary digits of the BDI are fed to the L-bit information input 13.0-1 of the intensity calculator 13.0 of the catastrophe analysis unit 13 (see Fig. 2).

The logical values corresponding to the values of the binary digits of the BDI are recorded in the memory cells of the intensity calculator 13.0 of the disaster analysis unit 13 (technically implemented as a programmable summing counter with a RAM function and controlled reset), and the number of these incoming binary digits of the BDI, in fact, is the initial data - the number w, which characterizes the number of elements of the incoming data stream (the number of BDIs to be analyzed) coming as part of the search for information at the input of the device and used in the intensity calculator 13.0 of the disaster analysis unit 13 to calculate the intensity of incoming BDIs when implementing search queries.

Thus, in the intensity calculator 13.0 of the disaster analysis block 13, the values of the intensity of incoming BDI (λ=w/t _nab ), i.e., the number w of incoming elements of the incoming data stream for the time interval t _nab , are calculated, and these values are taken into account in the interests of to control the drift of this number λ towards the boundary and catastrophic state. After the implementation in the disaster analysis unit 13 procedures for identifying and verifying the states of the boundary and emergency (catastrophic) intensity, through the L-bit information output 131 of the disaster analysis unit 13, the identified and verified (from the point of view of the catastrophe theory) values of the BDI binary digits are sent to the L-bit information inputs 21 ₁ - 21 _N selection blocks 2 ₁ - 2 _N .

At the same time, dynamic control of the permissible intensity value, when exceeded, the information retrieval system will most likely go into an emergency state from any other state, identification and verification of the states of the boundary (catastrophic) intensity, as well as the formation of signals of a logical zero or a logical unit (signal of prediction and warning ), characterizing respectively the absence or presence of a possible catastrophic state of intensity, is implemented within the framework of the catastrophe analysis block 13 as follows.

Formation of the control code sequence (consisting of elements of the set Λ _G (see expression (1), where Λ _g - admissible (threshold, maximum), and λ _g - real for each g-th (g=1,2,…,G ) successive time interval t nab _{of observation} , the intensity value, when exceeded, the information retrieval system will most likely go into an emergency state from any other state, as well as the formation of a logical zero or logical one signal, characterizing, respectively, the prohibition or permission of the administrator (user, operator) to issue an alert signal about the presence of a possible catastrophic state of the number of incoming messages per unit (period) of time (i.e., intensity), is carried out using the test RAM 13.9 and the counter 13.10 of the disaster analysis unit (see Fig. 2).

In this case, the formation of the control code sequence of permissible intensity values, as well as the formation of a signal of a logical zero or a logical unit, is performed as follows. From an external source through the input "Input of threshold values of intensity" 018 of the device, the control input 136 of the disaster analysis unit and the control input 13.9-4 in the memory cells of the test RAM 13.9 (see Fig. 2) is sequentially written in binary code of a set of valid (threshold) intensity values for each g-th (g=1,2,…,G) successive time interval t _{nab of} observation, as well as a record of a logical zero or a logical one, characterizing, respectively, the prohibition or permission to issue a warning signal about the presence of a possible catastrophic states.

Counts (t _set ), (t _set +1), (t _set +2), etc. steps (cycles) of observation within the framework of the cycle of implementation of search requests by the claimed device, and in fact, steps (cycles) of checking (controlling) the compliance of the permissible values of the intensity of the receipt of BDI with their real value, come from the output 101 of the clock generator 10, in addition to the clock inputs 25 ₁ - 25 _N selection blocks 2 ₁ - 2 _N , to the clock input 138 of the disaster analysis block 13, being the reference steps within the search query implementation cycle. These signals (counts) are received through the clock input 138 of the catastrophe analysis block 13 to the clock input 13.10-3 of the counter 13.10 (see Fig. 2) and are determined by coming from the clock output 13.10-1 of the counter 13.10 to the clock input 13.9-2 of the test RAM 13.9 , the start of sequential reading in binary code stored in the test RAM 13.9 Λ _G - a set of valid (threshold) intensity values for each g-th (g=1,2,...,G) successive time interval t _nab observation. Sequential reading of the allowable (threshold) intensity values for each g-th (g=1,2,...,G) successive time interval t _nab observation is carried out through the test output 13.9-1 of the test RAM 13.9 to the test input 13.2-2 of the ROM 13.2 block analysis of catastrophes 13 (Fig. 2), and also determines the start of reading stored in the test RAM 13.9 logical zero or logical unit. The reading of a logical zero or a logical one is also performed from the test output 13.9-1 of the test RAM 13.9 to the second input 13.7-2 of the element And 13.7 of the catastrophe analysis block 13 (Fig. 2). From the release output 13.9-3 of the test RAM 13.9 to the release input 13.10-2 of the counter 13.10 at the time of reading the sequence of allowable (threshold) intensity values and reading the logical zero (logical one), a signal is received that clears (frees, resets) the values of the counter 13.10 and gives the command counter 13.10 to start a new countdown for the newly introduced control actions (Λ _G - sequences of allowable (threshold) intensity values for each g-th time interval), and for the newly introduced control actions - logical zero (logical one).

The procedures for identifying and verifying the states of the boundary, emergency (catastrophic) intensity of incoming BDI with smooth changes in the parameters of external conditions and control actions, as well as for generating signals of a logical zero or a logical one (a prediction and warning signal), characterizing, respectively, the absence or presence of a possible catastrophic state are implemented in the disaster analysis block 13 (which can be implemented according to the scheme shown in Fig. 2), as follows.

Calculated in the intensity calculator 13.0 of the disaster analysis block 13, the intensity values of the incoming BDI λ are recorded in the memory cells of the calculator 13.0, which contains the values of λ at each step, i.e., at each g-th time interval (t _nab +g) of observation in the interests of implementation of search queries by the device.

From the output 101 of the clock generator 10 through the clock input 138 of the catastrophe analysis unit 13 to the clock input 13.0-3 of the intensity calculator 13.0 and to the clock input 13.2-1 of the ROM 13.2, a clock signal is received, initiating a sequential reading from the memory cells of the intensity calculator 13.0 of the intensity λ values of the BDI for each g-th (g=1,2,…,G) successive time interval t _{nab of} observation, characterizing (predetermining) the existence or absence of a smooth drift of intensity values towards the boundary and catastrophic state of the system, namely those values which correspond to set (1) and have the physical meaning of exceeding the threshold of the system's performance capabilities and, as a result, a high probability of the information retrieval system transitioning into an emergency, catastrophic state.

The intensity λ values of the analyzed BDI for each next, next (t _nab +1)-th observation time interval are sequentially read from each g-th of the corresponding executive outputs 13.0-4 ₁ - 13.0-4 _G of the intensity calculator 13.0 to the inputs of the corresponding g-th execution RAM 13.1 ₁ - 13.1 _G , as well as the first input 13.3-1 of the iterative comparison element 13.3 and the second input 13.4-2 of the comparison element 13.4 (see Fig. 2). In addition, from the output of each g-th of the executive RAM 13.1 ₁ - 13.1 _G to the second input 13.3-2 element of iterative comparison 13.3 in binary code are sequentially read stored in the memory cells of the value λ of the intensity of the BDI on (t _nab ), the previous time interval of observation .

In addition, the clock signals coming from the clock input 138 of block 13 to the clock input 13.2-1 of the ROM 13.2 initiate reading in binary code from the memory cells of the ROM 13.2 to the first input 13.4-1 of the comparison element 13.4 of a pre-recorded allowable intensity value of the incoming BDI for any g-th observation time interval, the value, when exceeded, in accordance with expression (2), the information retrieval system with a high probability will switch to an emergency (catastrophic) state from any other state. Based on the comparison element 13.4, the procedure for identifying the states of the boundary, emergency (catastrophic) intensity of incoming and analyzed BDI is carried out with smooth changes in the parameters of external conditions and control actions. This procedure is carried out by successive step-by-step (at each step, i.e., each g-th time interval (t _set + g) of observation) a priori analysis (estimation) and comparison of the values of each λ _g of the g-th element of the set Λ _g in order to determining the presence or absence of a possible excess of these values of the allowable threshold for the intensity of the intake of BDI in accordance with expression (2).

Thus, the comparison element 13.4 performs a step-by-step comparison of the sequentially received in binary code from the executive outputs 13.0-4 ₁ - 13.0-4 _G of the intensity calculator 13.0 values of each λ _g (t _set +g), with the value of Λ _g received from the output of the ROM 13.2 . Not exceeding any g-th λ _g (t _nab +g) value of this threshold Λ _g characterizes the non-fulfillment of condition (5) and the result of the identification (comparison) procedure is expressed in the appearance of a logical zero at the output of the comparison element 13.4, which is received in binary code on the first input 13.7-1 element And 13.7. The excess of any g-th λ _g (t _nab + g) value of this threshold Λ _g characterizes the fulfillment of condition (5) and the beginning of a smooth change in the intensity values of incoming BDI for this g-th time interval towards the boundary and catastrophic state, and the result of the procedure identification (comparison) is expressed in the appearance at the output of the comparison element 13.4 of a logical unit, which in binary code is fed to the first input 13.7-1 element And 13.7.

On the basis of the iterative comparison element 13.3, the intermediate RAM 13.5 and the intermediate element And 13.6, the procedure for verifying the boundary and emergency (catastrophic) states of the intensity of the incoming BDI is carried out. This procedure is carried out in two stages. The first stage is carried out on the basis of the iterative comparison element 13.3 by successive step-by-step a priori estimation and comparison of the value λ of the intensity of incoming BDI for each previous (t _nab )-th observation time interval with the value λ of the intensity of incoming BDI at (t _nab +1) the next, next time interval observation interval in accordance with expression (4). For this purpose, in the iterative comparison element 13.3, a step-by-step comparison is made in sequence and in pairs in binary code from the outputs G of the executive RAM 13.1 ₁ - 13.1 _G of each of the intensity values at a given (t _nab )-th step with the corresponding intensity values of the BDI at the next (t _Nab +1)-th step, coming from the executive outputs 13.0-4 ₁ - 13.0-4 _G of the intensity calculator 13.0. Any g-th value λ of the BDI intensity at the (t _nab +1)-th step does not exceed the corresponding g-th value λ of the BDI intensity at the (t _nab )-th (previous) step characterizes the non-fulfillment of condition (6) and the result of the first stage of the verification procedure (comparison of the subsequent value of the BDI intensity with the previous one) is expressed in the appearance of a logical zero at the output of the iterative comparison element 13.3, which in binary code enters the input of the intermediate RAM 13.5 and the second input 13.6-2 of the intermediate element AND 13.6. The excess of any g-th value of λ of the intensity of the BDI at the (t _nab +1)-th step of the corresponding g-th value of λ of the intensity of the BDI at the (t _nab )-th (previous) step, characterizes the fulfillment of condition (6), has the physical meaning of the identified tendency to increase the probability of changing the intensity values of the BDI towards the boundary and emergency (catastrophic) state, and the result of the first stage of the verification procedure (comparison of the subsequent intensity value with the previous one) is expressed in the appearance at the output of the iterative comparison element 13.3 of a logical unit, which in binary code is input intermediate RAM 13.5 and the second input 13.6-2 intermediate element And 13.6.

The second stage of the verification procedure in block 13 (see Fig. 2) is carried out on the basis of the intermediate RAM 13.5 and the intermediate element And 13.6 by comparing in the intermediate element And 13.6 obtained at (t _nab )-th, the previous step of the result of the first stage (logical zero or logical unit) stored in the intermediate RAM 13.5 and supplied to the first input 13.6-1 of the intermediate element AND 13.6, with the result of the first stage (logical zero or logical one) obtained at (t _nab +1) th, the next step of observation and incoming from the output of the iterative comparison element 13.3 to the second input 13.6-2 of the intermediate element And 13.6. The physical meaning of the second stage of verification is to confirm (verification itself) the identified trend of increasing the probability of changing the intensity of incoming BDI towards the boundary and emergency (catastrophic) state, and the result of the second stage of verification is expressed in the appearance of a logical unit or logical zero at the output of the intermediate element AND 13.6, which in binary code arrive at the first input 13.7-1 element And 13.7. If the result of the first stage (for example, logical one) obtained at the (t _nab )-th previous step coincides with the result of the first stage (logical one) obtained at the (t _nab +1)-th next observation step, the result of the second stage of the procedure verification is expressed in the appearance at the output of the intermediate element AND 13.6 logical unit. If the result of the first stage (for example, logical one) obtained at the (t _nab )-th previous step does not match the result of the first stage (logical one) obtained at the (t _nab +1)-th cycle, or obtained at the (t _nab )-th, previous step, the result of the first stage is a logical zero and coincides with the result of the first stage (logical zero) obtained at the (t _nab +1)-th, next observation step, the result of the second stage of the verification procedure is expressed in the appearance of an intermediate element AND 13.6 logical zero.

The signals characterizing the type of BDI and determined by the first and second bit masks come from the second L-bit outputs "Mask 1" 14 ₁ -14 _N and the second L-bit outputs "Mask 2" 15 ₁ - 15 _N N of the corresponding mask storage units 1 ₁ -1 _N through N corresponding second L-bit inputs "Mask 1" 132 ₁ - 132 _N and N corresponding second L-bit inputs "Mask 2" 133 ₁ - 133 _N to N corresponding second L-bit inputs "Mask 1" 13.8-2 ₁ - 13.8-2 _N and N corresponding second L-bit inputs "Mask 2" 13.8-3 ₁ - 13.8-3 _N additional RAM 13.8 of the disaster analysis unit 13. In the additional RAM 13.8 of the disaster analysis unit 13, these signals are stored until the moment of completion of the procedures for identification and verification (from the point of view of the theory of catastrophes) of the values of the intensity of the arrival of the elements of the data stream (the intensity of those arriving for the analysis of the BDI) and the receipt of a logical zero or a logical one from the output of the AND element 13.7 to the reading input 13.8-1 of the additional RAM 13.8.

The clock signal (see Fig. 2) arriving at the clock input 138 of the catastrophe analysis block 13 initiates reading from the test output 13.9-1 of the test RAM 13.9 to the second input 13.7-2 of the element AND 13.7 of a logical zero or a logical unit, characterizing respectively the prohibition or permission of the administrator (user, operator) to issue an alert signal about the presence of a possible catastrophic state of the intensity of the incoming elements of the data stream (incoming for analysis of the BDI).

If the second input 13.7-2 element AND 13.7 received a logical unit (notification of a possible catastrophic state of intensity) and a logical unit received on the first input 13.7-1 element And 13.7 from the output of the comparison element 13.4 or from the output of the intermediate element And 13.6, the result of identification and verification of boundary and emergency (catastrophic) states of the intensity of the incoming BDI is expressed in the appearance of a logical unit at the output of the element AND 13.7. The logical unit from the output of the element And 13.7 is fed to the reading input 13.0-5 of the intensity calculator 13.0 and locks the intensity calculator 13.0, not allowing the signals characterizing the identified and verified values of the BDI binary digits to be read from the L-bit information output 13.0-2 in a parallel code through L-bit information output 131 block 13 L-bit information inputs 21 ₁ - 21 _N selection blocks 2 ₁ - 2 _N . In addition, the logical zero from the output of the AND element 13.7 is fed to the reading input 13.8-1 of the additional RAM 13.8 and locks the additional RAM 13.8, not allowing the signals characterizing the type of BDI, determined by the first and second bit masks, to be read from N L-bit outputs "Mask 1" 13.8-4 ₁ - 13.8-4 _N and N L-bit outputs "Mask 2" 13.8-5 ₁ - 13.8-5 _N additional RAM 13.8 through the corresponding N second L-bit outputs "Mask 1" 134 ₁ - 134 _N and the corresponding N second L-bit outputs "Mask 2" 135 ₁ - 135 _N of the catastrophe analysis unit 13 to the second L-bit inputs "Mask 1" 22 ₁ - 22 _N and the second L-bit outputs "Mask 2" 23 ₁ - 23 _N N corresponding selection blocks 2 ₁ - 2 _N .

In addition, the logical unit from the output of the element AND 13.7 through the warning output 137 of the block 13 goes to the output "Threat of a catastrophe" 019 of the device, being a prediction and warning signal characterizing the presence of a possible catastrophic state of the intensity of the arrival of the elements of the data flow (incoming for the analysis of the BDI) and calling administrator (user, operator) to make an external correction of the intensity (for example, introduce a temporary ban or forced restriction on the receipt of BDI) or threshold, permissible values for the intensity of receipt of data flow elements (incoming for BDI analysis) in order to prevent an emergency (catastrophic) abrupt change in state information retrieval systems under small perturbations caused by external or internal influences [2].

If a logical zero is received from the test output 13.9-1 of the test RAM 13.9 to the second input 13.7-2 of the element And 13.7 of block 13 (see Fig. 2) (the administrator does not need to be notified of a possible catastrophic condition), and from the output of the comparison element 13.4 or from the output of the intermediate element AND 13.6, either a logical zero or a logical unit is received, the result of the identification and verification of the boundary and emergency (catastrophic) states of the intensity of the incoming elements of the data flow (the intensity of the input for the analysis of the BDI) is expressed in the appearance of a logical zero at the output of the element AND 13.7. If a logical unit is received from the test output 13.9-1 of the test RAM 13.9 to the second input 13.7-2 of the element AND 13.7 (notification of a possible catastrophic condition is necessary), and a logical zero is received from the output of the comparison element 13.4 or from the output of the intermediate element AND 13.6, the result identification and verification of the boundary and emergency (catastrophic) intensity of receipt of elements of the data flow (intensity of incoming for analysis of the BDI) is expressed in the appearance of a logical zero at the output of the element AND 13.7. The logical zero from the output of the AND element 13.7 is fed to the reading input 13.0-5 of the intensity calculator 13.0 and opens the intensity calculator 13.0, allowing the signals characterizing the identified and verified values of the BDI binary digits to be read from the L-bit information output 13.0-2 in a parallel code through L -bit information output 131 block 13 to L-bit information inputs 21 ₁ - 21 _N selection blocks 2 ₁ - 2 _N . In addition, the logical zero from the output of the AND element 13.7 is fed to the reading input 13.8-1 of the additional RAM 13.8 and opens the additional RAM 13.8, allowing the signals characterizing the type of BDI, determined by the first and second bit masks, to be read from N L-bit outputs "Mask 1 » 13.8-4 ₁ - 13.8-4 _N and N L-bit outputs "Mask 2" 13.8-5 ₁ - 13.8-5 _N additional RAM 13.8 through the corresponding N second L-bit outputs "Mask 1" 134 ₁ - 134 _N and the corresponding N second L-bit outputs "Mask 2" 135 ₁ - 135 _N of the catastrophe analysis unit 13 to the second L-bit inputs "Mask 1" 22 ₁ - 22 _N and the second L-bit outputs "Mask 2" 23 ₁ - 23 _N N corresponding selection blocks 2 ₁ - 2 _N .

In addition, the logical zero from the output of the element AND 13.7 through the warning output 137 of the block 13 goes to the output "Threat of a catastrophe" 019 of the device, being a prediction and warning signal characterizing the absence of a possible catastrophic state of the intensity of the arrival of the elements of the data flow (the intensity of the BDI arriving for analysis). Thus, the presence of a logical unit signal at the warning output 137 of the catastrophe analysis block 13 and at the output "Threat of a catastrophe" 019 of the device serves as an indicator of blocking the information search device, and the presence of a warning output 137 of the catastrophe analysis block 13 and at the output "Threat of a catastrophe" 019 of the device a logic zero signal serves as an indicator of unlocking the information retrieval device as a whole.

Thus, the signals characterizing the identified and verified (from the point of view of checking the intensity λ by the methods of catastrophe theory) values of the BDI binary digits come from the L-bit information output 131 of block 13 to the L-bit information inputs 21 ₁ - 21 _N selection blocks 2 ₁ - 2 _N , and the signals characterizing the type of BDI, determined by the first and second bit masks, come from the corresponding N second L-bit outputs "Mask 1" 134 ₁ - 134 _N and the corresponding N second L-bit outputs "Mask 2" 135 ₁ - 135 _N catastrophe analysis unit 13 to the second L-bit inputs "Mask 1" 22 ₁ - 22 _N and the second L-bit outputs "Mask 2" 23 ₁ - 23 _N N corresponding selection blocks 2 ₁ - 2 _N .

The moment of time corresponding to the installation of the identified and verified (from the point of view of the theory of catastrophes) values of the binary digits of the BDI at the L-bit information output 131 of the catastrophe analysis block 13, denoted as T ₁ . From the L-bit information output 131 of the catastrophe analysis block 13, the identified and verified (from the point of view of the catastrophe theory) values of the binary digits of the BDI are fed to the L-bit information inputs 21 ₁ - 21 _N of the selection blocks 2 ₁ - 2 _N . Each selection block identifies the BDI of the corresponding type.

After 13 procedures for identifying and verifying boundary and emergency (catastrophic) states implemented in the disaster analysis unit, which are characteristic of the emergency, critical position of the reliability and stability parameters of the search query implementation process, the type of incoming BDI is determined in the information retrieval device. The type of BDI is determined by the first and second bit masks coming from the corresponding N second L-bit outputs "Mask 1" 134 ₁ - 134 _N and the corresponding N second L-bit outputs "Mask 2" 135 ₁ - 135 _N of the catastrophe analysis block 13 to the second L-bit inputs "Mask 1" 22 ₁ - 22 _N and the second L-bit outputs "Mask 2" 23 ₁ - 23 _N N corresponding selection blocks 2 ₁ - 2 _N . Comparators 2.3 of each selection block 2 ₁ - 2 _N (see Fig. 4) compares the values of the identification bits of the received BDI with the values of the corresponding bits of the second bit mask. Identification bits are selected in the first and second groups of two-input elements AND 2.1 ₁ -2.1 _L , 2.2 ₁ - 2.2 _L of each selection block based on the corresponding first bit mask. If the compared values are equal, the output of the equality "A=B" 2.3-3 of the comparator 2.3 will set the value of a logical unit, otherwise the value of logical zero. The logical value corresponding to the comparison result from the output "A=B" 2.3-3 of the comparator 2.3 in the case when only the initial maximum search time is set and there is no external dynamic control of the search time (there are no signals at the bits of each of the N S-bit inputs "Correction maximum search time" 016 ₁ - 016 _N of the device), is inverted by the inverter 2.4 and fed to the output "Comparison result" 26 of the selection block 2.

In other words, if during the analysis of incoming identified and verified (from the point of view of catastrophe theory) BDI there is no external dynamic control of the search time for all N search queries, at the S-bit corrective inputs 27 ₁ - 27 _N selection blocks 2 ₁ - 2 _N , and hence the S-bit inputs 2.7 ₁ -1 - 2.7 _N -1 corrective registers 2.7 ₁ - 2.7 _N code signals are missing. In this case, corrective registers 2.7 ₁ - 2.7 _N selection blocks 2 ₁ - 2 _N (see Fig. 4) determine the initial search time codes as uncorrected and broadcast (overwrite) them each through their S outputs (2.7 ₁ -2 ₁ - 2.7 ₁ -2 _S ) - (2.7 _N -2 ₁ - 2.7 _N -2 _S ) to the corresponding S information inputs (D ₁ -D _S ) of the respective counters 2.5 ₁ - 2.5 _N selection blocks 2 ₁ - 2 _N .

поискового запроса набора конкретных типов БДИ, с внешнего устройства в S-разрядном коде (либо с помощью человека-оператора, либо с помощью специального управляющего устройства), через N S-разрядных входов «Коррекция максимального времени поиска» 016₁ - 016_N устройства на N S-разрядных входов 91₁ - 91_N главного контроллера времени поиска 9 (см. фиг. 10) поступают новые, дополнительно вводимые в динамике управления поиском, значения максимального времени поиска для конкретных поисковых запросов абонентов информационно-справочной (поисковой) системы.If, during the analysis of the elements of the incoming data stream, external dynamic search time control is initiated for any nth

search request for a set of specific types of BDI, from an external device in the S-bit code (either with the help of a human operator or with the help of a special control device), through N S-bit inputs "Correction of the maximum search time" 016 ₁ - 016 _N of the device on N S-bit inputs 91 ₁ - 91 _N of the main search time controller 9 (see Fig. 10) receive new, additionally entered in the search control dynamics, values of the maximum search time for specific search queries of subscribers of the information and reference (search) system.

Additionally entered in the search control dynamics, the values of the maximum search time for specific search queries of subscribers, in the S-bit code, enters through N S-bit inputs 91 ₁ - 91 _N of the main search time controller 9 (see Fig. 10) to N S- bit inputs 9.1-1 ₁ - 9.1-1 _N recording element of the search time 9.1 for control and registration. With N S-bit outputs 9.1-2 ₁ - 9.1-2 _N registering element of the search time 9.1 new values of the maximum search time are fed to the corresponding N S-bit inputs 9.2-1 ₁ - 9.2-1 _N storage element of the new value of the search time 9.2, which records and stores these values in the S-bit code until the next control action is introduced, as well as from its N S-bit outputs 9.2-2 ₁ - 9.2-2 _N , through the corresponding N S-bit outputs 92 ₁ - 92 _N of the main the search time controller 9 transmits these new values of the maximum search time to the corrective inputs 27 ₁ - 27 _N of the respective selection blocks 2 ₁ -2 _N and to the test inputs 82 ₁ -82 _N of the corresponding selection search time controllers 8 ₁ - 8 _N .

At the same time, at the S-bit corrective inputs 27 ₁ - 27 _N of the selection blocks 2 ₁ - 2 _N , and hence at the S-bit inputs 2.7 ₁ -1 - 2.7 _N -1 of the corrective registers 2.7 ₁ - 2.7 _N , there are S-bit code signals. Correction registers 2.7 ₁ - 2.7 _N selection blocks 2 ₁ - 2 _N (see Fig. 4) register the original code (recorded when preparing the device for operation, i.e. the initial maximum search time) and preliminarily compare it with the newly entered dynamic control S-bit code that comes through the corrective inputs 27 ₁ - 27 _N selection blocks 2 ₁ - 2 _N S-bit inputs 2.7 ₁ -1 - 2.7 _N -1 corrective registers 2.7 ₁ - 2.7 _N .

блока селекции 2_n. Если на S-разрядном входе 2.7_n-1 корректирующего регистра 2.7_n нет S-разрядного сигнала, обуславливающего новое, вводимое в динамике управления максимальное время поиска, то приоритетными признаются ранее записанные значения кода, задающего начальное максимальное время поиска.Correction (formation based on the results of a preliminary comparison) at S outputs 2.7 _n -2 ₁ - 2.7 _n -2 _S , for example, a corrective register 2.7 _n of a selection block 2 _n of a code characterizing a preliminary decision on the value of the maximum search time for each specific n-th query carried out as follows (see Fig. 4). If the S-bit input 2.7 _n -1 of the correction register 2.7 _n has an S-bit signal that determines the new maximum search time entered in the control dynamics, this signal is classified as a priority, and it is he with S outputs 2.7 _n -2 ₁ - 2.7 _n -2 _S correction register 2.7 _n is supplied to the corresponding S information inputs (D ₁ - D _S ) counter 2.5 _n n-th

selection block 2 _n . If the S-bit input 2.7 _n -1 of the correction register 2.7 _n does not have an S-bit signal that determines the new maximum search time entered in the control dynamics, then the previously recorded code values that specify the initial maximum search time are recognized as priority.

блока селекции 2_n либо изначально введенное, либо новое, вводимое в динамике управления процессом анализа элементов входящего потока данных, значение кода, задающие максимальное время поиска. Тем самым обеспечивается инициализация счетчиков 2.5₁ - 2.5_N блоков селекции 2₁ - 2_N. Причем наименьшему времени поиска соответствует наибольший код, являющийся дополнением до максимального числа, представимого в S-разрядном коде.Thus, with S outputs 2.7 _n -2 ₁ -2.7 _n -2 _S corrective register 2.7 _n is supplied to the corresponding S information inputs (D ₁ - D _S ) counter 2.5 _n n-th

selection block 2 _n either initially introduced, or new, introduced in the dynamics of the process control of the analysis of the elements of the incoming data stream, the value of the code that specifies the maximum search time. This ensures the initialization of the counters 2.5 ₁ - 2.5 _N selection blocks 2 ₁ - 2 _N . Moreover, the smallest search time corresponds to the largest code, which is the complement to the maximum number representable in the S-bit code.

In the initial period, when the BDI to be analyzed are not fed to the input of the device, the logical value corresponding to the results of the comparison is absent at the output "A=B" 2.3 _n -3 of the comparator 2.3 _n . The three-input elements And 2.6 ₁ - 2.6 _N of all selection blocks are closed, the clock pulses from the clock generator 10 through the three-input elements And 2.6 ₁ - 2.6 _N to the counting inputs Z of the counters 2.5 ₁ - 2.5 _N of the selection blocks 2 ₁ - 2 _N are not received. At the outputs "Comparison result" 26 ₁ - 26 _N selection blocks 2 ₁ - 2 _N signal is missing.

счетчиков 2.5₁ - 2.5_N через установленный интервал времени, определяемый кодами начального заполнения счетчиков и частотой тактовых импульсов.If the logical values corresponding to the comparison results are received from the outputs "A=B" 2.3 ₁ -3 - 2.3 _N -3 of the comparators 2.3 ₁ - 2.3 _N of the selection blocks 2 ₁ - 2 _N , at the outputs 26 ₁ - 26 _N of these blocks through the inverters 2.4 ₁ - 2.4 _N pre-set low level signals. On the counting inputs Z counters 2.5 ₁ - 2.5 _N selection blocks 2 ₁ - 2 _N , pulses are received from the output 101 of the clock generator 10 through the circuit: clock inputs 25 ₁ - 25 _N selection blocks 2 ₁ - 2 _N , open three-input elements AND 2.6 ₁ - 2.6 _N selection blocks 2 ₁ - 2 _N . Counters 2.5 ₁ - 2.5 _N of each selection block perform the function of timers that control the expiration of the allowable time (initial or entered within the control) of the search by summing the clock pulses received at their counting input Z and form an overflow signal at the inverse outputs

counters 2.5 ₁ - 2.5 _N after a set time interval, determined by the codes of the initial filling of the counters and the frequency of clock pulses.

The signals from the outputs "Comparison result" 26 ₁ - 26 _N selection blocks 2 ₁ - 2 _N are fed through the inputs "Comparison result" 81 ₁ - 81 _N selection search time controllers 8 ₁ - 8 _N (see Fig. 9) to the signal inputs 8.2 ₁ -1 - 8.2 _N -1 registers of comparison-correction of the maximum search time 8.2 ₁ - 8.2 _N , to test inputs 8.2 ₁ -2 - 8.2 _N -2 which are received in binary code from test outputs 8.1 ₁ -2 - 8.1 _N - 2 decoders of the corrected code of the maximum search time 8.1 ₁ - 8.1 _N signals characterizing the new value (boundaries) of the maximum search time entered in the control process for a specific request. In this case, the decoders of the corrected code of the maximum search time 8.1 ₁ - 8.1 _N convert the S-bit code that determines the new values (boundaries) of the maximum search time entered in the control process into a binary code and transmit this code for verification of truth to the comparison-correction registers of the maximum search time 8.2 ₁ - 8.2 _N .

The comparison-correction registers of the maximum search time 8.2 ₁ - 8.2 _N carry out an additional check (comparison) of meeting the timeliness requirements in accordance with the initial and newly introduced maximum search time and, acting as relay nodes, form 8.2 ₁ -3 - 8.2 at their outputs _N -3 and at the outputs "Comparison result" 83 ₁ - 83 _N selective search time controllers 8 ₁ - 8 _N sequence of signals describing the logical values corresponding to the comparison results.

Thus, at the outputs "Comparison result" 83 ₁ - 83 _N selective search time controllers 8 ₁ - 8 _N we have logical values corresponding to the comparison results and obtained taking into account the dynamic correction of the maximum search time.

After satisfying the need to search for information, the n-th search request is removed, there is no signal at the output "A \u003d B" 2.3 _n -3 of the comparator 2.3 _n , the counter 2.5 _n of the corresponding selection block 2 _n is reset at the n-th input "Zeroing" 017 _n device and the input "Reset" 24 _n of the corresponding selection block 2 _n .

а, следовательно и на выходах «Результат сравнения» 26₁ - 26_N соответствующих блоков селекции 2₁ - 2_N сигнала переполнения, инициирующего блокирование выхода 2.4_n-2 инвертора 2.4_n и разовое поступление с данного инверсного выхода логических значений, соответствующих результатам сравнения на данный, конкретный момент времени, без возможности продолжения поиска, что соответствует процедуре динамического управления поиском - переносу поисковых запросов заново в очередь на места, соответствующие их приоритетам.If one or more search queries have reached the maximum search time or the maximum search time has decreased as a result of dynamic correction (control), the counters 2.5 ₁ - 2.5 _N of the corresponding selection blocks 2 ₁ - 2 _N are overflowed, and an overflow is formed on their inverse outputs

and, consequently, at the outputs "Comparison result" 26 ₁ - 26 _N of the corresponding selection blocks 2 ₁ - 2 _N of the overflow signal that initiates blocking of the output 2.4 _n -2 of the inverter 2.4 _n and a one-time receipt from this inverse output of logical values corresponding to the comparison results on a given, specific point in time, without the possibility of continuing the search, which corresponds to the procedure for dynamic search management - transferring search requests again to the queue to the places corresponding to their priorities.

This locks the corresponding three-input elements And 2.6 ₁ - 2.6 _N (Fig. 4), prohibiting the receipt of clock pulses on the counting inputs Z of the respective counters 2.5 ₁ - 2.5 _N selection blocks 2 ₁ - 2 _N .

Logic values corresponding to the results of the comparison and obtained taking into account the dynamic correction of the maximum search time from the outputs "Result of comparison" 83 ₁ - 83 _N of the selection controllers of the search time 8 ₁ - 8 _N are fed to the corresponding inputs "Result of comparison" 51 ₁ - 51 _N of the strategy register search 5 and the corresponding inputs "Comparison result" 61 ₁ - 61 _N of the block for generating the address of the transition mask 6.

Upon receipt of an identified and verified (from the point of view of catastrophe theory) BDI, the type of which corresponds to one of the types of BDI provided in the search scenario, at the output "Result of comparison" 26 _n , for example, the selection block 2 _n and at the corresponding output "Result of comparison" 83 _n of the selection search time controller 8 _n , where a match is found between the values of the recognized bits of the incoming BDI and the corresponding values of the second bit mask, the value of logical zero will be set, and the outputs of all other selection blocks and, as a result, the selection search time controllers will have the value of a logical unit. When a BDI not provided for by the search scenario arrives, the value of a logical unit will be set at the “Comparison Result” outputs of all selection blocks and, as a result, the “Comparison Result” outputs of the search time selection controllers.

The search strategy register 5 (see Fig. 6) checks whether the recognized by the selection blocks 2 ₁ - 2 _N and verified, taking into account the dynamic correction of the maximum search time, by the selection controllers of the search time 8 ₁ - 8 _N , the type of identified and verified (with from the point of view of catastrophe theory) BDI to the type expected according to the search scenario. The verification is carried out regardless of the results of recognition of the BDI in the blocks of selection 2 ₁ - 2 _N and correction (verification) in the selection controllers of the search time 8 ₁ - 8 _N . The type of expected identified and verified (from the point of view of catastrophe theory) BDI is determined by the mask of the beginning of the search script or transition masks stored in the random access memory 5.2. The mask, which is checked against, is determined by the K-bit address set at the address inputs 5.2-2 (inputs A ₁ - A _K ) random access memory 5.2. The address is the code coming from the bits of the K-bit transit output 113 of the fuzzy search scenario analysis unit 11 and from the corresponding bits of the K-bit output 122 of the fuzzy search scenario conversion unit 12 and set at the K-bit input "Event Code" 52 of the strategy register search 5 (see Fig. 1). From the K-bit input "Event Code" 52, the specified code enters the first group of information inputs 5.1-1 (A ₁ - A _K ) of the selector-multiplexer a 5.1 (see Fig. 6), where, if there is a control input 5.1-3 (SE) selector-multiplexer 5.1 values of logical zero, switched to address inputs 5.2-2 RAM 5.2. The corresponding mask is read by setting a logic zero value at the "Crystal Select" input 56 of the search strategy register 5. Setting the logic zero value at the "Crystal Select" input 56 of the search strategy register 5 must be carried out with a time delay relative to the time T ₁ determined by the delay time signal in the selection block and in the selection search time controller. Let us denote the time of setting a logical zero at the input "Crystal Select" 56 register search strategy 5 as T ₂ . Checking the conformity of the recognized BDI type to the type expected according to the search scenario is carried out by three-input elements OR-NOT 5.3 ₁ - 5.3 _N . At the same time, the first inputs 5.3 ₁ -1 - 5.3 _N -1 of the three-input elements OR NOT 5.3 ₁ - 5.3 _N receive logical values from the outputs 83 ₁ - 83 _N of the corresponding selection controllers of the search time 8 ₁ - 8 _N , and the second inputs are logical values corresponding to the mask read from the random access memory 5.2. The test results come from the outputs 5.3 ₁ -4 - 5.3 _N -4 of the three-input elements OR-NOT 5.3 ₁ - 5.3 _N to the corresponding inputs 5.4-1 ₁ - 5.4-1 _N of the N-input element OR 5.4 after setting the value of logical zero at the enabling input 014 devices. Setting the value of the logical zero at the enabling input 014 of the device must be carried out with a time delay relative to time T ₂ determined by the time of reading information from random access memory 5.2. Let's denote the time of setting a logical zero on the enabling input 014 of the device as T ₃ . If the recognized type of the BDI matches one of the types expected according to the search scenario, the output of the corresponding three-input element OR-NOT, and, consequently, the signal output 59 of the search strategy register 5, will be set to the value of a logical unit, which is fed to the signal input 65 the block for generating the address of the transition mask 6 and to the signal input 45 of the shaper of time intervals 4.

шифратора 6.1 (нулевой вход

шифратора 6.1 не используется, при этом на нем всегда должно быть установлено значение логической единицы «1»). Если поступивший БДИ распознан одним из блоков селекции 2₁ - 2_N с учетом максимального времени поиска, на инверсном входе шифратора 6.1, номер которого соответствует номеру блока селекции, распознавшего БДИ и номеру соответствующего селекционного контроллера времени поиска, будет установлено значение логического нуля, а на всех остальных инверсных входах - значение логической единицы.In the case of finding, for a time not exceeding the specified (maximum search time), the type of incoming identified and verified (from the point of view of catastrophe theory) BDI is found to match the type expected according to the search scenario, in the transition mask address generation block 6 based on the results of recognition of the incoming BDI the formation of the address is carried out, at which in the random access memory 5.2 of the register of the search strategy 5 the mask of transitions is stored, which determines the type of BDI following it. Logic values corresponding to the results of recognition of the BDI from the outputs "Result of comparison" 83 ₁ - 83 _N of the selection controllers of the search time 8 ₁ - 8 _N are fed to the corresponding inputs "Result of comparison" 61 ₁ - 61 _N of the block for generating the address of the transition mask 6 and further (see Fig. 7) - to the corresponding inverse inputs 6.1-1 ₁ - 6.1-1 _N

encoder 6.1 (zero input

encoder 6.1 is not used, while it must always be set to the value of the logical unit "1"). If the incoming BDI is recognized by one of the selection blocks 2 ₁ - 2 _N , taking into account the maximum search time, at the inverse input of the encoder 6.1, the number of which corresponds to the number of the selection block that recognized the BDI and the number of the corresponding selection search time controller, the value of logical zero will be set, and on all other inverse inputs - the value of a logical unit.

дешифратора 6.1 установится код, соответствующий инверсному представлению номера входа дешифратора, на котором установлено значение логического нуля. Инверторами 6.2₁ - 6.2_K данный код преобразуется в код типа БДИ, то есть код, соответствующий номеру блока селекции, распознавшего поступивший БДИ и номеру соответствующего селекционного контроллера времени поиска.At the same time, at the inverse outputs 6.1-2 ₁ - 6.1-2 _K

decoder 6.1 will set the code corresponding to the inverse representation of the input number of the decoder, which is set to the value of logical zero. Inverters 6.2 ₁ - 6.2 _K this code is converted into a BDI type code, that is, a code corresponding to the number of the selection block that recognized the incoming BDI and the number of the corresponding search time selection controller.

This type code BDI, identified and verified (from the point of view of the theory of catastrophes), is used as the address at which the appropriate transition mask is stored in the random access memory 5.2 of the search strategy register 5.

In addition to the fact that this code of the BDI type, identified and verified (from the point of view of catastrophe theory), is subject to analysis (identification) of fuzzy given (observed) values, as well as computational fuzzy transformation and recognition of these values using the mathematics of fuzzy sets to a form that allows its clearly (quantitatively, unambiguously, reliably) define and interpret the values of this code characterizing the type of specific expected BDI.

From the outputs 6.2 ₁ -2 - 6.2 _K -2 of the inverters 6.2 ₁ - 6.2 _K , the K-bit code of the BDI type, identified and verified (from the point of view of catastrophe theory), enters the corresponding information inputs 6.4-2 ₁ - 6.4-2 _K ( D ₁ -D _K ) register 6.4. The entry of the BDI type code into register 6.4 is carried out only when a logical unit value is received at the signal input 65 of the transition mask address generation block 6, that is, only if in the search strategy block 5 a preliminary correspondence of the type of the incoming BDI to the type expected according to the scenario is found search. Otherwise, the previous value of the BDI type code is stored in register 6.4. Thus, at the K-bit output "Event Code" 62 of the block for generating the address of the transition mask 6, a code that requires additional verification is always set, corresponding to the address at which the mask of transitions is stored in the random access memory 5.2 of the search strategy register 5, which determines the type of expected, according to the scenario search, BDI.

The values of the BDI type code, identified and verified (from the point of view of catastrophe theory), which are a priori (initially) defined as the address at which the corresponding transition mask is stored in the search strategy register 5, from the bits of the K-bit output "Event Code" 62 of the formation unit the addresses of the transition mask 6 are received and written in binary code through the K-bit information input 111 to the K-bit input 11.2-1 of the counter 11.2 of the analysis block of the fuzzy search scenario 11 (see Fig. 11).

The block for analyzing the fuzzy search scenario 11 can be implemented in accordance with the scheme proposed in FIG. 11. The algorithm of its operation is described in detail in the prototype (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published on 06/25/2020, 2020, Bull. 18). The BDI type code, identified and verified (from the point of view of catastrophe theory), which is a priori (initially) defined as the address at which the corresponding transition mask is stored in the search strategy register 5, is fed to the K-bit input 11.2-1 of the counter 11.2, which is calculated to register in each cell (bit) one binary number (bit) of incoming information.

Sequential comparison (according to the number of binary numbers characterizing any of the K bits of the BDI type code) of the input data in the binary code and making a decision about their mathematical nature - the BDI type code, identified and verified (from the point of view of catastrophe theory), is set parametrically or with using the membership function characteristic of fuzzy sets is carried out in the analysis unit of the fuzzy search scenario 11 in the sequence described in detail in the prototype (see RF patent No. 2020, Bull. 18) as follows.

If the number of binary numbers characterizing any of the K bits of the BDI type code, identified and verified (from the point of view of catastrophe theory), exceeds one, then, from the point of view of fuzzy mathematics, this code sequence contains redundancy (because it contains, among other things, , values of membership functions of fuzzy sets), which causes the fuzziness of data describing the features of the search scenario (addresses of bit masks).

In other words, initially qualitative and quantitative information coming from the K-bit output "Event Code" 62 of the transition mask address generation unit 6 may differ in the number of binary numbers characterizing any of the K bits of the BDI type code, identified and verified (from the point of view of catastrophe theory): to record quantitative information in a binary code, one binary number is enough, while fuzzy (qualitative) information carries, in addition to the usual number, the characteristic of the membership function, which objectively requires the use of more than one binary number to record and store fuzzy information ( ambiguous (fuzzy) given values of the BDI type code). If one binary number is received that characterizes any of the K bits of the BDI type code, then this code does not need to be checked, clearly (unambiguously) corresponds to the type of the expected block of binary information and reliably determines the address at which the corresponding transition mask is stored in the search strategy register 5.

Taking into account this fact, a storage register 11.1 and a counter 11.2 of block 11 were built, which are described in detail in the prototype (see RF patent No. ). Both of these elements of the circuit are designed for registration, analysis and storage of one binary number characterizing any of the K bits of the BDI type code, if the number of binary numbers exceeds one, then, from the point of view of fuzzy mathematics, this information (values of the elements of the BDI type code) enters fuzzy form. In this case, both the storage register 11.1 and the counter 11.2 of block 11 perform the functions of a transit node, moreover, with K test outputs 11.1-1 ₁ - 11.1-1 _K of the storage register 11.1, this information in binary code immediately, through test outputs 112 ₁ - 112 _K block 11 is supplied to the corresponding K test inputs 121 ₁ - 121 _K of the conversion block of the fuzzy search scenario 12.

If the K-bit information input 111 and the K-bit input 11.2-1 of the counter 11.2 of the analysis block of the fuzzy search scenario 11 receive information in binary code (code values of the BDI type, identified and verified from the point of view of catastrophe theory) in the amount of one binary number characterizing any of the K bits of this code, then this information comes in a clear form, has a quantitative meaning and the counter 11.2 registers this information and from its K outputs 11.2-2 ₁ - 11.2-2 _K sends it to the corresponding K inputs 11.1-2 ₁ - 11.1-2 _K of the storage register 11.1, which, through its transit K-bit output 11.1-3 and the transit output 113 of the analysis unit of the fuzzy search scenario 11, sends this information (the initial data specified in a clear form, quantitatively specified, not requiring additional verification of the value of the type code BDI) to the K-bit inputs "Event Code" 52 and 71 of the search strategy register 5 and the display unit 7, respectively.

Information that requires additional verification (given in a fuzzy form, data fuzzy values of the BDI type code, identified and verified from the point of view of catastrophe theory), in a binary code through K verification inputs 121 ₁ - 121 _K enters the corresponding inputs 12.1-1 ₁ - 12.1 -1 _K register 12.1 of the fuzzy search script transformation block 12, which can be implemented in accordance with the scheme proposed in FIG. 12. Transformation (transformation) of the source data given in a fuzzy form to a form suitable for obtaining clear, unambiguous (reliable) results of a specific search scenario occurs in the fuzzy search scenario transformation block 12 in accordance with the sequence described in detail in the prototype (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published on 06/25/2020, 2020, Bull. 18) as follows. The essence of the operation of the conversion block of the fuzzy search scenario 12, from the point of view of mathematics, is the correct calculation of the integrated opinion of experts and the decision on the maximum value of the membership function (degree of confidence) of this integrated opinion, which determines the unambiguous choice of the quantitative value of the fuzzy parameter - the quantitative value of a specific type code BDI.

(c_k) - функции принадлежности значения конкретного k-ого нечетко заданного (наблюдаемого) кода типа БДИ

, к пространству истинных (однозначно, четко определенных) значений, в двоичном коде поступают на соответствующие входы 12.8-1₁ - 12.8-1_K вычислителя 12.8, дополнительные входы 12.2-5₁ - 12.2-5_K элемента расчета дополнения 12.2 и дополнительные входы 12.4-3₁ - 12.4-3_K вспомогательного элемента хранения 12.4 (см. патент РФ №2724788, «Устройство поиска информации» МПК G06F 9/46, опубликован 25.06.2020 г., 2020, Бюл. 18).The fuzzy code sequence (fuzzy code values of the BDI type, identified and verified from the point of view of catastrophe theory), enters the inputs 12.1-1 ₁ - 12.1-1 _K of the register 12.1 of the conversion block of the fuzzy search scenario 12 (see Fig. 12). Register 12.1 registers and sorts information into two components, in accordance with the number of expert opinions (number of experts) on the degree of belonging of the value of a particular k-th code of the BDI type to the space of true (unambiguously, well-defined) values. Primary 12.1-2 ₁ - 12.1-2 _K and secondary 12.1- 3 ₁ - 12.1-3 _K outputs of the register 12.1 correspond to the data from the first U ₁ and second U ₂ experts, from these outputs information in binary code is supplied respectively to the inputs 12.3-1 ₁ - 12.3-1 _K of the main storage element 12.3 and direct inputs 12.4-1 ₁ - 12.4-1 _K of the auxiliary storage element 12.4, as well as respectively to the primary inputs 12.2- 1 ₁ - 12.2-1 _K and secondary inputs 12.2-2 ₁ - 12.2-2 _K of the calculation element of Addendum 12.2. Addition 12.2 calculation element implements the function of arithmetic subtraction from unity of the values of membership functions of fuzzy sets, in accordance with the algorithm described in the prototype (see RF patent No. 2020, Bull. 18). The main 12.3 and auxiliary 12.4 storage elements store fuzzy information from the expert U ₁ and U ₂ and through their respective outputs 12.3-2 ₁ - 12.3-2 _K and 12.4-2 ₁ - 12.4-2 _K in binary code give out the values of membership functions of fuzzy sets to the main inputs 12.5-1 ₁ - 12.5-1 _K of the main analyzer 12.5 and to the main inputs 12.6-1 ₁ - 12.6-1 _K of the auxiliary analyzer 12.6, respectively. Each of the main 12.5 and auxiliary 12.6 analyzers, receiving in binary code on their additional inputs 12.5-2 ₁ - 12.5-2 _K and 12.6-2 ₁ - 12.6-2 _K , respectively, the values of the complement elements of fuzzy sets from the primary 12.2-3 ₁ - 12.2 -3 _K and secondary 12.2-4 ₁ - 12.2-4 _K outputs of the calculation element of addition 12.2, performs the function of crossing fuzzy sets, as described in the prototype (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published June 25, 2020, 2020, Bul. 18). From the outputs 12.5-3 ₁ - 12.5-3 _K of the main analyzer 12.5 and the outputs 12.6-3 ₁ - 12.6-3 _K of the auxiliary analyzer 12.6, the obtained values in binary code are respectively sent to the primary 12.7-1 ₁ - 12.7-1 _K and secondary 12.7- 2 ₁ - 12.7-2 _K inputs of the calculation element of the union 12.7, which performs the final cycle of disjunctive summation. From the outputs 12.7-3 ₁ - 12.7-3 _K of the calculation element of the union 12.7, the resulting total values (generalized opinion of experts about the value)

(c _k ) - membership functions of the value of a specific k-th fuzzy (observable) code of the BDI type

, to the space of true (unambiguously, well-defined) values, in binary code, are fed to the corresponding inputs 12.8-1 ₁ - 12.8-1 _K of the calculator 12.8, additional inputs 12.2-5 ₁ - 12.2-5 _K of the calculation element of the supplement 12.2 and additional inputs 12.4 -3 ₁ - 12.4-3 _K of the auxiliary storage element 12.4 (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published 06/25/2020, 2020, Bull. 18).

The calculator 12.8 of the conversion unit of the fuzzy search scenario 12 performs an unambiguous selection (assignment) of quantitative values of the fuzzy given values of a specific k-th fuzzy given (observed) code of the BDI type, identified and verified from the point of view of catastrophe theory. Being, in fact, a programmable comparison circuit (a programmable TTL comparator of type 74LS85), in which the binary code compares the values of the membership function of a specific k-th fuzzy (observed) code of the BDI type obtained from the union calculation element 12.7, calculator 12.8, based on determining the maximum of the membership function, unambiguously, clearly and finally assigns the numerical value of this code, which determines the address at which the mask of transitions is stored in the random access memory of the search strategy register, which determines the type of BDI expected according to the search scenario, as described in the prototype (see RF patent No. 2724788, "Information retrieval device" IPC G06F 9/46, published 06/25/2020, 2020, Bull. 18).

The transmission of information to additional inputs 12.2-5 ₁ - 12.2-5 _K of the calculation element of the add-on 12.2 and additional inputs 12.4-3 ₁ - 12.4-3 _K of the auxiliary storage element 12.4 is intended for the case when the number of experts is more than two. In this case, the addition of the calculation element of the union 12.7 of the fuzzy set obtained from the outputs 12.7-3 ₁ - 12.7-3 _K in the calculation element of the complement 12.2 and the values obtained from the outputs 12.7-3 ₁ - 12.7-3 _K of the calculation element of the union 12.7 are overwritten in the auxiliary element storage 12.4, playing the role of information from the first expert. Information from a new (for example, third) expert is written through register 12.1 to the main storage element 12.3 and the calculation cycle is repeated again.

Thus, an unambiguous determination of the quantitative values of the fuzzy parameters of the search scenario is carried out, the calculator 12.8 outputs at its K-bit output 12.8-2 in binary code the quantitative value of a specific k-th fuzzy (observable) code of the BDI type, which is the address where in the random access memory The device of the register of the search strategy stores a mask of transitions that determines the type of BDI identified and verified from the point of view of the catastrophe theory and expected according to the search scenario.

In other words, there is a transformation of certain (recognized) fuzzy initial data characterizing the type of BDI, identified and verified from the point of view of catastrophe theory, to a form suitable for unambiguously making a reliable decision about the address at which the corresponding transition mask is stored in the search strategy register 5.

As a result, on the corresponding digits of the K-bit output of the calculator 12.8 and on the corresponding digits of the K-bit output 122 of the conversion unit of the fuzzy search scenario 12, we obtain information characterizing (based on the analysis of the integrated expert opinion obtained in the framework of fuzzy sets mathematics) the true type of expected BDI, identified and verified in terms of catastrophe theory.

The calculator 12.8 writes, stores and outputs from the corresponding bits of its K-bit output 12.8-2 through the corresponding bits of the K-bit output 122 of the conversion unit of the fuzzy search scenario 12 to the corresponding bits of the K-bit inputs "Event Code" 52 and 71 of the search strategy register 5 and display unit 7, respectively, a code (not a zero code) containing the results of an analysis of the type of expected BDI, identified and verified from the point of view of catastrophe theory - a code that clearly and unambiguously determines the address at which the corresponding transition mask characterizing the scenario is stored in the search strategy register 5 search.

If this code is found to correspond to the type of received BDI (identified and verified from the point of view of catastrophe theory) with the expected code (type), the value of the logical unit from the signal output 59 of the search strategy register 5 is fed to the corresponding signal input 45 of the time interval generator 4 (see Fig. . 5). From the signal input 45 of the time interval generator 4, the value of the logical unit is fed to the write enable input V of the counter 4.7 and the first information input 4.2-1 (input J) of the JK flip-flop 4.2, as described in the prototype (see RF patent No. 2724788, "Search device Information” IPC G06F 9/46, published 06/25/2020, 2020, Bull. 18). In this case, the waiting time code of the next BDI, set at the M-bit input "Waiting time code" 06 of the device, is written to the counter 4.7 and the output 4.2-3 of the JK-flip-flop 4.2 is generated with the value of a logical unit (since its second information input 4.2 -2 (input K) set to logic zero). The value of the logical unit output 4.2-3 JK-flip-flop 4.2, acting on the second input 4.3-2 of the first two-input element And 4.3, allows the receipt of clock pulses from the second clock input 41 shaper time intervals 4 to the counting input With the counter 4.7. Thus, in the generator of time intervals 4, the countdown of the waiting time of the next BDI, identified and verified from the point of view of the theory of catastrophes, is initiated.

The end of the analysis of the received BDI, identified and verified from the point of view of the catastrophe theory, and the transition to waiting for the next BDI for a given time, determined by the code of the maximum search time, is carried out by setting the value of the logical unit at the enabling input 014 of the device, which leads to the unconditional setting of the value of logical zero at the signal output 59 of the search strategy register 5. For the correct operation of the device, setting the value of a logical unit at the enable input 014 of the device must be carried out with a time delay ΔT relative to the time T ₃ , determined by the delay time of the parallel operation of the three-input elements OR-NOT 5.3 ₁ - 5.3 _N and N-input element OR 5.4 search strategy register 5, the maximum of the delay time of writing to the register 6.4 of the block for generating the address of the transition mask 6, the delay time of the JK-flip-flop 4.2 and the delay time of writing to the counter 4.7 of the time interval shaper 4:

where: ΔT _5.3 - delay time of parallel operation of three-input elements OR-NOT 5.3 ₁ - 5.3 _N search strategy register 5;

ΔT _5.4 - delay time N-input element OR 5.4 search strategy register 5;

ΔT _6.4 - the delay time of writing to the register 6.4 of the block for generating the address of the transition mask 6;

ΔT _4.2 - the delay time of the JK-flip-flop 4.2 shaper time intervals 4;

ΔT _4.7 - the delay time of writing to the counter 4.7 of the shaper of time intervals 4.

Simultaneously with the setting of the value of the logical unit at the enabling input 014 of the device, the value of the logical unit at the input "Crystal selection" 012 of the device is set.

If before the expiration of the waiting time for the next BDI, identified and verified from the point of view of the catastrophe theory, but within the maximum search time specified through the main controller of the search time 9, the next BDI arrives, the type of which corresponds to the type expected according to the search scenario, then the value of the logical unit on the signal output 59 of the search strategy register 5 will lead to the re-initialization of the counter 4.7 of the shaper of the time intervals 4 (re-writing the timeout code of the next BDI in the counter). In this case, the logical value at the output 4.2-3 of the JK flip-flop 4.2 of the time interval generator 4 will not change (since its second information input 4.2-2 (input K) is set to a logical zero value), and the clock pulses will continue to flow to the counting input C counter 4.7. Thus, the countdown of the waiting time for the next BDI, identified and verified from the point of view of catastrophe theory within the specified maximum search time, will start from the beginning.

If the next BDI, identified, verified from the point of view of catastrophe theory and corresponding to the search scenario, does not arrive before the expiration of the waiting time, but within the specified maximum search time, then the counter 4.7 of the time interval generator 4 will overflow. shaper time intervals 4 will set the value of the logical unit, which will go to the second input 4.5-2 of the second two-input element And 4.5 and through the first two-input element OR 4.1 - to the second information input 4.2-2 (input K) JK-flip-flop 4.2. The values of the logical unit at the output of the overflow P counter 4.7 and the signal input 45 shaper time intervals 4 can be set at arbitrary times, but within the specified maximum search time. In this regard, at the information inputs 4.2-1 (input J) and 4.2-2 (input K) of the JK-flip-flop 4.2 of the time interval generator 4, the following combinations of logical values may appear:

the first information input 4.2-1 (input J) JK flip-flop 4.2 is set to a logic zero, and its second information input 4.2-2 (input K) is set to a logic one;

on the first 4.2-1 (input J) and the second 4.2-2 (input K) information inputs of the JK flip-flop 4.2 set to the value of the logical unit.

In the first case, the output 4.2-3 of the JK flip-flop 4.2 of the time interval generator 4 will be set to a logical zero value, which will be fed to the second input 4.3-2 of the first two-input element And 4.3 and to the input 4.4-1 of the inverter 4.4. The value of logical zero at the second input 4.3-2 of the first two-input element And 4.3 will stop the receipt of clock pulses from the second clock input 41 of the time interval generator 4 to the counting input C of the counter 4.7 (see Fig. 5). The value of a logical zero at the input 4.4-1 of the inverter 4.4 will lead to setting the value of a logical unit at the first input 4.5-1 of the second two-input element AND 4.5, which, if there is a value of a logical unit at its second input 4.5-2, will lead to setting the value of a logical unit at the second input 4.6-2 of the second two-input element OR 4.6 and at the output "Reset" 44 shaper time intervals 4. The value of the logical unit at the second input 4.6-2 of the second two-input element OR 4.6 will reset the counter 4.7, and the value of the logical unit at the output "Reset" 44 will reset the register 6.4 of the block for generating the address of the transition mask 6. Resetting the counter 4.7 of the time interval shaper 4 will lead to setting its overflow output P to a logical zero value, which will go to the second information input 4.2-2 (input K) of the JK flip-flop 4.2 and to the second input 4.5-2 of the second two-input element And 4.5. In this case, the output "Reset" 44 shaper time intervals 4 will be set to the value of logical zero. Resetting the register 6.4 of the block for generating the address of the transition mask 6 (see Fig. 8) will lead to the installation of address inputs 5.2-2 (inputs A ₁ - A _K ) of the random access memory 5.2 of the search strategy register 5 (see Fig. 6) of the zero address . Thus, the search for the next BDI, identified and verified from the point of view of the catastrophe theory and determined by the current, within the specified maximum search time, transition mask, is interrupted and the device proceeds to waiting for the BDI, the types of which are determined by the mask of the beginning of the search scenario (initial BDI).

In the second case, the logical value at the output 4.2-3 JK-flip-flop 4.2 shaper time intervals 4 will not change. In this case, the clock pulses will continue to arrive at the counting input C of the counter 4.7, and at the output 4.5-3 of the second two-input element And 4.5, the value of logical zero will remain (despite the value of the logical unit at its second input 4.5-2). The value of the logical unit at the signal input 45 of the time interval generator 4 (see Fig. 5) will lead to writing the waiting code for the next BDI, identified and verified from the point of view of the catastrophe theory and set at the M-bit input "Waiting time code" 06 of the device to the counter 4.7. At the same time, the overflow output P of the counter 4.7 will be set to a logical zero value, which will come through the first two-input element OR 4.1 to the second information input 4.2-2 (input K) of the JK flip-flop 4.2 and to the second input 4.5-2 of the second two-input element And 4.5. Thus, the reset of the register 6.4 of _the block _for generating the address of the transition mask 6 (see Fig. 7) will not occur and the address corresponding to the address next verified (from the point of view of fuzziness elimination) transition mask. In this regard, the operation of the device to search for the next (according to the search scenario) BDI identified and verified from the point of view of catastrophe theory within the specified maximum search time will continue.

инверсном выходе 7.1-2_n (выходе

из

выходов) дешифратора 7.1, номер n которого соответствует коду типа БДИ, установится значение логического нуля, а на всех остальных выходах дешифратора - значение логической единицы. Логические значения с выходов 7.1-2₁ - 7.1-2_N (выходов

) дешифратора 7.1 блока индикации 7 поступают на первые входы 7.2₁-1 - 7.2_N-1 соответствующих двухвходовых элементов ИЛИ 7.2₁ - 7.2_N, где происходит их сравнение со значениями маски окончания сценария поиска, установленными на N-разрядном входе «Правило завершения поиска» 07 устройства. При совпадении логических значений на входах двухвходового элемента ИЛИ, номер которого соответствует коду типа БДИ, на его выходе установится значение логического нуля, которое через N-входовый элемент И 7.3 поступит на выход «Результат поиска» 015 устройства, что соответствует четкому (достоверному) обнаружению во входящем потоке БДИ, идентифицированного и верифицированного с точки зрения теории катастроф и соответствующего сценарию поиска.Upon receipt of the BDI, the type of which corresponds to one of the types of BDI identified and verified from the point of view of the catastrophe theory and indicated in the end mask of the search scenario (final BDI), a logical zero value is formed at the output “Search result” 015 of the device. The formation of a logical zero value at the output "Search result" 015 of the device is carried out as follows (see Fig. 8). From the bits of the K-bit input "Event code" 71 of the display unit 7, the verified results of fuzzy analysis of the code of the type of the expected BDI, identified and verified from the point of view of catastrophe theory, are fed to the corresponding inputs 7.1-1 ₁ - 7.1-1 _K (inputs F ₁ - F _K ) decoder 7.1. At the same time, on the nth

inverted output 7.1-2 _n (output

from

outputs) of the decoder 7.1, the number n of which corresponds to the BDI type code, the value of the logical zero will be set, and on all other outputs of the decoder - the value of the logical unit. Logic values from outputs 7.1-2 ₁ - 7.1-2 _N (outputs

) of the decoder 7.1 of the indication block 7 are fed to the first inputs 7.2 ₁ -1 - 7.2 _N -1 of the corresponding two-input elements OR 7.2 ₁ - 7.2 _N , where they are compared with the search script end mask values set at the N-bit input "Search completion rule » 07 devices. If the logical values at the inputs of the two-input OR element, the number of which corresponds to the BDI type code, match, the value of the logical zero will be set at its output, which, through the N-input element AND 7.3, will go to the output “Search result” 015 of the device, which corresponds to a clear (reliable) detection in the incoming flow of the BDI, identified and verified from the point of view of the catastrophe theory and corresponding to the search scenario.

Thus, the proposed information retrieval device provides an increase in the reliability and stability of operation under conditions inherent in the real process of receipt (circulation) of large arrays (streams) of heterogeneous data in SPD, ISPS and in SIAOD. This device is capable of both performing search queries and identifying and verifying the boundary and emergency (catastrophic) values of the intensity of the analyzed information flows with smooth changes in external conditions and control actions (intensity of incoming BDI). The increase in reliability and stability occurs due to the procedures implemented in the disaster analysis unit for identifying and verifying the states of the boundary, emergency (catastrophic) intensity of incoming blocks of binary information with smooth changes in the parameters of external conditions and control actions, as well as timely notification (warning) of the user (operator, system administrator) of such a search system about its possible emergency state, based on the received identification and verification data.

This result is due to the receipt from the L-bit information output 131 of block 13 to the L-bit information inputs 21 ₁ - 21 _N selection blocks 2 ₁ - 2 _N signals characterizing the identified and verified (in terms of checking the intensity λ by methods of catastrophe theory) binary values bits of the BDI, as well as signals characterizing the type of the BDI, determined by the first and second bit masks, coming from the corresponding N second L-bit outputs "Mask 1" 134 ₁ - 134 _N and the corresponding N second L-bit outputs "Mask 2" 135 ₁ - 135 _N catastrophe analysis unit 13 to the second L-bit inputs "Mask 1" 22 ₁ - 22 _N and the second L-bit outputs "Mask 2" 23 ₁ - 23 _N N corresponding selection blocks 2 ₁ - 2 _N .

An analysis of the principle of operation of the claimed device shows the obviousness of the fact that, along with the capabilities stored and described in the prototype for searching and recognizing information, the features of which can be set (defined) both quantitatively and qualitatively - fuzzy, with the involvement of a linguistic variable, the device is able to carry out implementation of procedures for identifying and verifying the states of the boundary and emergency (catastrophic) intensity of incoming elements of the data stream (intensity of incoming BDI), as well as controlled formation of values of permissible intensity, which leads to an increase in the reliability and stability of the claimed device in real conditions in which it has to function.

This device provides an increase in the reliability and stability of operation under conditions inherent in the real process of implementing search queries in large arrays (streams) of heterogeneous data circulating in modern data networks, in information and reference and search systems and in intelligent analytical data processing systems - when the intensity of incoming data, the intensity of blocks of binary information can smoothly change under the influence of control actions (search time for current search queries) or external factors, creating a potential threat of blocking (collapse) of the information retrieval process for systems of this class. This significantly expands the scope of the device, expands its functionality, allowing identification, verification and a priori warning of emergency (catastrophic) conditions for the implementation of the search process in large amounts of information, allowing you to reliably, stably, timely and dynamically search for information in real conditions, in streams of heterogeneous data of large volume, when there is a threat of reaching a critical (on the border of the possibilities of implementing search queries) intensity of BDI arrival with smooth changes in the parameters of external conditions and control actions that affect systems and networks where the claimed information retrieval device will be used.

INFORMATION SOURCES

1. Gilmore R. Applied Catastrophe Theory. Book 2. - M.: Mir, 1984. - 285 p.

2. Arnold V.I. Theory of catastrophes. - M.: Editorial URSS, 2004. - 128 p.

3. Poston T., Stuart I. Catastrophe theory and its applications. Per. from English. - M.: Mir, 1980.-608 p.

4. Petrov Yu.P., Petrov L.Yu. The Unexpected in Mathematics and Its Connection with Accidents and Catastrophes of Recent Years. - St. Petersburg: NIIKh St. Petersburg State University, 1999. - 108 p.

5. Parashchuk I.B., Dyakov S.V. Mathematics of the theory of catastrophes in relation to the problems of analyzing the reliability of communication network elements. / Communication systems. Analysis. Synthesis. Management./ Ed. prof. Postyushkova V.P. Issue 5. - St. Petersburg: Publishing House "Theme", 2001. - 84 p., S. 47-49.

6. Parashchuk I.B., Dyakov S.V. Prospects for assessing the stability of telecommunication networks using the methods of catastrophe theory. // 56th scientific and technical conference dedicated to the Radio Day. Conference materials. St. Petersburg: SPbGETU "LETI", 2001, pp. 56-57.

7. Gonzalez R., Tu J. Principles of pattern recognition. - M.: Mir, 1978. - 411 p.

Claims (2)

1. An information retrieval device containingNmask storage units (1₁-1 _N ), WhereN ≥ 2 - the number of types of binary information blocks (BDI),N selection blocks (2₁-2 _N ,), frequency divider (3), time interval shaper (4), search strategy register (5), transition mask address generation unit (6), indication unit (7),N selection search time controllers (8₁-8 _N ), the main search time controller (9), the clock generator (10), the fuzzy search scenario analysis unit (11), the fuzzy search scenario conversion unit (12), and the clock input (38) of the frequency divider (3) is the first clock input (08) of the device, and the output (39) of the frequency divider (3) is connected to the clock input (41) of the time interval generator (4), the recording enable inputs (16₁-16 _N )N mask storage units (1₁-1 _N ) are combined and are the write enable input (01) of the device, the firstL-bit, whereL ≥ 2 - the maximum number of bits in the block of binary information, inputs "Mask 1" (17 _n ) and "Mask 2" (18 _n )n-th mask storage unit (1 _n ), Wheren = 1, 2, …,N, aren - mi first L - bit inputs respectively "Mask 1" (02 _n ) and "Mask 2" (03 _n ) of the device, the input "Initial reset" (42) of the time interval generator (4) is connected to the input "Initial reset" (63) of the transition mask address generation unit (6) and is the input "Initial reset" (05) of the device, whileM- bit, whereM ≥ 2 - bit length of the waiting time code, the input "Waiting time code" (43) of the time interval generator (4) isM-bit input "Wait time code" (06) of the device, and the output (44) of the time interval generator (4) is connected to the "Reset" input (64) of the transition mask address generation unit (6), the signal output (59) of the search strategy register (5) connected to the signal inputs (45) and (65) of the time interval generator (4) and the transition mask address generation unit (6), respectively,K- bit, whereK = (log₂ N) + 1 – bit length of the transition mask address code, address input (53), control input (54),N-bit information input (55) and enabling input (58) of the search strategy register (5) are respectivelyK-bit address input (09), control input (010),N-bit information input (011) and enabling input (014) of the device, the inputs "Crystal selection" (56) and "Read / write" (57) of the search strategy register (5) are respectively the inputs "Crystal selection" (012) and " Read/Write (013) devices,N-bit input "Search completion rule" (72) and output "Search result" (73) of the display unit (7) are respectivelyN-bit input "Search completion rule" (07) and output "Search result" (015) of the device, outputs "Comparison result" (26₁-26 _N ) selection blocks (2₁-2 _N ) are connected to the corresponding inputs "Comparison result" (81₁–81 _N ) of the corresponding search time selection controllers (8₁-8 _N ), "Comparison Result" outputs (83₁–83 _N ) of which are connected to the corresponding inputs "Comparison result" (51₁-51 _N ) search strategy register (5) and with the corresponding inputs "Comparison result" (61₁-61 _N ) block for generating the address of the transition mask (6), the output (101) of the clock generator (10) is connected to the clock inputs (25₁-25 _N ) of eachN selection blocks (2₁-2 _N ), inputs "Zeroing" (24₁-24 _N ) which are the corresponding inputs "Reset" (017₁–017 _N ) devices, andS- bit, whereS ≥ 2 – bit length of the search time correction code, corrective input (27 _n )nth selection block (2 _n ) is connected toS-bit test input (82 _n )n-th selection search time controller (8 _n ) and connected ton-muS- bit output (92 _n ) main search time controller (9),N S- bit inputs (91₁–91 _N ) of which are the correspondingNS- bit inputs "Correction of the maximum search time" (016) of the device, whileK- bit output "Event code" (62) of the transition mask address generation unit (6) is connected toK- bit information input (111) of the block for analyzing the fuzzy search scenario (11),Ktest outlets (112₁ – 112 _K ) which are connected to the correspondingK test inputs (121₁–121 _K ) fuzzy search script conversion block (12),K- bit output (122) of which andK- bit transit output (113) of the fuzzy search scenario analysis block (11) are combined and areK-bit inputs "Event code" (52) and (71) of the search strategy register (5) and indication block (7), respectively, characterized in that an additional catastrophe analysis block (13) is introduced, designed to carry out procedures for identifying and verifying the states of the boundary , emergency (catastrophic) intensity of incoming blocks of binary information with smooth changes in the parameters of external conditions and control actions, as well as for generating signals of a logical zero or a logical unit (prediction and warning signal), characterizing, respectively, the absence or presence of a possible catastrophic state of the system, andL-bit, whereL ≥ 2 is the maximum possible number of bits in a block of binary information used in the search scenario, information inputs (21₁-21 _N )N selection blocks (2₁-2 _N ) are combined and areL- bit information output (131) of the disaster analysis unit (13),L-bit information input (130) which isL- bit information input (04) of the device, while the secondL-bit outputs "Mask 1" (14 _n ) and "Mask 2" (15 _n )n-th mask storage unit (1 _n ) are connected to the correspondingn-th secondL-bit inputs "Mask 1" (132 _n ) and "Mask 2" (133 _n ) catastrophe analysis unit (13),n-th secondL-bit outputs "Mask 1" (134 _n ) and "Mask 2" (135 _n ) which are connected to the corresponding secondL-bit inputs "Mask 1" (22 _n ) and "Mask 2" (23 _n )n-th selection block (2 _n ), the output (101) of the clock generator (10) is connected to the clock input (138) of the catastrophe analysis unit (13), the control input (136) of which is the input "Input intensity threshold values" (018) of the device, warning output (137) of the catastrophe analysis block (13) is the “Disaster Threat” output (019) of the information retrieval device. 2. The device according to claim 1, characterized in that the disaster analysis unit (13) consists of an intensity calculator (13.0), G executive random access memories (RAM) (13.1 ₁ –13.1 _G ), where G ≥ 1 is the number of autonomous, successive observation time intervals when implementing search queries, read-only memory (ROM) (13.2), iterative comparison element (13.3), comparison element (13.4), intermediate RAM (13.5), intermediate AND element (13.6), AND element (13.7), additional RAM (13.8), test RAM (13.9) and counter (13.10), moreover, the clock input (13.0-3) of the intensity calculator (13.0) is connected to the clock inputs (13.2-1) and (13.10-3) of the ROM (13.2) and counter (13.10), respectively, and is the clock input (138) of the disaster analysis unit (13), L -bit information input (13.0-1) of the intensity calculator (13.0) is the L -bit information input (130) of the disaster analysis unit (13) and L -bit information input ohm (04) of the device, the L -bit information output (13.0-2) of the intensity calculator (13.0) is the L -bit information output (131) of the disaster analysis unit (13), g -th, where g = 1, 2, ..., G, the executive output (13.0-4 _g ) of the intensity calculator (13.0) is connected to the input of the corresponding g -th executive RAM (13.1 _g ), the G inputs of the executive RAM (13.1 ₁ -13.1 _G ) are combined and connected to the first input (13.3-1 ) iterative comparison element (13.3) and the second input (13.4-2) of the comparison element (13.4), the outputs G of the execution RAM (13.1 ₁ -13.1 _G ) are combined and connected to the second input (13.3-2) of the iterative comparison element (13.3), the output of the ROM (13.2) is connected to the first input (13.4-1) of the comparison element (13.4), the output of the iterative comparison element (13.3) is connected to the input of the intermediate RAM (13.5) and the second input (13.6-2) of the intermediate element AND (13.6), the first input (13.6-1) of which is connected to the output of the intermediate RAM (13.5), the output of the intermediate element AND ( 13.6) is connected to the output of the comparison element (13.4) and connected to the first input (13.7-1) of the AND element (13.7), the second input (13.7-2) of which is connected to the test input (13.2-2) of the ROM (13.2) and is a test output (13.9-1) of the test RAM (13.9), the output of the AND element (13.7) is connected to the reading inputs (13.0-5) and (13.8-1) of the intensity calculator (13.0) and additional RAM (13.8), respectively, and is a warning output ( 137) of the catastrophe analysis block (13) and the output "Disaster threat" (019) of the device, while N L -bit inputs "Mask 1" (13.8-2 ₁ -13.8-2 _N ) and N L -bit inputs "Mask 2 » (13.8-3 ₁ –13.8-3 _N ) of additional RAM (13.8) are the corresponding N second L -bit inputs “Mask 1” (132 ₁ –132 _N ) and the corresponding N second L -bit inputs “Mask 2” (133 ₁ –133 _N ) block (13), N L -bit outputs "Mask 1" (13.8-4 ₁ -13.8-4 _N ) and N L -bit outputs "Mask 2" (13.8-5 ₁ -13.8-5 _N ) of additional RAM (13.8) are N second L -bit outputs "Mask 1" (134 ₁ -134 _N ) and the corresponding N second L -bit outputs "Mask 2" (135 ₁ -135 _N ) of block (13), clock output (13.10-1) of the counter (13.10) is connected to the clock input (13.9-2) of the test RAM (13.9), the release output (13.9-3) of which is connected to the release input (13.10-2) of the counter (13.10), the control input (13.9-4) of the test RAM ( 13.9) is the control input (136) of the disaster analysis unit (13) and the input "Input of intensity threshold values" (018) of the information retrieval device.

Images