CN111784533B

CN111784533B - Information analysis method based on artificial intelligence and big data and cloud computing platform

Info

Publication number: CN111784533B
Application number: CN202010550357.2A
Authority: CN
Inventors: 洪江川
Original assignee: Infinity Network Technology Beijing Co ltd
Current assignee: Infinity Network Technology Beijing Co ltd
Priority date: 2020-06-16
Filing date: 2020-06-16
Publication date: 2021-04-27
Anticipated expiration: 2040-06-16
Also published as: CN111784533A; CN112991095A; CN112991094A

Abstract

According to the information analysis method and the cloud computing platform based on the artificial intelligence and the big data, firstly, the running state information of the power equipment is obtained, secondly, each group of the obtained running state information is analyzed to obtain the state change track and the plurality of information labels of each group of the running state information in each set time period, the state change tracks and the plurality of information labels are packaged to obtain the state change set, then the state change set is transmitted to the artificial intelligence server, the identification result returned by the artificial intelligence server is obtained, and finally whether at least one group of power utilization networks formed by the power equipment and each power equipment is abnormal or not is determined according to the identification result. Therefore, the operation state information can be analyzed in a multi-dimensional mode, faults possibly occurring in the power equipment can be comprehensively and completely analyzed, adverse effects caused by equipment loss of the power equipment are fully considered in combination with cooperativity between the power equipment, accurate and reliable state monitoring on the power equipment can be guaranteed, and large-area power grid production accidents are avoided.

Description

Information analysis method based on artificial intelligence and big data and cloud computing platform

Technical Field

The application relates to the technical field of big data analysis and mining, in particular to an information analysis method and a cloud computing platform based on artificial intelligence and big data.

Background

The information and digital age is a big trend of the current age development, and the social form is gradually developed from the industrial society to the information society. The industrial production and daily life of society have advanced toward intelligence, safety and efficiency by means of informatization technology and digitization technology. By utilizing digital information to serve all walks of society, labor costs can be effectively released and productivity optimized. Taking the electric power thing networking as an example, the problem that the delay is too long and the flow is too big is collected in the construction of present digital power grid can effectively be solved to realize the full life cycle management of electric power thing networking, ensure the safe high-efficient operation of electric power thing networking. However, when the digital power grid is in operation, cooperativity between different power devices and device loss are often ignored, which may cause errors in condition monitoring of the power devices, and may cause a large-area power grid production accident.

Disclosure of Invention

The application aims to provide an information analysis method and a cloud computing platform based on artificial intelligence and big data so as to solve the technical problem that errors occur when the state of power equipment is monitored in the prior art.

The information analysis method based on artificial intelligence and big data is applied to a cloud computing platform which is communicated with an artificial intelligence server and a plurality of electric power devices, and at least comprises the following steps:

acquiring the running state information of the corresponding power equipment through an information transmission interface pre-established with each power equipment; the running state information is acquired when the sensor deployed in the corresponding power equipment runs and is uploaded to the cloud computing platform in real time through the information transmission interface;

periodically analyzing each set of acquired running state information to obtain a state change track and a plurality of information labels of each set of running state information in each set time period;

packaging the state change track and the plurality of information labels of each group of running state information in each set time period to obtain a state change set of each group of running state information; transmitting each state change set to the artificial intelligence server when detecting that the vacancy rate of the identification threads in the artificial intelligence server reaches a set rate;

acquiring an identification result returned by the artificial intelligence server based on the state identification of each state change set through the identification thread; the identification result comprises abnormal state parameters determined based on the equipment loss weight of the electric equipment;

and determining whether each electrical device and at least one group of electricity utilization networks formed by the electrical devices have abnormality according to the identification result.

Providing a cloud computing platform in communication with an artificial intelligence server and a plurality of electrical devices, the cloud computing platform to:

According to the information analysis method and the cloud computing platform based on the artificial intelligence and the big data, firstly, the running state information of the power equipment is obtained, each group of the obtained running state information is periodically analyzed, the state change track and the plurality of information labels of each group of the running state information in each set time interval are obtained and packaged to obtain the state change set, then the state change set is transmitted to the artificial intelligence server, the identification result returned by the artificial intelligence server based on the state identification of each state change set through the identification thread is obtained, and finally whether each power equipment and at least one group of power utilization networks formed by the power equipment are abnormal or not is determined according to the identification result.

Therefore, the operation state information can be analyzed in a multi-dimensional mode, faults possibly occurring in the power equipment can be comprehensively and completely analyzed, adverse effects caused by equipment loss of the power equipment are fully considered in combination with cooperativity between the power equipment, accurate and reliable state monitoring on the power equipment can be guaranteed, and large-area power grid production accidents are avoided.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and together with the description, serve to explain the principles of the application.

FIG. 1 is a schematic diagram of an information analysis system based on artificial intelligence and big data provided by the invention.

FIG. 2 is a flow chart of an information analysis method based on artificial intelligence and big data provided by the invention.

Fig. 3 is a functional block diagram of an information analysis apparatus based on artificial intelligence and big data according to the present invention.

Fig. 4 is a hardware structure diagram of a cloud computing platform provided in the present invention.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

In order to solve the technical problems that errors occur when the state of the power equipment is monitored and large-area power grid production accidents are caused in the prior art, the embodiment of the invention provides an information analysis method and a cloud computing platform based on artificial intelligence and big data, which can carry out multi-dimensional analysis on the collected running state information of the power equipment, so that possible faults of the power equipment can be comprehensively and completely analyzed, adverse effects caused by equipment loss among the power equipment are fully considered based on the cooperativity of the power equipment, accurate and reliable state monitoring on the power equipment can be ensured, and large-area power grid production accidents are avoided.

To achieve the above object, please first refer to fig. 1, which is a schematic diagram of a communication connection architecture of an artificial intelligence and big data based information analysis system 100 according to an embodiment of the present invention, where the information analysis system 100 may include a cloud computing platform 110, an artificial intelligence server 120, and a plurality of power devices 130. Further, the cloud computing platform 110 is respectively in communication with the artificial intelligence server 120 and the plurality of power devices 130, and is configured to collect operation state information from the power devices 130, and analyze the operation state information in combination with the artificial intelligence server 120, so as to monitor whether an electricity network formed by the power devices 130 and the power devices 130 is abnormal, and ensure normal operation of the digital power grid.

On the basis, please refer to fig. 2 in combination, a schematic flow chart of an information analysis method based on artificial intelligence and big data is provided, the information analysis method may be applied to the cloud computing platform 110 in fig. 1, and the method may specifically include the contents described in the following steps S110 to S150.

Step S110, acquiring running state information of corresponding electric equipment through an information transmission interface pre-established with each electric equipment; the running state information is acquired when the sensor deployed in the corresponding power equipment runs and is uploaded to the cloud computing platform in real time through the information transmission interface.

In step S110, the operating status information of different electrical devices is different, and in this embodiment, the electrical devices include, but are not limited to, utility boilers, steam turbines, gas turbines, water turbines, generators, transformers, contactors, and the like, which are not limited herein.

And step S120, periodically analyzing each set of acquired running state information to obtain a state change track and a plurality of information labels of each set of running state information in each set time period.

In step S120, the setting time period may be adjusted according to the actual operating condition, for example, during the peak period of power utilization, the setting time period may be reduced appropriately, and during the valley period of power utilization, the setting time period may be increased appropriately. Further, the state change trajectory may be a curve for characterizing the stability of the power device 130, and the information tag may be a plurality of tags extracted from the state change trajectory and reflecting the state dimension of the power device 130.

Step S130, packaging the state change track and the plurality of information labels of each group of running state information in each set time period to obtain a state change set of each group of running state information; transmitting each state change set to the artificial intelligence server upon detecting that an idle rate of identified threads in the artificial intelligence server reaches a set rate.

In step S130, the status information change set may be a result of encapsulating the information tag of each set of operating status information with the status change trace, for example, the information tag may be implanted at a different position in the status change trace to mark the status change trace.

Step S140, obtaining the identification result returned by the artificial intelligence server based on the state identification of each state change set through the identification thread; the identification result comprises abnormal state parameters determined based on the equipment loss weight of the electric equipment.

And S150, determining whether each electric device and at least one group of power utilization networks formed by the electric devices are abnormal or not according to the identification result.

It can be understood that, when the contents described in steps S110 to S150 are applied, the operation state information of the electrical equipment is first obtained, each set of the obtained operation state information is periodically analyzed, a state change track and a plurality of information tags of each set of the operation state information in each set time period are obtained and encapsulated to obtain a state change set, then the state change set is transmitted to the artificial intelligence server, an identification result returned by the artificial intelligence server based on the state identification of each state change set through the identification thread is further obtained, and finally, whether each electrical equipment and at least one set of the electricity utilization networks formed by the electrical equipment are abnormal or not is determined according to the identification result.

Therefore, the operation state information can be analyzed in a multi-dimensional mode, faults possibly occurring in the power equipment can be comprehensively and completely analyzed, adverse effects caused by equipment loss of the power equipment are fully considered, accurate and reliable state monitoring on the power equipment can be guaranteed, and large-area power grid production accidents are avoided.

In specific implementation, the inventor finds that, when analyzing the operating state information, the coupling relationship between different electrical devices is often easily ignored, so that the cooperativity between different electrical devices is ignored, which may cause a deviation in the analysis result of the operating state information, and the analysis of the operating state information cannot be completely realized, which is directly related to the accuracy and reliability of the subsequent state monitoring of the electrical devices. Therefore, in order to solve the above technical problem, in step S120, each set of acquired operating state information is periodically analyzed to obtain a state change track and a plurality of information tags of each set of operating state information in each set time period, which may specifically include the contents described in steps S121 to S125 below.

Step S121, acquiring information attribute parameters corresponding to each set of running state information in each set time period and port information of a coupling port of the power equipment corresponding to each set of running state information, determining associated attribute parameters corresponding to the information attribute parameters according to the port information of the coupling port of the power equipment corresponding to each set of running state information, and extracting device identifications of other power equipment having coupling relation with the power equipment corresponding to each set of running state information from the associated attribute parameters; the information attribute parameter is a parameter sequence determined based on the attribute identifier in the operating state information, and sequence coding logic corresponding to the parameter sequence is fixed and invariable.

Step S122, determining the wiring configuration information between the power equipment corresponding to each group of running state information and other power equipment having coupling relation with the power equipment, and determining a state offset coefficient between an information attribute parameter corresponding to each group of running state information and the wiring configuration information corresponding to each group of running state information based on the associated attribute parameter; and the state offset coefficient is used for representing a coordination coefficient between the electric equipment corresponding to each group of running state information and other electric equipment.

Step S123, determining whether a state deviation coefficient between the information attribute parameter corresponding to each set of running state information and the wiring configuration information corresponding to each set of running state information is smaller than a preset coefficient.

Step S124, if the state offset coefficient between the information attribute parameter corresponding to each set of running state information and the wiring configuration information corresponding to each set of running state information is smaller than the preset coefficient, mapping the associated attribute parameter corresponding to each set of running state information to the port information corresponding to each set of running state information to obtain the mapping attribute parameter of the associated attribute parameter in the port information, extracting the category of the mapping attribute parameter, and determining the number of information tags of each set of running state information according to the first number of the category; determining target fields matched with each group of information fields in each group of running state information from a preset field extraction list, determining description information of each group of target fields, mapping each group of target fields to a preset coordinate plane according to a time sequence grade included in the description information to obtain corresponding field nodes, fitting the field nodes to obtain state change tracks corresponding to each group of running state information, and determining a plurality of information labels according to track characteristics of the state change tracks.

Step S125, if a state offset coefficient between the information attribute parameter corresponding to each set of operating state information and the connection configuration information corresponding to each set of operating state information is greater than or equal to the preset coefficient, calculating a port loss percentage of a coupling port of the power equipment corresponding to each set of operating state information, correcting each set of information fields corresponding to each set of operating state information according to the port loss percentage to obtain a target field, and determining the number of information tags according to the similarity between the target field and the corresponding information field; and mapping the target field to a preset coordinate plane to obtain a corresponding field node, fitting the field nodes to obtain a state change track corresponding to each group of running state information, and determining a plurality of information labels according to the track characteristics of the state change track.

It can be understood that by performing the above steps S121 to S125, the following technical effects can be achieved: by analyzing and processing the coupling relation and the cooperativity between the power equipment, the state change track and the information label can be generated by adopting different methods based on the size of the state offset coefficient between the information attribute parameter corresponding to each group of running state information and the wiring configuration information corresponding to each group of running state information, so that the deviation of the analysis result of the running state information can be avoided, the running state information can be completely analyzed, and the accuracy and the reliability of the subsequent state monitoring of the power equipment are ensured.

In practical applications, the inventor finds that the operation state information of some electric power equipment is easy to be confused in the process of acquiring the operation state information. The reason for this is that some electrical devices are relatively close in geographical location and have a high-correlation electrical connection relationship, which may cause confusion due to statistical errors occurring when obtaining the operation state information of these electrical devices, thereby affecting subsequent state identification. In order to improve the above technical problem, the operation state information of the corresponding power device acquired through the information transmission interface pre-established with each power device in step S110 may specifically include the contents described in the following steps S111 to S114.

Step S111, reading geographical position information and electrical position information of corresponding electric power equipment through each information transmission interface, extracting direction information between the geographical position information and the electrical position information corresponding to each electric power equipment, and constructing a connection topology of the association degree of the electric power equipment through the direction information; the association degree connection topology is a multi-area network, each area network corresponds to one area electrical level, each area network is provided with at least one electric power device, one device in each area network is electrically connected with one electric power device in at least one other area network, and the area electrical levels have a sequence from high to low.

Step S112, when the running state information of the corresponding power equipment is acquired through each information transmission interface, if two groups of running state information are in the same area network, determining a first similarity of geographic position information and a second similarity of electrical position information between the two groups of running state information; and when the first similarity and the second similarity are both greater than a set similarity, determining a plurality of first information codes of the first running state information and a plurality of second information codes of the second running state information in the two sets of running state information.

Step S113, determining a coincidence ratio between each second information code in the second operating state information and the first information code at the same code position in the first operating state information according to each first information code in the first operating state information and the code registration information thereof, and marking the first target information code and the second target information code, of which the coincidence ratio is greater than a preset ratio.

Step S114, determining a first degree of association of the first target information code in the first running state information and a second degree of association of the second target information code in the second running state information, setting a first signature for the first running state information based on the first degree of association and setting a second signature for the second running state information based on the second degree of association; wherein the first signature and the second signature are different digital signatures.

When the contents described in the above steps S111 to S114 are executed, the similarities between the geographic locations and the electrical locations of the electrical devices can be analyzed, so that different signatures are set for the operating state information with high similarity, and confusion caused by statistical errors occurring when the operating state information of the electrical devices is acquired is avoided. Therefore, the running state information corresponding to the power equipment can be accurately acquired.

When the state change track and the information tag are encapsulated, the inventor finds that if the time sequence consistency and the coding heterogeneity of the state change track and the information tag are not considered, a phenomenon that the information tag is misplaced in the state change track in the encapsulation process is easy to occur, which affects the integrity of the state change set and may cause the loss of part of information of the state change set. For this reason, in step S130, the state change trajectory of each set of operating state information in each set period and the plurality of information tags are encapsulated to obtain the state change set of each set of operating state information, which may specifically include the contents described in steps S131 to S134 below.

Step S131, determining a first time sequence feature array of the state change track and a second time sequence feature array of each information tag of the state change track; each second time sequence feature array has different label weights, the label weights are used for representing the encoding heterogeneous degree of each time sequence feature array and the first time sequence feature data, the first time sequence feature array and each second time sequence feature array have the same number of time sequence feature fields, and each time sequence feature field corresponds to one field encoding string.

Step S132, sorting the second time sequence feature arrays according to the sequence of the label weights from big to small to obtain a feature array sorting sequence, sequentially extracting a feature field distribution sequence of each second time sequence feature array in the sorting sequence, and transforming each feature field distribution sequence through a preset sequence conversion list to obtain a corresponding target distribution sequence; determining a first sequence topology corresponding to a feature field distribution sequence of the first time sequence feature array and a second sequence topology of the target distribution sequence; the first sequence topology and the second sequence topology have the same logical directional connection information.

Step S133, calculating the topological structure similarity of the first sequence topology and each second sequence topology; and the topological structure similarity is the weighted sum of the node similarity and the directed line similarity of the first sequence topology and each second sequence topology.

And S134, sequentially implanting the information labels corresponding to the second sequence topologies into the track node containers corresponding to the state change tracks according to the sequence of the topological structure similarity from small to large so as to obtain the state change set of each group of running state information.

When the contents described in steps S131 to S134 are applied, the information tags can be sequentially embedded in the state change trace based on the above characteristics while taking into account the time sequence consistency and the encoding heterogeneity of the state change trace and the information tags, so that a phenomenon that the information tags are misaligned in the state change trace during the packaging process can be avoided, thereby ensuring the integrity of the state change set and avoiding the loss of part of information of the state change set.

In transmitting the state change sets, in order to improve the identification efficiency of the artificial intelligence server 120, in step S130, each state change set is transmitted to the artificial intelligence server, which may specifically include the contents described in the following steps (1) - (5).

(1) And acquiring a plurality of identification threads in an idle state in the artificial intelligence server.

(2) And judging whether the number of the identification threads is more than or equal to the number of the state change sets.

(3) And transmitting each state change set to one of the identification threads of the artificial intelligence server on the premise that the number of the identification threads is greater than or equal to the number of the state change sets.

(4) On the premise that the number of the identification threads is smaller than the number of the state change sets, acquiring the information throughput of each identification thread; sequencing the identification threads according to the sequence of high and low information throughput to obtain a first sequencing sequence; and sorting the state change sets according to the size sequence of the information capacity of the state change sets to obtain a second sorting sequence.

(5) And sequentially transmitting at least part of state change sets in the second sorting sequence to corresponding identification threads in the first sorting sequence respectively.

For example, if the number of elements in the first sorting sequence is 3, and the number of elements in the second sorting sequence is 10, the first 3 state change sets in the second sorting sequence may be respectively imported into the identification threads in the first sorting sequence, and then, after one of the identification threads in the first sorting sequence completes the identification of the state change set, the 4 th state change set in the second sorting sequence is imported, and so on.

It is understood that through the contents described in the above steps (1) to (5), the state change sets can be identified in parallel based on the throughput of the identification thread, thereby improving the identification efficiency of the artificial intelligence server 120.

In the present embodiment, the artificial intelligence server 120 can consider the device loss of the power device when recognizing the state change set, thereby ensuring the reliability of the recognition result. To achieve the above object, in step S140, the artificial intelligence server 120 can identify the state change set through the following steps S210 to S250.

Step 210, determining path information of each information label in the state change set in a track node container of the corresponding state change track.

Step S220, determining a first state list before the information tag is implanted and a second state list after the information tag is implanted for the state change track corresponding to each state change set according to the path information.

Step S230, determining a device loss weight of the power device corresponding to each state change set based on the list contents at the same list position in the first state list and the second state list.

Step S240, comparing the state change trajectory, in which the information tag is implanted, corresponding to the state change set with a preset trajectory to obtain a comparison result.

Step S250, calculating abnormal state parameters corresponding to each state change track according to the equipment loss weight, and weighting the comparison result by adopting the abnormal state parameters to obtain an identification result; wherein the identification result is a series of numerical results.

It is understood that by performing the above steps S210 to S250, the device loss of the power device can be considered when identifying the state change set, thereby ensuring the reliability of the identification result.

In one possible embodiment, in order to ensure the foresight monitoring of the abnormal state of the power equipment 130, not only the identification result needs to be analyzed, but also the operation of the power equipment 130 needs to be simulated according to the identification result to predict the abnormal state of the power equipment 130, so as to deploy the state maintenance measures in advance. In order to achieve the above object, the determination of whether there is an abnormality in each electrical device and at least one group of power networks formed by the electrical devices according to the identification result described in step S150 may be specifically implemented by the following method described in step S151 to step S154.

In step S151, the failure occurrence rate of the power equipment corresponding to each identification result is determined based on the difference between every two adjacent identification values in the value result corresponding to each identification result.

Step S152, when the fault occurrence rate exceeds a first reference value, determining that the electrical equipment corresponding to the fault occurrence rate is abnormal; when the fault occurrence rate does not exceed the first reference value, determining a geographical distribution diagram corresponding to the fault occurrence rate according to the geographical position of each power device, and when the mean value of the fault occurrence rates in a set area in the geographical distribution diagram exceeds a second reference value, determining that the power utilization network corresponding to the set area is abnormal.

Step S153, when the mean value of the failure occurrence rates in the set area in the geographic distribution map does not exceed the second reference value, performing multi-dimensional feature clustering on the identification result to obtain a plurality of cluster sets.

Step S154, determining a simulation parameter set corresponding to each cluster set, and inputting the simulation parameter set into a preset state simulation thread to simulate the simulation running state of each power device; and acquiring a simulation identification result of each electric device and executing a step similar to the step of determining the fault occurrence rate of the electric device corresponding to each identification result based on the difference value between every two adjacent identification values in the numerical value result corresponding to each identification result.

It can be understood that through the above steps S151 to S154, not only the recognition result can be analyzed to determine whether the electrical equipment 130 is abnormal at the current moment, but also the operation of the electrical equipment 130 can be simulated according to the recognition result to predict an abnormal state that may occur to the electrical equipment 130, so as to deploy the state maintenance measures in advance according to the simulated recognition result.

In an alternative embodiment, in order to ensure the safe and reliable operation of the information analysis system 100, the resource configuration information of the artificial intelligence server 120 needs to be periodically detected, so as to avoid delay or error in the identification of the state change set caused by unreasonable resource configuration of the artificial intelligence server 120. To achieve the above object, on the basis of the above steps S110 to S150, the method may further include the following steps S310 to S350.

Step S310, sending a request instruction for calling the resource configuration information of the artificial intelligence server to the artificial intelligence server; the request instruction carries authentication information generated by the cloud computing platform according to a verification protocol established with the artificial intelligence server in advance.

Step S320, when receiving an authorization code sent by the artificial intelligence server based on the authentication information in the request instruction, packaging the authorization code into a call instruction and sending the call instruction to the artificial intelligence server.

Step S330, acquiring resource configuration information fed back by the artificial intelligence server based on the calling instruction, extracting configuration parameters and configuration time corresponding to each configuration record in the resource configuration information, and establishing response curves of the configuration parameters and the configuration time; the response curve is used for describing the resource configuration stability of the artificial intelligence server.

Step S340, segmenting the response curve based on the number of configuration records in the resource configuration information to obtain a plurality of curve segments, and extracting curve characteristics of each curve segment; wherein the curve characteristics include curve slope information and curve trajectory information.

Step S350, calculating a curve description value of an evaluation factor for representing each curve characteristic, and judging whether the current ratio of the median in the curve description value reaches a preset ratio; when the current occupation ratio reaches the preset occupation ratio, judging that a resource configuration thread of the artificial intelligence server is normal; and when the current ratio does not reach the preset ratio, judging that the resource configuration thread of the artificial intelligence server is abnormal and sending early warning information to the artificial intelligence server.

It can be understood that through the contents described in the above steps S310 to S350, the resource configuration information of the artificial intelligence server 120 can be periodically detected, and the early warning information can be sent to the artificial intelligence server 120 when the resource configuration thread of the artificial intelligence server 120 is abnormal, so as to avoid delay or error in the identification of the state change set caused by unreasonable resource configuration of the artificial intelligence server 120, and further ensure safe and reliable operation of the information analysis system 100.

On the basis of the above, please refer to fig. 3, a functional block diagram of the artificial intelligence and big data based information analysis apparatus 300 is provided, and the artificial intelligence and big data based information analysis apparatus 300 is described in detail as follows.

A1. An artificial intelligence and big data based information analysis apparatus 300 applied to a cloud computing platform communicating with an artificial intelligence server and a plurality of electric power devices, the apparatus at least comprising the following modules:

an information obtaining module 310, configured to obtain operating state information of a corresponding power device through an information transmission interface pre-established with each power device; the running state information is acquired when the sensor deployed in the corresponding power equipment runs and is uploaded to the cloud computing platform in real time through the information transmission interface;

the information analysis module 320 is configured to periodically analyze each set of acquired running state information to obtain a state change track and a plurality of information tags of each set of running state information in each set time period;

the information encapsulation module 330 is configured to encapsulate a state change trajectory of each set of running state information in each set time period and a plurality of information tags to obtain a state change set of each set of running state information; transmitting each state change set to the artificial intelligence server when detecting that the vacancy rate of the identification threads in the artificial intelligence server reaches a set rate;

a result obtaining module 340, configured to obtain a recognition result returned by the artificial intelligence server based on state recognition of each state change set by the recognition thread; the identification result comprises abnormal state parameters determined based on the equipment loss weight of the electric equipment;

and an anomaly monitoring module 350, configured to determine whether each electrical device and at least one group of power networks formed by the electrical devices are anomalous according to the identification result.

A2. The information analysis apparatus according to a1, the anomaly monitoring module 350 is configured to:

determining the fault occurrence rate of the power equipment corresponding to each identification result based on the difference value between every two adjacent identification values in the value result corresponding to each identification result;

when the fault occurrence rate exceeds a first reference value, determining that the electrical equipment corresponding to the fault occurrence rate is abnormal; when the fault occurrence rate does not exceed the first reference value, determining a geographical distribution diagram corresponding to the fault occurrence rate according to the geographical position of each power device, and when the mean value of the fault occurrence rate in a set area in the geographical distribution diagram exceeds a second reference value, determining that an electric network corresponding to the set area is abnormal;

when the mean value of the fault occurrence rates in the set area in the geographic distribution map does not exceed the second reference value, carrying out multi-dimensional feature clustering on the identification result to obtain a plurality of cluster sets;

determining a simulation parameter set corresponding to each cluster set, and inputting the simulation parameter set into a preset state simulation thread to simulate the simulation running state of each power device; and acquiring a simulation identification result of each electric device and executing a step similar to the step of determining the fault occurrence rate of the electric device corresponding to each identification result based on the difference value between every two adjacent identification values in the numerical value result corresponding to each identification result.

A3. The information analysis apparatus according to a1, the information packaging module 330, configured to:

acquiring a plurality of identification threads in an idle state in the artificial intelligence server;

judging whether the number of the identification threads is larger than or equal to the number of the state change sets;

on the premise that the number of the identification threads is larger than or equal to the number of the state change sets, transmitting each state change set to one identification thread of the artificial intelligence server;

on the premise that the number of the identification threads is smaller than the number of the state change sets, acquiring the information throughput of each identification thread; sequencing the identification threads according to the sequence of high and low information throughput to obtain a first sequencing sequence; sorting the state change sets according to the size sequence of the information capacity of the state change sets to obtain a second sorting sequence;

and sequentially transmitting at least part of state change sets in the second sorting sequence to corresponding identification threads in the first sorting sequence respectively.

A4. The information analysis apparatus according to a1, the information packaging module 330, configured to:

determining a first time sequence feature array of the state change track and a second time sequence feature array of each information tag of the state change track; each second time sequence feature array has different label weights, the label weights are used for representing the encoding heterogeneous degree of each time sequence feature array and the first time sequence feature data, the first time sequence feature array and each second time sequence feature array have the same number of time sequence feature fields, and each time sequence feature field corresponds to one field encoding string;

sequencing the second time sequence feature arrays according to the sequence of the label weights from large to small to obtain a feature array sequencing sequence, sequentially extracting a feature field distribution sequence of each second time sequence feature array in the sequencing sequence, and transforming each feature field distribution sequence through a preset sequence conversion list to obtain a corresponding target distribution sequence; determining a first sequence topology corresponding to a feature field distribution sequence of the first time sequence feature array and a second sequence topology of the target distribution sequence; the first sequence topology and the second sequence topology have the same logical directional connection information;

calculating topological structure similarity of the first sequence topology and each second sequence topology; the topological structure similarity is the weighted sum of the node similarity and the directed connecting line similarity of the first sequence topology and each second sequence topology;

and sequentially implanting the information labels corresponding to the second sequence topology into the track node containers corresponding to the state change tracks according to the sequence of the similarity of the topological structures from small to large so as to obtain the state change set of each group of running state information.

A5. The information analysis device according to a1, wherein the information analysis module 320 is configured to:

acquiring information attribute parameters corresponding to each set of running state information in each set time period and port information of a coupling port of the power equipment corresponding to each set of running state information, determining associated attribute parameters corresponding to the information attribute parameters according to the port information of the coupling port of the power equipment corresponding to each set of running state information, and extracting equipment identifications of other power equipment which have coupling relation with the power equipment corresponding to each set of running state information from the associated attribute parameters; the information attribute parameter is a parameter sequence determined based on an attribute identifier in the running state information, and a sequence coding logic corresponding to the parameter sequence is fixed and invariable;

determining wiring configuration information between the power equipment corresponding to each group of running state information and other power equipment having a coupling relation with the power equipment, and determining a state offset coefficient between an information attribute parameter corresponding to each group of running state information and the wiring configuration information corresponding to each group of running state information based on the association attribute parameter; the state offset coefficient is used for representing a coordination coefficient between the electric power equipment corresponding to each group of running state information and other electric power equipment;

judging whether a state offset coefficient between the information attribute parameter corresponding to each group of running state information and the wiring configuration information corresponding to each group of running state information is smaller than a preset coefficient or not;

if the state offset coefficient between the information attribute parameters corresponding to each group of running state information and the wiring configuration information corresponding to each group of running state information is smaller than the preset coefficient, mapping the associated attribute parameters corresponding to each group of running state information to the port information corresponding to each group of running state information to obtain the mapping attribute parameters of the associated attribute parameters in the port information, extracting the category of the mapping attribute parameters and determining the number of information labels of each group of running state information according to the first number of the category; determining target fields matched with each group of information fields in each group of running state information from a preset field extraction list, determining description information of each group of target fields, mapping each group of target fields to a preset coordinate plane according to a time sequence grade included in the description information to obtain corresponding field nodes, fitting the field nodes to obtain state change tracks corresponding to each group of running state information, and determining a plurality of information labels according to track characteristics of the state change tracks.

A6. The information analysis device according to a5, wherein the information analysis module 320 is further configured to:

if the state offset coefficient between the information attribute parameter corresponding to each group of running state information and the wiring configuration information corresponding to each group of running state information is greater than or equal to the preset coefficient, calculating the port loss percentage of the coupling port of the power equipment corresponding to each group of running state information, correcting each group of information fields corresponding to each group of running state information according to the port loss percentage to obtain a target field, and determining the number of information labels according to the similarity between the target field and the corresponding information field; and mapping the target field to a preset coordinate plane to obtain a corresponding field node, fitting the field nodes to obtain a state change track corresponding to each group of running state information, and determining a plurality of information labels according to the track characteristics of the state change track.

A7. The information analysis apparatus according to a1, wherein the information acquisition module 310 is configured to:

reading geographical position information and electrical position information of corresponding power equipment through each information transmission interface, extracting direction information between the geographical position information and the electrical position information corresponding to each power equipment, and constructing a relevance connection topology of the power equipment through the direction information; the association degree connection topology is a multi-area network, each area network corresponds to one area electrical level, each area network is provided with at least one electric power device, one device in each area network is electrically connected with one electric power device in at least one other area network, and the area electrical levels have a sequence from high to low;

when the running state information of the corresponding power equipment is acquired through each information transmission interface, if two groups of running state information are in the same area network, determining a first similarity of geographic position information and a second similarity of electrical position information between the two groups of running state information; when the first similarity and the second similarity are both greater than a set similarity, determining a plurality of first information codes of first running state information and a plurality of second information codes of second running state information in the two groups of running state information;

determining the coincidence rate between each second information code in the second running state information and the first information code at the same code position in the first running state information according to each first information code in the first running state information and the code registration information thereof, and marking a first target information code and a second target information code, of which the coincidence rate is greater than a preset rate;

determining a first degree of association of the first target information code in the first running state information and a second degree of association of the second target information code in the second running state information, setting a first signature for the first running state information based on the first degree of association and setting a second signature for the second running state information based on the second degree of association; wherein the first signature and the second signature are different digital signatures.

A8. According to the information analysis apparatus described in a1, the artificial intelligence server identifies the state change set by:

determining path information of each information label in the state change set in a track node container of a corresponding state change track;

determining a first state list before the information tag is implanted and a second state list after the information tag is implanted of a state change track corresponding to each state change set according to the path information;

determining a device loss weight for the power device corresponding to each state change set based on the list content at the same list position in the first and second state lists;

comparing the state change track implanted with the information tag corresponding to the state change set with a preset track to obtain a comparison result;

calculating abnormal state parameters corresponding to each state change track according to the equipment loss weight, and weighting the comparison result by adopting the abnormal state parameters to obtain an identification result; wherein the identification result is a series of numerical results.

A9. The information analysis apparatus according to a1, further comprising a configuration detection module 360 for:

sending a request instruction for calling the resource configuration information of the artificial intelligence server to the artificial intelligence server; the request instruction carries authentication information generated by the cloud computing platform according to a verification protocol established with the artificial intelligence server in advance;

when receiving an authorization code sent by the artificial intelligence server based on the authentication information in the request instruction, packaging the authorization code into a calling instruction and sending the calling instruction to the artificial intelligence server;

acquiring resource configuration information fed back by the artificial intelligence server based on the calling instruction, extracting configuration parameters and configuration time corresponding to each configuration record in the resource configuration information, and establishing response curves of the configuration parameters and the configuration time; the response curve is used for describing the resource configuration stability of the artificial intelligence server;

segmenting the response curve based on the number of configuration records in the resource configuration information to obtain a plurality of curve segments, and extracting curve characteristics of each curve segment; wherein the curve characteristics include curve slope information and curve trajectory information;

calculating a curve description value of an evaluation factor for representing each curve characteristic, and judging whether the current occupation ratio of a median in the curve description value reaches a preset occupation ratio or not; when the current occupation ratio reaches the preset occupation ratio, judging that a resource configuration thread of the artificial intelligence server is normal; and when the current ratio does not reach the preset ratio, judging that the resource configuration thread of the artificial intelligence server is abnormal and sending early warning information to the artificial intelligence server.

Based on the same inventive concept, an information analysis system based on artificial intelligence and big data is also provided, which is described in detail as follows.

B1. An information analysis system based on artificial intelligence and big data comprises a cloud computing platform, an artificial intelligence server and a plurality of electric power devices, wherein the cloud computing platform is communicated with the artificial intelligence server and the plurality of electric power devices;

the cloud computing platform is to:

the artificial intelligence server is configured to:

performing state recognition on each state change set through the recognition thread to obtain a recognition result and returning the recognition result to the cloud computing platform;

the cloud computing platform is to:

B2. The information analysis system of B1, the cloud computing platform further to:

B3. The information analytics device system of B1, the cloud computing platform being further to:

B4. The information analysis system of B1, the cloud computing platform further to:

B5. The information analysis system of B1, the cloud computing platform further to:

B6. The information analysis system of B5, the cloud computing platform further to:

B7. The information analysis system of B1, the cloud computing platform further to:

B8. According to the information analysis system of B1, the artificial intelligence server specifically identifies the set of state changes by:

B9. The information analysis system of B1, the cloud computing platform further to:

On the basis of the method, the device and the system, a cloud computing platform 110 as shown in fig. 4 is further provided, and includes a processor 111 and a memory 112 which are communicated with each other, and the processor 111 reads a computer program from the memory 112 and implements the method by running the computer program. Further, a readable storage medium applied to a computer is provided, and the readable storage medium is burned with a computer program, and the computer program realizes the above method when running in the processor 111 of the cloud computing platform 110.

Claims

1. An information analysis method based on artificial intelligence and big data is applied to a cloud computing platform which is communicated with an artificial intelligence server and a plurality of electric power devices, and the method at least comprises the following steps:

wherein:

the running state information of different power equipment is different;

the power equipment comprises a power station boiler, a steam turbine, a gas turbine, a water turbine, a generator, a transformer, a mutual inductor and a contactor;

determining whether each electrical device and at least one group of electricity utilization networks formed by the electrical devices have abnormality according to the identification result;

wherein, determining whether each electric device and at least one group of power utilization networks formed by the electric devices have abnormality according to the identification result comprises: determining the fault occurrence rate of the power equipment corresponding to each identification result based on the difference value between every two adjacent identification values in the value result corresponding to each identification result; when the fault occurrence rate exceeds a first reference value, determining that the electrical equipment corresponding to the fault occurrence rate is abnormal; when the fault occurrence rate does not exceed the first reference value, determining a geographical distribution diagram corresponding to the fault occurrence rate according to the geographical position of each power device, and when the mean value of the fault occurrence rates in a set area in the geographical distribution diagram exceeds a second reference value, determining that the power utilization network corresponding to the set area is abnormal.

2. The information analysis method of claim 1, wherein transmitting each set of state changes to the artificial intelligence server comprises:

3. A cloud computing platform in communication with an artificial intelligence server and a plurality of electrical devices, the cloud computing platform to:

wherein:

the running state information of different power equipment is different;

wherein the cloud computing platform is specifically configured to:

when the fault occurrence rate exceeds a first reference value, determining that the electrical equipment corresponding to the fault occurrence rate is abnormal; when the fault occurrence rate does not exceed the first reference value, determining a geographical distribution diagram corresponding to the fault occurrence rate according to the geographical position of each power device, and when the mean value of the fault occurrence rates in a set area in the geographical distribution diagram exceeds a second reference value, determining that the power utilization network corresponding to the set area is abnormal.

4. The cloud computing platform of claim 3, wherein the cloud computing platform is specifically configured to: