CN115858316B - Multi-Agent-based networked software system reliability modeling simulation method - Google Patents
Multi-Agent-based networked software system reliability modeling simulation method Download PDFInfo
- Publication number
- CN115858316B CN115858316B CN202211469080.6A CN202211469080A CN115858316B CN 115858316 B CN115858316 B CN 115858316B CN 202211469080 A CN202211469080 A CN 202211469080A CN 115858316 B CN115858316 B CN 115858316B
- Authority
- CN
- China
- Prior art keywords
- simulation
- reliability
- software
- model
- software system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention discloses a networking software system reliability modeling simulation method based on multiple agents, which comprises the following steps: identifying services and connectors of the networked software system, abstracting the services into nodes, and abstracting the connectors into edges; establishing individual agents for each service and each connecting piece respectively; according to the static structure of each Agent, the executed software service and the interaction relation with other agents, obtaining the network topology relation among the agents, and establishing a reliability model of the networked software system; and simulating and evaluating the fault propagation condition, fault repair capability and system reliability of the reliability model of the networked software system by using the simulation platform to obtain the quality and reliability level of the networked software system. The invention can dynamically simulate the structure and the behavior of the networked software system, find out the software defects and weak links, and promote the quality and the reliability level of the networked software system.
Description
Technical Field
The invention relates to the technical field of software reliability engineering, in particular to a multi-Agent-based networking software system reliability modeling simulation method.
Background
A networked software system is a complex system that exhibits layering, openness, and nonlinearity in terms of functionality, behavior, and architecture. The single subsystem is autonomous and reactive, i.e., has relative independence, initiative, adaptivity, and the ability to perceive external operating and use environments and provide useful information to the system evolution; the subsystems have cooperativity and evolution, namely, the subsystems are interconnected, intercommunicated, cooperated and allied, and the capability of dynamically adjusting functions, structures and behaviors according to requirements, and finally, the networked system with the ordered structure can be formed in a self-organizing mode in space, time, behaviors and functions.
The reliability of the networked software system mainly comprises research methods based on state analysis, path and test data. State-based reliability models represent the architecture of the system and component transitions through control flow graphs, and because of component independence, the future behavior conditions of the system are usually independent of past behavior, so that the software inter-component transitions have Markov characteristics, and the reliability is analyzed and calculated by using Markov theory; the reliability is evaluated based on the reliability model of the path according to all possible execution paths and related frequencies of the system; the reliability model based on the test is based on failure data of components in the system, the software reliability growth model is applied, the system structure of the networked software system is not considered, and meanwhile, the test result is also influenced by different test environments and methods.
Foreign and other scholars have proposed a reliability calculation method based on the composition of structural components, and the application of the composition of the structural components reduces the complexity of reliability calculation, but does not consider the analysis and acquisition of the networked software structure. And a learner introduces monitoring and feedback to realize the reliability evaluation of the networked software, but the calculation process is complex, and the effective formalized analysis modeling and the system reliability calculation are lacked. Considering that more intensive researches on component failure correlation, fault tolerant configuration and the like have appeared, a software reliability model in a wider sense is described, but obvious characteristics of networked software reliability analysis are not reflected. The domestic research is only limited to the reliability evaluation and improvement based on the state model, and extensive and intensive research has not been developed yet.
Through the analysis of the research results, the reliability is difficult to accurately calculate by a test-based method, the reliability is influenced by the system operation and the test environment, and the accuracy of the results is low; the reliability analysis and calculation method based on the structure analysis sets the system structure information and the transition probability to be fixed, which does not accord with the actual situation of the networked software, lacks further research on the acquisition of the structure and the calculation of the transition probability, and does not consider the influence of the network environment from the calculation method. In summary, the existing architecture model building mode is mainly aimed at single software, the system structure and the behavior dynamic evolution characteristics of the networked software cannot be described, and a new modeling simulation method and a new building mechanism need to be researched. These problems restrict the reliability analysis, test and evaluation work of the networked software system, and influence the performance improvement of the active equipment and the development of new generation digital armed equipment.
Therefore, how to provide a multi-Agent-based networking software system reliability modeling simulation method capable of significantly improving the quality and reliability level of the networking software system is a problem that needs to be solved by those skilled in the art.
Disclosure of Invention
In view of the above, the invention provides a multi-Agent-based networking software system reliability modeling simulation method, which can dynamically simulate the structure and the behavior of a networking software system and analyze and evaluate the quality and the reliability level of the networking software system.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a networking software system reliability modeling simulation method based on multiple agents comprises the following steps:
identifying services and connectors of the networked software system, abstracting the services into nodes, and abstracting the connectors into edges;
establishing individual agents for each service and each connection piece respectively;
according to the static structure of each Agent, the executed software service and the interaction relation with other agents, obtaining the network topology relation among the agents, and establishing a reliability model of the networked software system;
and simulating and evaluating the fault propagation condition, fault repair capability and system reliability of the networked software system reliability model by using a simulation platform to obtain the quality and reliability level of the networked software system.
Further, the Agent includes: service agents and connector agents;
the service Agent receives the needed information from the external network environment, integrates or judges the received information according to the internal self state, and generates a description for modifying the current state of the service Agent; then matching information in the knowledge base to make a plan, and generating a series of actions acting on the external network environment according to the target;
the connector Agent receives needed information from the external network environment, integrates the information, performs action reaction according to an internal rule base, and feeds the processed information back to the external network environment.
Further, the service Agent is modeled using a cognitive Agent, the formal description of which is represented by a six-tuple:
Ser_Agent::＝<Seri,Ser_S,Ser_P,Ser_Per,Ser_Trans,Ser_KB,Ser_PS,Ser_GS,Ser_AS>
wherein, seri represents the serial number of the service Agent; ser_s represents a self state set; ser_p represents a perception set; ser_Per represents a sensing function of E-Ser_P, the environmental state is mapped into sensing input, and E represents an environmental state set; ser_Trans represents a decision function of Ser_P, ser_S- & gt Ser_S, and service change is realized according to the current state of the sensing input; ser_KB represents a knowledge base, which is a knowledge set of service agents, and at least contains knowledge of the running environment, action knowledge of the agents, and target knowledge of the service agents; ser_ps represents a planned set of services; ser_gs represents a service target set; ser_as represents the action set of the service.
Further, the connector Agent is modeled by using the structure of the reactive Agent, and the formal description of the connector Agent is represented by a quadruple:
Con_Agent::＝<Coni,Con_set,Con_KB,Con_AS>
wherein Coni represents the serial number of the connector Agent; con_set represents a state set of the connector Agent and at least comprises reliability and state information of the connector; con_KB represents the knowledge base of connector agents; con_AS represents the set of actions of the connector Agent.
Further, the establishing process of the reliability model of the networked software system comprises the following steps:
establishing an environment operation model of the networked software system: the environment information to be established comprises a networked software system use profile, a software target use probability, hardware operation information and actions of a user;
establishing a generating model of a software target: any one Agent participating in the networked software system acts, and different software targets are issued by the same Agent or issued by a plurality of agents together;
establishing a software service call model: according to the requirements of the software target, each Agent decides whether to participate in the establishment of the software target according to the functions and the capabilities of the Agent; when the software target appears or reaches an Agent, if the Agent can independently complete the software target, the software target is completed, and the next software target is accessed; when the software target appears or reaches an Agent, if the Agent can not independently complete the software target, other agents related to the Agent are analyzed, and whether the software target can be jointly solved is judged; if the two agents can be completed together, forming the agents into a set, and starting the next step to perform a dynamic coordination process; if the software target cannot be completed together, discarding the software target, and feeding back that the software target is not completed;
establishing a dynamic coordination relation model: acquiring related information between an Agent set of the established software target and other agents, and coordinating and interacting to jointly complete the software target;
establishing a reliability measurement model: measuring the specific number of software targets that the networked software system can complete a user within a period of time;
and integrating the five models as the reliability model of the networked software system.
Furthermore, the simulation platform simulates the number of software targets used by a user of the networked software system by using a random number generation method, and specifically comprises the following steps: in the running time of the simulation platform, dividing the simulation time into six time periods, and respectively corresponding to the actual running time of the networked software system; the number of software targets provided by each stage of the networked software system is divided into three cases of small quantity, medium quantity and large quantity, and the number of the software targets completed in unit time is unequal.
Furthermore, the simulation platform also simulates the number of software services forming the software business by using a Monte Carlo method; the method comprises the following steps: the software services are divided into three types of simple services, general services and complex services, wherein the number of software services contained in each type of services is different, the number of software services contained in the complex services is the largest, and the number of software services contained in the simple services is the smallest.
Further, the simulation model adopted by the simulation platform is as follows: a fault propagation simulation model and a reliability simulation model;
the fault propagation simulation model is modeling based on a virus propagation model, and single-time or cyclic simulation is carried out on the fault propagation process in the networked software system by setting the rule and the strength of fault propagation in the networked software system; in the simulation process, parameters to be set at least comprise fault propagation probability, node fault detection probability and fault node recovery probability; obtaining reliability related data of the networked software system after the simulation is finished, and simulating and evaluating the capability of the networked software system for bearing software faults and repairing;
the reliability simulation model performs single or cyclic simulation on the fault condition of the internal node of the software when the networked software system executes various different tasks according to the reliability definition, calculates the structural reliability and the task reliability of the software, and simulates and evaluates the reliability of the networked software system when executing different tasks in the use process.
Further, the simulation platform further comprises: the system comprises a parameter setting module, a service simulation and state management module, a visual display module and a simulation process display module;
setting the type of the simulation model and part of index parameters in the simulation process by the parameter setting module;
simulating a development environment through the service simulation and state management module, and starting and controlling corresponding simulation functions of the selected simulation model according to the set parameters;
checking the values and states of key indexes of the reliability model of the networked software system at each simulation time point through the visual display module;
and displaying each parameter set for the selected simulation model, state change chart information of the reliability model of the networked software system in the simulation process, simulation result information when the simulation is completed and the state of the reliability model of the networked software system through the simulation process display module.
Further, after the single simulation of the fault propagation simulation model is finished, the node proportion of faults generated in the networked software system, the node proportion with fault resistance, the node proportion of faults still possible to occur, and the value and the change condition of each simulation time point are displayed through the visual display module;
after the cyclic simulation of the reliability simulation model is finished, the structural reliability and task reliability of the networked software system after the simulation is started are displayed through the visual display module, the development and change conditions of the structural reliability and the task reliability along with the simulation time are output, and the task sequence when the simulation is executed, the node sequence of each task in the execution sequence and the fault condition of the node sequence are output.
Compared with the prior art, the invention discloses a multi-Agent-based networking software system reliability modeling simulation method, which can model the structure and reliability of the networking software system, conduct simulation analysis, effectively develop the analysis of the defect propagation and failure mechanism of the networking software system and the reliability modeling simulation analysis, complete the reliability analysis and evaluation work of the networking software system, complete the discovery and exposure of the software defects, and remarkably improve the quality and reliability level of the networking software system.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a multi-Agent-based networked software system reliability modeling simulation method provided by the invention;
FIG. 2 is a single simulation run interface of a reliability model of a networked software system employing a fault propagation model in accordance with an embodiment of the present invention;
FIG. 3 is a configuration area interface in a single simulation process using a fault propagation model provided by the present invention;
FIG. 4 is a data output area interface in a single simulation process using a fault propagation model provided by the present invention;
FIG. 5 is a variation interface of various statistical indexes and partial statistical indexes in a single simulation process using a fault propagation model;
FIG. 6 is a state display interface of a reliability model of a networked software system in a single simulation process using a fault propagation model provided by the invention;
FIG. 7 is a diagram showing an interface for single simulation results using a fault propagation model provided by the present invention;
FIG. 8 is a diagram of a loop simulation run interface employing a fault propagation model provided by the present invention;
FIG. 9 is a view showing an interface for a simulation result of a cycle using a fault propagation model according to the present invention;
FIG. 10 is a diagram of a reliability model status display interface for a networked software system simulated using a reliability simulation model provided by the present invention;
FIG. 11 is a diagram of an interface for operating a simulation using a reliability simulation model provided by the present invention;
FIG. 12 is a diagram of a data output area interface for simulation using a reliability simulation model provided by the present invention;
FIG. 13 is a diagram showing the reliability change of the simulation using the reliability simulation model according to the present invention;
fig. 14 is a simulation result display interface for simulation using a reliability simulation model provided by the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, the embodiment of the invention discloses a multi-Agent-based networking software system reliability modeling simulation method, which comprises the following steps:
s1, identifying services and connecting pieces of a networked software system, abstracting the services into nodes and abstracting the connecting pieces into edges;
establishing individual agents for each service and each connecting piece respectively;
s2, obtaining a network topological relation among the agents according to the static structure of each Agent, the executed software service and the interaction relation with other agents, and establishing a reliability model of the networked software system;
s3, simulating and evaluating the fault propagation condition, fault repair capability and system reliability of the networked software system reliability model by using the simulation platform to obtain the quality and reliability level of the networked software system reliability model.
The networked software system is composed of software services for realizing specific targets, is a basis for reliability analysis and is mainly measured from the aspects of continuity, reliability and the like of the services provided by the system to the user from the viewpoint of the user.
The architecture model for service oriented architecture is made up of a large number of services and connectors. In the modeling process, services are abstracted into nodes, and connectors are abstracted into edges, so that a topological network is formed. The subsystems are connected with each other by connecting pieces to form a complex network. Meanwhile, the service and the connecting piece are abstracted into agents with different structure types for representation, so that the nodes and edges of the complex network have intelligence, and the multi-Agent macroscopic mode is effectively analyzed.
The invention takes the Agent as a basic abstract unit of the target networked software system, adopts a related abstract technology, firstly establishes an Agent model of each individual forming the networked software system, then adopts a proper topological relation to assemble the individual agents, and finally establishes a system model of the whole networked software system.
The behavior of each Agent can affect the reliability of the networked software system, and can change the state of other agents. By researching static structures and dynamic relations among different kinds of Agent entities, the running mechanism and reliability change of the networked software system can be analyzed.
Thus, a networked software system can be used as a target system and is decomposed into a system composed of a plurality of individual services and connectors, and the name of each individual Agent, the executed software service and the interaction relation between the individual agents and other agents can be determined.
The above steps are further described below.
In S1, the Agent includes: service agents and connector agents;
(1) Modeling process of service Agent:
the service Agent receives the needed information from the external network environment, integrates or judges the received information according to the internal self state, and generates a description for modifying the current state of the service Agent; and then matching information in the knowledge base to make a plan, and generating a series of actions acting on the external network environment according to the target. Service agents are modeled using cognitive agents, whose formal description is represented by a six-tuple:
Ser_Agent::＝<Seri,Ser_S,Ser_P,Ser_Per,Ser_Trans,Ser_KB,Ser_PS,Ser_GS,Ser_AS>
wherein, seri represents the serial number of the service Agent; ser_s represents a self state set; ser_p represents a perception set; ser_Per represents a sensing function of E-Ser_P, the environmental state is mapped into sensing input, and E represents an environmental state set; ser_Trans represents a decision function of Ser_P, ser_S- & gt Ser_S, and service change is realized according to the current state of the sensing input; ser_KB represents a knowledge base, which is a knowledge set of service agents, and at least contains knowledge of the running environment, action knowledge of the agents, and target knowledge of the service agents; ser_ps represents a planned set of services; ser_gs represents a service target set; ser_as represents the action set of the service.
(2) Modeling process of connector Agent:
the connector Agent is modeled by the structure of the reactive Agent, and the formal description of the connector Agent is represented by a four-tuple:
Con_Agent::＝<Coni,Con_set,Con_KB,Con_AS>
wherein Coni represents the serial number of the connector Agent; con_set represents a state set of the connector Agent and at least comprises reliability and state information of the connector; con_KB represents the knowledge base of connector agents; con_AS represents the set of actions of the connector Agent.
In a specific embodiment, in S2, the reliability change of the networked software system is analyzed by studying interaction relations between agents. Each Agent behavior can affect the reliability of the networked software system, and can change the states of other agents. In the embodiment of the invention, the modeling process of the networked system software mainly builds a model from the aspects of operation environment, software service generation and service measurement, and the specific modeling process is as follows:
s21, establishing an environment operation model of the networked software system: the environment information to be established includes the network software system usage profile, the probability of software target usage, the hardware running information and the actions of the user.
S22, establishing a generating model of the software target: any one Agent participating in the networked software system acts, and different software targets are issued by the same Agent or commonly issued by a plurality of agents.
S23, establishing a software service call model: according to the requirements of the software target, each Agent decides whether to participate in the establishment of the software target according to the functions and the capabilities of the Agent; when the software target appears or reaches an Agent, if the Agent can independently complete the software target, the software target is completed, and the next software target is accessed; when the software target appears or reaches an Agent, if the Agent can not independently complete the software target, other agents related to the Agent are analyzed, and whether the software target can be jointly solved is judged; if the two agents can be completed together, forming the agents into a set, and starting the next step to perform a dynamic coordination process; if the software objects cannot be completed together, the software objects are abandoned and incomplete software objects are fed back.
S24, establishing a dynamic coordination relation model: and acquiring related information between the Agent set of the established software target and other agents, and coordinating and interacting to jointly complete the software target.
S25, establishing a reliability measurement model: the networked software system can measure a specific number of software objects that can be completed by the user within a period of time.
And integrating the five models to be used as a reliability model of the networked software system.
In a specific embodiment, after the reliability model of the networked software system is built, the simulation platform is utilized to simulate the reliability model of the networked software system in the following two aspects:
(1) The simulation platform simulates the number of software targets used by a user of the networked software system by utilizing a random number generation method, and specifically comprises the following steps: in the running time of the simulation platform, dividing the simulation time into six time periods, and respectively corresponding to the actual running time of the networked software system; the number of software targets provided by each stage of the networked software system is divided into three cases of small quantity, medium quantity and large quantity, and the number of the software targets completed in unit time is unequal. Namely: in the time of executing a plurality of software targets, the number of the software targets executed by the system concurrently is more; the number of software objects that the system concurrently executes is small in a small amount of time to execute the software objects.
(2) The simulation platform also simulates the number of software services forming the software business by using a Monte Carlo method; the method comprises the following steps: the software services are divided into three types of simple services, general services and complex services, wherein the number of software services contained in each type of services is different, the number of software services contained in the complex services is the largest, and the number of software services contained in the simple services is the smallest. A system modeling and simulation lacking reliability is meaningless, and the reliability guarantee is more important for a complex simulation system.
Specifically, the invention realizes the platform for modeling the reliability of the networked software system based on the Netlogo development environment, can perform multi-agent modeling on the networked software structure, designs various simulation models and strategies for performing multi-angle reliability simulation analysis on the networked software system, and is an integrated software platform for fault propagation analysis, reliability modeling simulation and evaluation of the networked software system. The method can carry out fault propagation modeling simulation and reliability modeling simulation on the networked software system based on the multi-Agent technology of the Netlogo platform to obtain the fault propagation condition and the repair capability of the networked software system and the change condition of the system structure reliability and the task reliability, and evaluate and predict the quality and the reliability level of the target networked software system. The platform mainly uses two types of simulation models, namely a fault propagation simulation model and a reliability simulation model, and is specifically as follows:
(1) The fault propagation simulation model is modeling based on a virus propagation model, and single or cyclic simulation is carried out on the fault propagation process in the networked software system by setting the rule and the strength of fault propagation in the networked software system; in the simulation process, parameters to be set at least comprise fault propagation probability, node fault detection probability and fault node recovery probability; and after the simulation is finished, obtaining the reliability related data of the networked software system, and simulating and evaluating the capability of the networked software system for bearing the software fault and repairing.
(2) The reliability simulation model performs single or cyclic simulation on the fault condition of the internal node of the software when the networked software system executes various different tasks according to the reliability definition, calculates the structural reliability and the task reliability of the software, and performs simulation and evaluation on the reliability of the networked software system when executing different tasks in the use process.
In other embodiments, the simulation platform further comprises: the system comprises a parameter setting module, a service simulation and state management module, a visual display module and a simulation process display module;
and setting the type of the simulation model and part of index parameters in the simulation process by a parameter setting module.
And simulating a development environment through the service simulation and state management module, and starting and controlling the corresponding simulation function of the selected simulation model according to the set parameters.
And checking the values and the states of key indexes of the reliability model of the networked software system at each simulation time point through a visual display module. After the single simulation of the fault propagation simulation model is finished, the node proportion of faults generated in the networked software system, the node proportion with fault resistance, the node proportion of faults still possible to happen, and the value and the change condition of each simulation time point are displayed through a visual display module. After the cyclic simulation of the reliability simulation model is finished, the structural reliability and task reliability of the networked software system after the simulation is started are displayed through a visual display module along with the development and change conditions of simulation time, and a task sequence when the simulation is executed, a node sequence of each task in the execution sequence and the fault condition of the node sequence are output.
And displaying each parameter set for the selected simulation model, state change chart information of the reliability model of the networked software system in the simulation process, simulation result information when the simulation is completed and the state of the reliability model of the networked software system through a simulation process display module.
The modeling simulation method of the present invention will be further described with specific examples.
According to the method, the application implementation of the case is completed aiming at a weapon system command networking software system of a certain model, the connection relation among nodes of the weapon system is analyzed according to the topological diagram of the networking software system structure, a networking software system structure model and a reliability simulation model are constructed, the simulation analysis is carried out by utilizing a fault propagation model and a software reliability model, the analysis and evaluation of the software reliability are completed, and the rationality and feasibility of the proposed method are verified.
1. Single-pass simulation using fault propagation model
(1) After the simulation platform is opened and the menu bar selects the required simulation model, the main interface below will switch to the corresponding parameter setting interface (left side) and system state interface (right side), as shown in fig. 2. Firstly, selecting a fault propagation model, and setting various model parameters in the model on a parameter setting interface. The right system state interface is initially displayed as an initial state before the simulation of the reliability model of the networked software system, and is used for displaying the reliability model structure of the networked software system.
After setting the model parameters, clicking a button of the 'do simulation' type can open the Netlogo model and start simulation, executing single simulation of the fault propagation model, clicking the 'do single simulation', and opening a Netlogo model interface and automatically starting simulation after 10-20 s.
(2) The upper left is a model arrangement area, and as shown in fig. 3, the model parameters set in the model parameter setting section and the simulation function buttons included in the model can be seen. In general, the model can be automatically simulated after being opened, and if the model needs to be simulated again, the model can be simulated by clicking a setup button for initialization and clicking a corresponding simulation model button.
(3) The lower left is the data output area as shown in fig. 4. The part outputs various information of the simulation process, and in the single simulation process of the fault propagation model, the area outputs the node name of the injected fault when the simulation is started and the node name of the fault state in each simulation time step system.
(4) The middle of the interface is the second part model data output area, as shown in fig. 5. The main output of the part is the variation of various statistical indexes and part of statistical indexes in the model simulation process. In a single simulation of the fault propagation model, the output of this region is the number of faulty nodes, the proportion of faulty nodes, the number of resistant nodes, the proportion of resistant nodes, the number of nodes that can fail, and the proportion of nodes that can fail. In the lower statistical chart, three analogy node ratios are shown as the simulation progresses, and it can be seen that the simulation ends after the failed node ratio has decreased to 0.
(5) The portion to the right of the interface is a system status presentation area, as shown in fig. 6. The area can display the current state of the simulation system, taking a single simulation of the fault propagation model as an example, after the simulation is finished, the green nodes in the reliability model of the networked software system are nodes with fault resistance, the blue nodes are nodes which can still generate faults, and if red nodes occur, the nodes are fault nodes.
(6) After the simulation is completed, clicking a button for outputting the simulation result on the main interface, and popping up a simulation result display interface on the right side of the main interface, as shown in fig. 7. The statistical diagram in the simulation process is generally reproduced above the simulation result display interface, but the data display is clearer, and the statistical diagram can also perform functions such as zooming in and out, information display and the like. In a single simulation process of the fault propagation model, the data in the statistical graph is the change condition of the node duty ratio of three states in the system. The specific information display of each node in the upper statistical chart is shown below, and the state change condition of each simulation time step system can be checked in the table through adjusting the scroll bar. The above is the use flow of the networked software system in the simulation platform, and the use flow is similar in different models, but the input, output and function of each part are different.
2. Loop simulation using fault propagation model
Similar to the single simulation, the fault propagation model is still selected at the parameter setting interface (left side) and the system state interface (right side), and each model parameter built in the model is set. After setting the model parameters, clicking a button of the 'simulation' type can open the Netlogo model and start simulation, at the moment, carrying out cyclic simulation, clicking the 'cyclic simulation', opening a Netlogo model interface and automatically starting simulation after 10-20s, and completing the interface as shown in figure 8.
In the cyclic simulation process of the fault propagation model, the output of the data output area is the node name of the injection fault when the simulation is started, and the time steps and the real time which are passed from the injection fault to the maintenance completion of each round of simulation are calculated, so that the real time required by each time step is calculated. The statistics show that the average repair time of the system changes with the increase of the cycle number, and the simulation can be seen to end after the simulation of the designated cycle number is completed. Clicking the "output simulation result" button of the corresponding function on the main interface can pop up the simulation result display interface on the right side of the main interface, as shown in fig. 9.
In the cyclic simulation process of the fault propagation model, the data in the statistical graph is the change condition of the average repair time of the system, and the state change condition of the system in each simulation time step can be checked in the table by adjusting the rolling bar.
3. Software model reliability simulation analysis
The reliability modeling simulation platform of the networked software system is opened, a reliability simulation model is selected in a menu bar, and a main interface becomes a model parameter setting interface and a system state display interface of the reliability simulation model, as shown in fig. 10. In the parameter setting interface, various model parameters built in the model can be set, such as reliability of various nodes of each organization, task failure judgment standards and the like. The right system state interface is initially displayed as an initial state before the simulation of the reliability model of the networked software system and is used for displaying the reliability model structure of the networked software system.
(2) After setting the model parameters, clicking a button for 'simulation' class can open the NetLogo model and start simulation, and the completion interface is shown in fig. 11.
(3) In the cyclic simulation process of the reliability simulation model, the data output area outputs the number of nodes participating in task execution, the number of nodes having faults when executing the tasks, the total number of the executed tasks and the number of the fault tasks in the whole simulation process, as shown in fig. 12, and meanwhile, the structural reliability and the task reliability of the reliability model of the networked software system are calculated according to the structural reliability and the task reliability.
(4) The middle part is a second part simulation result output area, a task sequence being executed in the simulation process, the names of nodes executing the tasks and the node fault conditions are output, after the execution of each task is finished, execution summary of the task is output, the execution summary comprises the number of nodes executing the task, the number of fault nodes, the names of the fault nodes, the execution time of the tasks and the number of simulation steps, and meanwhile, the time required by single-step simulation is calculated according to the calculation result. The lower statistical chart shows the situation that the system structure reliability and the task reliability change along with the simulation process, and the simulation is finished after the simulation of the designated task number is executed as shown in fig. 13.
(5) After the simulation is completed, clicking a button for outputting the simulation result on the main interface, and popping up a simulation result display interface on the right side of the main interface, as shown in fig. 14. In the cyclic simulation process of the reliability simulation model, the data in the statistical graph is the change condition of the reliability of the structure and the reliability of the task of the reliability model of the networked software system. The specific information display of each node in the upper statistical chart is shown below, and the reliability change condition of each simulation time step system can be checked in the table through adjusting the scroll bar. At the bottom, the state of the networked software system reliability model at the end of the simulation is shown, along with the simulation process description above.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (7)
1. A multi-Agent-based networking software system reliability modeling simulation method is characterized by comprising the following steps:
identifying services and connectors of the networked software system, abstracting the services into nodes, and abstracting the connectors into edges;
establishing individual agents for each service and each connection piece respectively;
according to the static structure of each Agent, the executed software service and the interaction relation with other agents, obtaining the network topology relation among the agents, and establishing a reliability model of the networked software system;
simulating and evaluating the fault propagation condition, fault repair capability and system reliability of the networked software system reliability model by using a simulation platform to obtain the quality and reliability level of the networked software system;
the establishment process of the reliability model of the networked software system comprises the following steps:
establishing a software service call model: according to the requirements of the software target, each Agent decides whether to participate in the establishment of the software target according to the functions and the capabilities of the Agent; when the software target appears or reaches an Agent, if the Agent can independently complete the software target, the software target is completed, and the next software target is accessed; when the software target appears or reaches an Agent, if the Agent can not independently complete the software target, other agents related to the Agent are analyzed, and whether the software target can be jointly solved is judged; if the two agents can be completed together, forming the agents into a set, and starting the next step to perform a dynamic coordination process; if the software target cannot be completed together, discarding the software target, and feeding back that the software target is not completed;
the simulation platform simulates the number of software targets used by a user of the networked software system by utilizing a random number generation method, and specifically comprises the following steps: in the running time of the simulation platform, dividing the simulation time into six time periods, and respectively corresponding to the actual running time of the networked software system; dividing the number of software targets provided by each stage of the networked software system into three cases of small quantity, medium quantity and large quantity, wherein the number of the software targets completed in unit time is unequal;
the simulation platform also simulates the number of software services forming the software service by using a Monte Carlo method; the method comprises the following steps: dividing the software services into three types of simple services, general services and complex services, wherein the number of software services contained in each type of services is unequal, the number of software services contained in the complex services is the largest, and the number of software services contained in the simple services is the smallest;
the simulation model adopted by the simulation platform is as follows: a fault propagation simulation model and a reliability simulation model;
the fault propagation simulation model is modeling based on a virus propagation model, and single-time or cyclic simulation is carried out on the fault propagation process in the networked software system by setting the rule and the strength of fault propagation in the networked software system; in the simulation process, parameters to be set at least comprise fault propagation probability, node fault detection probability and fault node recovery probability; obtaining reliability related data of the networked software system after the simulation is finished, and simulating and evaluating the capability of the networked software system for bearing software faults and repairing;
the reliability simulation model performs single or cyclic simulation on the fault condition of the internal node of the software when the networked software system executes various different tasks according to the reliability definition, calculates the structural reliability and the task reliability of the software, and simulates and evaluates the reliability of the networked software system when executing different tasks in the use process.
2. The method for modeling and simulating reliability of a networked software system based on multiple agents according to claim 1, wherein the agents comprise: service agents and connector agents;
the service Agent receives the needed information from the external network environment, integrates or judges the received information according to the internal self state, and generates a description for modifying the current state of the service Agent; then matching information in the knowledge base to make a plan, and generating a series of actions acting on the external network environment according to the target;
the connector Agent receives needed information from the external network environment, integrates the information, performs action reaction according to an internal rule base, and feeds the processed information back to the external network environment.
3. The method for simulating reliability modeling of a multi-Agent-based networked software system according to claim 2, wherein the service Agent is modeled by using a cognitive Agent, and the formal description is represented by a six-tuple:
Ser_Agent::=<Seri,Ser_S,Ser_P,Ser_Per,Ser_Trans,Ser_KB,Ser_PS,Ser_GS,Ser_AS>
wherein, seri represents the serial number of the service Agent; ser_s represents a self state set; ser_p represents a perception set; ser_Per represents a sensing function of E-Ser_P, the environmental state is mapped into sensing input, and E represents an environmental state set; ser_Trans represents a decision function of Ser_P, ser_S- & gt Ser_S, and service change is realized according to the current state of the sensing input; ser_KB represents a knowledge base, which is a knowledge set of service agents, and at least contains knowledge of the running environment, action knowledge of the agents, and target knowledge of the service agents; ser_ps represents a planned set of services; ser_gs represents a service target set; ser_as represents the action set of the service.
4. The method for modeling and simulating reliability of a networked software system based on multiple agents according to claim 2, wherein the connector agents are modeled by using the structure of the reactive agents, and the formal description is represented by a quadruple:
Con_Agent::＝<Coni,Con_set,Con_KB,Con_AS>
wherein Coni represents the serial number of the connector Agent; con_set represents a state set of the connector Agent and at least comprises reliability and state information of the connector; con_KB represents the knowledge base of connector agents; con_AS represents the set of actions of the connector Agent.
5. The method for modeling and simulating reliability of a networked software system based on multiple agents according to claim 1, wherein the process for establishing the reliability model of the networked software system further comprises:
establishing an environment operation model of the networked software system: the environment information to be established comprises a networked software system use profile, a software target use probability, hardware operation information and actions of a user;
establishing a generating model of a software target: any one Agent participating in the networked software system acts, and different software targets are issued by the same Agent or issued by a plurality of agents together;
establishing a dynamic coordination relation model: acquiring related information between an Agent set of the established software target and other agents, and coordinating and interacting to jointly complete the software target;
establishing a reliability measurement model: measuring the specific number of software targets that the networked software system can complete a user within a period of time;
and integrating an environment operation model of the networked software system, a generation model of the software target, the software service call model, the dynamic coordination relation model and the reliability measurement model to serve as a reliability model of the networked software system.
6. The method for modeling and simulating reliability of a multi-Agent-based networked software system according to claim 1, wherein the simulation platform further comprises: the system comprises a parameter setting module, a service simulation and state management module, a visual display module and a simulation process display module;
setting the type of the simulation model and part of index parameters in the simulation process by the parameter setting module;
simulating a development environment through the service simulation and state management module, and starting and controlling corresponding simulation functions of the selected simulation model according to the set parameters;
checking the values and states of key indexes of the reliability model of the networked software system at each simulation time point through the visual display module;
and displaying each parameter set for the selected simulation model, state change chart information of the reliability model of the networked software system in the simulation process, simulation result information when the simulation is completed and the state of the reliability model of the networked software system through the simulation process display module.
7. The method for modeling and simulating the reliability of the networked software system based on the multiple agents according to claim 6, wherein after the single simulation of the fault propagation simulation model is finished, the node proportion of faults, the node proportion with fault resistance and the node proportion of faults which are still possible to occur in the networked software system, and the value and the change condition of each simulation time point are displayed through the visual display module;
after the cyclic simulation of the reliability simulation model is finished, the structural reliability and task reliability of the networked software system after the simulation is started are displayed through the visual display module, the development and change conditions of the structural reliability and the task reliability along with the simulation time are output, and the task sequence when the simulation is executed, the node sequence of each task in the execution sequence and the fault condition of the node sequence are output.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211469080.6A CN115858316B (en) | 2022-11-22 | 2022-11-22 | Multi-Agent-based networked software system reliability modeling simulation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211469080.6A CN115858316B (en) | 2022-11-22 | 2022-11-22 | Multi-Agent-based networked software system reliability modeling simulation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115858316A CN115858316A (en) | 2023-03-28 |
| CN115858316B true CN115858316B (en) | 2023-08-08 |
Family
ID=85665074
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211469080.6A Active CN115858316B (en) | 2022-11-22 | 2022-11-22 | Multi-Agent-based networked software system reliability modeling simulation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115858316B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117891566A (en) * | 2024-03-18 | 2024-04-16 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Reliability evaluation method, device, equipment, medium and product of intelligent software |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111898227A (en) * | 2020-07-24 | 2020-11-06 | 中国兵器工业信息中心 | Networked reliability simulation analysis method based on multiple intelligent agents |
| WO2021213293A1 (en) * | 2020-04-24 | 2021-10-28 | 西北工业大学 | Ubiquitous operating system oriented toward group intelligence perception |
| CN113746674A (en) * | 2021-08-27 | 2021-12-03 | 中车长春轨道客车股份有限公司 | TCMS network fault propagation analysis method based on SIR model |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10275545B2 (en) * | 2013-03-15 | 2019-04-30 | Ventana Systems, Inc. | Modeling and simulation |
| US20180227167A1 (en) * | 2017-02-08 | 2018-08-09 | Macau University Of Science And Technology | System, method, computer program and data signal for fault detection and recovery of a network |
-
2022
- 2022-11-22 CN CN202211469080.6A patent/CN115858316B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021213293A1 (en) * | 2020-04-24 | 2021-10-28 | 西北工业大学 | Ubiquitous operating system oriented toward group intelligence perception |
| CN111898227A (en) * | 2020-07-24 | 2020-11-06 | 中国兵器工业信息中心 | Networked reliability simulation analysis method based on multiple intelligent agents |
| CN113746674A (en) * | 2021-08-27 | 2021-12-03 | 中车长春轨道客车股份有限公司 | TCMS network fault propagation analysis method based on SIR model |
Non-Patent Citations (1)
| Title |
|---|
| 韩霜.列车网络控制系统软件可靠性分析及研究.制造业自动化.2017,(04),第139页至第142页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115858316A (en) | 2023-03-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kalaji et al. | Generating feasible transition paths for testing from an extended finite state machine (EFSM) | |
| Matinnejad et al. | Search-based automated testing of continuous controllers: Framework, tool support, and case studies | |
| Pandalai et al. | Template languages for fault monitoring of timed discrete event processes | |
| CN115858316B (en) | Multi-Agent-based networked software system reliability modeling simulation method | |
| Gregoriades et al. | Scenario-based assessment of nonfunctional requirements | |
| CN112487592B (en) | Bayesian network-based task reliability modeling analysis method | |
| Xia et al. | Property preservation of Petri synthesis net based representation for embedded systems | |
| Liu et al. | An approach towards the verification of expert systems using numerical Petri nets | |
| CN112084668B (en) | Simulation test method and device and electronic equipment | |
| CN107943008A (en) | Automated diagnostic test method based on VT systems | |
| CN112380084B (en) | Fault injection and simulation verification method | |
| CN111898227B (en) | Multi-agent-based networked reliability simulation analysis method | |
| Bai et al. | Semantic-based test oracles | |
| CN111709139A (en) | Complex electronic equipment failure risk transfer relation analysis method based on model system | |
| Last | Data mining for software testing | |
| Shuo et al. | Research on fault mechanism modeling method of virtual maintenance training equipment system | |
| Timo et al. | Fault detection in timed FSM with timeouts by SAT-solving | |
| Hall | A method and tools for large scale scenarios | |
| CN113067726B (en) | Network node failure determination method based on double logic layer agents | |
| Jones | Automated support for test-driven specification | |
| Hu et al. | DEVS systems-theory framework for reusable testing of I/O behaviors in service oriented architectures | |
| Rotondo | A Survey for Testing Self-organizing, Adaptive Systems in Industry 4.0 | |
| Matinnejad | Automated Testing of Simulink/Stateflow Models in the Automotive Domain | |
| Tadj et al. | Dynamic verification of an Object-Rule knowledge base using Colored Petri Nets | |
| Letichevsky et al. | Symbolic modelling in white-box model-based testing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |