CN114237207B - Multi-agent distributed fault diagnosis method under influence of communication noise - Google Patents
Multi-agent distributed fault diagnosis method under influence of communication noise Download PDFInfo
- Publication number
- CN114237207B CN114237207B CN202111570502.4A CN202111570502A CN114237207B CN 114237207 B CN114237207 B CN 114237207B CN 202111570502 A CN202111570502 A CN 202111570502A CN 114237207 B CN114237207 B CN 114237207B
- Authority
- CN
- China
- Prior art keywords
- agent
- model
- fault
- rule
- fault diagnosis
- 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
- 238000003745 diagnosis Methods 0.000 title claims abstract description 113
- 238000004891 communication Methods 0.000 title claims abstract description 53
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000005259 measurement Methods 0.000 claims abstract description 30
- 230000009471 action Effects 0.000 claims abstract description 11
- 239000003795 chemical substances by application Substances 0.000 claims description 178
- 239000013598 vector Substances 0.000 claims description 22
- 230000004913 activation Effects 0.000 claims description 20
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000010276 construction Methods 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000012544 monitoring process Methods 0.000 abstract description 18
- 230000008569 process Effects 0.000 description 8
- 239000011159 matrix material Substances 0.000 description 5
- 238000005457 optimization Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 206010033799 Paralysis Diseases 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
- G05B23/0262—Confirmation of fault detection, e.g. extra checks to confirm that a failure has indeed occurred
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24065—Real time diagnostics
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Computer And Data Communications (AREA)
- Test And Diagnosis Of Digital Computers (AREA)
Abstract
The invention discloses a multi-agent distributed fault diagnosis method under the influence of communication noise. The method comprises the following steps: establishing an agent related model; the intelligent agent correlation model comprises an intelligent agent system model, a measurement output information model and an information communication model; establishing a consistency protocol based on an agent related model to obtain the self output of the agent and the measurement output of the neighbor thereof under the action of the consistency protocol; determining fault characteristics based on the output of the agent itself and the measured output of its neighbors; determining fault diagnosis model parameters based on fault characteristics, and constructing a fault diagnosis model; and performing multi-agent distributed fault diagnosis through a fault diagnosis model. The invention can effectively reduce the communication cost between the intelligent agents by only using the monitoring information of the intelligent agents and the neighbors thereof, thereby realizing accurate and efficient diagnosis of system faults.
Description
Technical Field
The invention relates to the technical field of multi-agent control, in particular to a multi-agent distributed fault diagnosis method under the influence of communication noise.
Background
In the current technical development and social services, due to the improvement of the complexity of task execution, a single intelligent agent cannot complete more and more complex task demands, and a large system or a combat platform is formed by a plurality of intelligent agents to acquire comprehensive reconnaissance and measurement information, so that the whole situation is quickly and comprehensively examined, the complex tasks are cooperatively completed, and continuous work is realized more quickly and efficiently. For example, a military multi-unmanned aerial vehicle system can realize the perception of the whole battlefield situation through the information acquired by different unmanned aerial vehicles, and the battlefield contribution degree is higher than that of a single unmanned aerial vehicle. The distributed method aims at jointly completing a task through different agents through local network communication, and has the complex capability of far exceeding the capability of the agents after large-scale group formation, and has stronger flexibility and reliability, thereby being an effective strategy for solving the problem of cooperative control of multiple agents. However, in the distributed multi-agent, since there is no intermediate node to plan the behaviors of all agents, the system is affected by factors such as malicious attack, communication interference, internal component failure, etc., so that part of agents fail, and the whole system may be paralyzed when serious. Therefore, for the distributed multi-agent system, a set of safe and efficient fault diagnosis scheme is designed, so that the system fault can be accurately found and positioned, a foundation is provided for maintenance, and the system is an urgent work with wide application prospect. At present, the existing distributed fault diagnosis method is mainly developed aiming at a multi-agent system in an ideal communication environment, and the situation that communication noise exists when an agent frequently occurs in actual engineering application to acquire neighbor node information is not considered. When unpredictable communication noise and interference exist in the multi-agent system, the existing method cannot work normally.
Disclosure of Invention
The invention aims to provide a multi-agent distributed fault diagnosis method under the influence of communication noise, which is used for solving the problem of fault diagnosis of a multi-agent system under the influence of unpredictable communication noise and interference in engineering practice, thereby realizing accurate and efficient diagnosis of system faults.
In order to achieve the above object, the present invention provides the following solutions:
a multi-agent distributed fault diagnosis method under the influence of communication noise comprises the following steps:
establishing an agent related model; the intelligent agent related model comprises an intelligent agent system model, a measurement output information model and an information communication model;
establishing a consistency protocol based on the relevant model of the intelligent agent to obtain the self output of the intelligent agent and the measurement output of the neighbor of the intelligent agent under the action of the consistency protocol;
determining fault characteristics based on the output of the agent itself and the measured output of its neighbors;
determining fault diagnosis model parameters based on the fault characteristics, and constructing a fault diagnosis model;
and performing multi-agent distributed fault diagnosis through the fault diagnosis model.
Optionally, the establishing a consistency protocol based on the agent correlation model specifically includes:
and according to the information communication model, establishing a consistency protocol by using the output information of the intelligent agent in the intelligent agent system model and the measurement output information of the neighbor intelligent agent in the measurement output information model.
Optionally, determining the fault diagnosis model parameter based on the fault feature specifically includes:
calculating the matching degree of the fault characteristics relative to each rule in the BRB-R based on the fault characteristics;
and calculating the activation weight of each rule based on the matching degree.
Optionally, the performing multi-agent distributed fault diagnosis through the fault diagnosis model specifically includes:
calculating an output fault feature vector of the fault diagnosis model according to the activation weight of each rule;
and carrying out fault diagnosis on the intelligent agent according to the output fault characteristic vector.
Optionally, the matching degree of the fault characteristics with respect to each rule in the BRB-R is calculated as follows:
wherein,matching degree in the j rule after the conversion of the ith fault feature of the ith intelligent agent, R lk And R is l(k+1) Reference class in the k and k+1 rules for agent first failure feature, +.>For the first fault feature at time t, L i Is the regular number of the ith agent.
Optionally, the calculation formula of the activation weight of each rule is as follows:
wherein,activation weight, alpha, for the ith agent at the kth rule k 、α l The matching degree of the fault characteristics relative to the kth rule and the first rule respectively, theta k 、θ l The weights of the kth rule and the first rule, respectively.
Optionally, a calculation formula of the output fault feature vector of the fault diagnosis model is as follows:
wherein,output fault feature vector of the ith agent output for fault diagnosis model,/for the fault diagnosis model>Activation weight, beta, for the ith agent at the kth rule n,k 、β j,k The nth and the jth output fault feature vectors in the kth rule, L i And H is the type of fault diagnosis results, wherein H is the rule number of the ith agent.
The invention also provides a multi-agent distributed fault diagnosis system under the influence of communication noise, which comprises:
the intelligent agent related model building module is used for building an intelligent agent related model; the intelligent agent related model comprises an intelligent agent system model, a measurement output information model and an information communication model;
the consistency protocol establishing module is used for establishing a consistency protocol based on the relevant model of the intelligent agent to obtain the self output of the intelligent agent and the measurement output of the neighbor of the intelligent agent under the action of the consistency protocol;
the fault characteristic determining module is used for determining fault characteristics based on the output of the intelligent agent and the measurement output of the neighbor of the intelligent agent;
the fault diagnosis model construction module is used for determining fault diagnosis model parameters based on the fault characteristics and constructing a fault diagnosis model;
and the fault diagnosis module is used for performing multi-agent distributed fault diagnosis through the fault diagnosis model.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention can effectively reduce the communication cost between the intelligent agents by only using the monitoring information of the intelligent agents and the neighbors thereof, thereby realizing accurate and efficient diagnosis of system faults. The method can effectively improve the robustness of the distributed fault diagnosis method of the multi-intelligent system in the actual communication noise environment and ensure the safe and reliable operation of the method.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a multi-agent distributed fault diagnosis method under the influence of communication noise according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a multi-agent distributed fault diagnosis method under the influence of communication noise according to an embodiment of the present invention;
FIG. 3 is a topology of a group communication of unmanned aerial vehicles;
FIG. 4 is data of unmanned aerial vehicle group monitoring without failure;
fig. 5 is a case where node 1 has a failure;
FIG. 6 is a failure condition of node 2;
FIG. 7 is a failure condition of node 3;
fig. 8 is a result of fault diagnosis of the multi-unmanned aerial vehicle system.
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.
The invention aims to provide a multi-agent distributed fault diagnosis method under the influence of communication noise, which is used for solving the problem of fault diagnosis of a multi-agent system under the influence of unpredictable communication noise and interference in engineering practice, thereby realizing accurate and efficient diagnosis of system faults.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
As shown in fig. 1-2, the multi-agent distributed fault diagnosis method under the influence of communication noise provided by the invention comprises the following steps:
step 101: establishing an agent related model; the intelligent agent correlation model comprises an intelligent agent system model, a measurement output information model and an information communication model.
Step 102: and establishing a consistency protocol based on the relevant model of the intelligent agent to obtain the self output of the intelligent agent and the measurement output of the neighbor of the intelligent agent under the action of the consistency protocol.
Step 103: and determining fault characteristics based on the output of the intelligent agent and the measurement output of the neighbor of the intelligent agent.
Step 104: and determining fault diagnosis model parameters based on the fault characteristics, and constructing a fault diagnosis model.
Step 105: and performing multi-agent distributed fault diagnosis through the fault diagnosis model.
The step 101 specifically includes:
and (3) establishing an agent system model, and describing the motion state of the node (namely the agent) by adopting a general linear differential equation. And (5) establishing a measurement output information model of the intelligent agent to the neighbor of the intelligent agent under the communication noise. The information communication model between the nodes is described through the directed graph, namely, the nodes are in one-way communication, and the intelligent agents perform information interaction according to the directed graph.
Consider a multi-agent system in which there are N agents, where the system model of an agent is:
wherein x is i (t)，u i (t)，y i (t) is the system state, control input, and system output of the ith agent, respectively. w (w) i And (t) is the white noise interference of the unknown environment suffered by the ith agent. A, B, C, E w Is a system matrix of agent i.
In a multi-agent system, each agent can only measure the output signals of neighbor agents in a noisy environment. Thus the measured output of agent j received by agent i is
Wherein eta ji And (t) is communication white noise between the agent i and the agent j.
Communication topology between agents in a multi-agent systemRepresentation, where v= {1,..n } is a node in the figure, also represents each agent in the multi-agent system. E represents the road between nodes, if agent i can send information to agent j,/if it can send information to agent j>The set of adjacent nodes of node i consists of +.>And (3) representing. The adjacency matrix corresponding to the communication graph can be expressed as +.>Wherein when->When a is ij =1; otherwise, a ij ＝0。
All consider the case where there is an actuator failure in a multi-intelligent system. Assume a total of H failure modes, with m actuators per agent. Thus agent i is in failure mode H, h=1,..h, i=1..m actuators can be written as follows:
wherein the method comprises the steps ofFor unknown fault parameters, satisfy +.>When there is no fault ++>Definitions->Wherein xi i ＝diag{ξ ii ,...,ξ il ,...,ξ im }，/>Thus, inUnder the fault of the actuator, the system model (1) of the intelligent agent i can be rewritten as follows
The step 102 specifically includes:
in order to achieve failure diagnosis of a multi-agent system actuator failure, first, a consistency controller is designed to enable the system to achieve consistent output without failure. Then, a fault diagnosis model is designed based on a closed-loop system under the action of a designed controller, so that fault diagnosis can be effectively performed, and meanwhile, interference of measurement noise and environment on a detection result of the system is avoided.
For a multi-agent system under the influence of measurement noise, a distributed consistency controller based on relative measurement output is designed as follows:
where c > 0, and the matrix K is the controller gain. c and K are selected to satisfy the matrix A+cλ i BKC is Schur stable.
Under the action of the controller (5), the follower's closed-loop system model can be written as:
according to the closed loop system model (6) of the follower under the action of the controller, the information sent by the follower i received the neighbor j is obtained under the noise environment
For convenience of subsequent description, the invention receives |N received by the agent i i Information sent by i neighborsRedefined as->
Step 104 specifically includes: calculating the matching degree of the fault characteristics relative to each rule in the BRB-R based on the fault characteristics; and calculating the activation weight of each rule based on the matching degree.
Output information y of intelligent body under action of controller i (t) and its neighbor's measurement output informationAccording to the construction mode of the rule in BRB-R, the kth rule in the fault diagnosis model of the intelligent agent i is expressed as:
wherein,and y i (t) is |N of agent i at time t i Monitoring information of the +1 fault signature,is the weight of the index,/->Corresponding reference levels are input for the fault diagnosis model, aiming at converting the multiple information into a unified framework. F (t) represents the output failure diagnosis result, { D 1 ,…D H [ beta ] is the H fault states of the intelligent agent 1,k ,β 2,k …β H,k ]Outputting characteristic vector theta for the corresponding fault k Is a rule weight. k=1, 2,.. i 。
In the multi-agent distributed fault diagnosis method, the fault diagnosis process is divided into four steps, namely feature information conversion, activation weight calculation, rule fusion and parameter optimization.
Firstly, because the acquired multi-element monitoring information has different formats, the multi-element monitoring information cannot be directly used, and the multi-element monitoring information needs to be converted into a unified frame by the following formula:
wherein,matching degree in the j rule after the conversion of the ith fault feature of the ith intelligent agent, R lk And R is l(k+1) Reference class in the k and k+1 rules for agent first failure feature, +.>For the first fault feature at time t, L i Is the regular number of the ith agent.
After the matching degree of each index in each rule is obtained, the matching degree of all fault features in the kth rule can be obtained through the following formula:
wherein,representing the relative weight of the characteristics of the intelligent agent, r l The method for calculating the reliability of the communication information among the intelligent agents comprises the following steps:
wherein T is li Is the total number of communication information between the ith intelligent agent and the ith intelligent agent, v li For the number of unreliable information in the acquired information, i.e. when the communication data is reliable, v liτ =1; otherwise, v liτ =0. The reliable judgment of the communication information needs to be determined by the set fluctuation interval, and the fluctuation interval is expressed as And v li Respectively mean and variance of communication information in a period of time epsilon l And adjusting the coefficient for the width of the fluctuation interval. When the communication information is in the fluctuation interval, the interference degree of the communication information at the current moment is in the accepted range, and the information is reliable and available; otherwise, the communication information at the current moment is considered to be interfered to a large extent, and the information is unreliable.
Based on the calculation process, the characteristic information of the intelligent agent can be converted into the matching degree under the unified framework, and then the activation weight of the kth rule in the ith intelligent agent fault diagnosis model can be obtained through the following formula:
wherein,activation weight, alpha, for the ith agent at the kth rule k 、α l The matching degree of the fault characteristics relative to the kth rule and the first rule respectively, theta k 、θ l The weights of the kth rule and the first rule, respectively.
The activated rule may generate a feature vector of the system fault that represents the result of the rule diagnosis. The fault feature vectors output by all rules can be fused through a evidence reasoning (Evidential Reasoning, ER) algorithm to obtain the final output fault feature vector. The ER algorithm resolution format is as follows:
wherein,output fault feature vector of the ith agent output for fault diagnosis model,/for the fault diagnosis model>Activation weight, beta, for the ith agent at the kth rule n,k 、β j,k The nth and the jth output fault feature vectors in the kth rule, L i And H is the type of fault diagnosis results, wherein H is the rule number of the ith agent.
Step 105 specifically includes:
after the output fault feature vector of the model is obtained, faults of the intelligent agents in the multi-intelligent system are judged through the following formula.
Wherein the obtained multi-agent system fault is the eta.
The above process is a derivation process of the distributed fault diagnosis model of the multi-agent system. Since BRB-R is one of expert systems, the construction of the initial model needs to be given by means of expert knowledge. Under the influence of uncertainty of expert knowledge, the accuracy of the initial model for fault diagnosis of the intelligent agent cannot meet the requirement, and the initial model needs to be trained according to the acquired monitoring data.
In the aspect of model parameter updating, the BRB-R model has strict requirements on the physical meaning of parameters.
Therefore, the following constraints need to be obeyed in the model parameter optimization process:
0≤θ k ≤1 (17)
0≤δ i ≤1,i＝1,2,...,M (18)
0≤β n,k ≤1,n＝1,2,...,H,k＝1,2,...,L' (19)
the fault diagnosis model is optimized to have the maximum diagnosis accuracy, and the calculation method comprises the following steps:
wherein, gamma i And (3) the diagnosis rate of the constructed ith intelligent agent fault diagnosis model. R is R i For diagnosing the accurate sample number, T i The total number of samples obtained. When the diagnosis is accurate, R it =1; otherwise, R it ＝0。
The invention can effectively reduce the communication cost between the intelligent agents by only using the monitoring information of the intelligent agents and the neighbors thereof, thereby being widely applied to a large-scale system. In addition, the method can effectively improve the robustness of the distributed fault diagnosis method of the multi-intelligent system in the actual communication noise environment and ensure the safe and reliable operation of the method.
In order to verify the effectiveness of the invention, experimental verification is carried out in a multi-agent system consisting of 3 vertical take-off and landing airplanes, and the method mainly comprises the following steps:
step one: problem description and acquisition of monitoring information
The multi-unmanned aerial vehicle group is used as a typical multi-agent system, has huge application prospect and enabling potential in the key fields of national defense, military, aerospace, civil industry and the like, and can realize ground reconnaissance and striking in military, report performances in important places such as full-freight and the like. In the process of the cooperative flight of multiple unmanned aerial vehicles, the unmanned aerial vehicles are influenced by malicious attack, communication interference, internal component faults and other reasons, so that part of unmanned aerial vehicle intelligent bodies break down, and the whole system cannot complete the expected tasks. The multi-unmanned aerial vehicle group is composed of three vertical take-off and landing (VTOL) planes, the form of the longitudinal dynamics equation after linearization and discretization is shown as formula (1), wherein the state x is as follows i (t)＝[x i1 (t),x i2 (t),x i3 (t),x i4 (t)] T I=1, 2,3 represent the horizontal speed, vertical speed, pitch rate, pitch angle, respectively, of the unmanned aerial vehicle i. The system parameters are
The communication topology between the drones is shown in figure 3.
The controller parameter c=2 is selected,the monitored data obtained for the multiple unmanned aerial vehicle system in the absence of a fault is shown in fig. 4. The monitoring data when the nodes 1,2 and 3 are in failure are respectively obtained by simulating the failure of the unmanned aerial vehicle components and are respectively shown in figures 5,6 and 7.
Step two: construction and training of fault diagnosis model
In the process of carrying out fault diagnosis on the multi-unmanned aerial vehicle system, errors among monitoring data of three unmanned aerial vehicles are used as fault diagnosis characteristic input, and fault states of the three unmanned aerial vehicles are used as fault diagnosis model output results. In the BRB-R based multiple unmanned system fault diagnosis model, the reference levels of errors between three unmanned monitoring data are set as shown in tables 1 and 2, respectively, and are marked as low (L), medium (M), slightly High (SH) and high (H), respectively. The number of reference levels of the two fault diagnosis features is 4, the number of confidence rule bases in the initial fault diagnosis model constructed based on the reference levels is 16, and as shown in table 3, output results are divided into four types, namely normal, node 1 fault, node 2 fault and node 3 fault.
In the experimental process, 800 groups of monitoring data are obtained altogether, wherein 200 groups are monitoring data of a multi-unmanned aerial vehicle system in a normal state, 200 groups are monitoring data of a node 1 fault, 200 groups are monitoring data of a node 2 fault, and 200 groups are monitoring data of a node 3 fault. In the fault diagnosis model training stage, 400 groups are randomly selected from four states to serve as training data, and the rest 400 groups serve as test data. The optimization algorithm selected in the experiment is a covariance matrix self-adaptive optimization strategy (P-CMA-ES) taking a projection operator into consideration, and the optimization iteration number is set to 300. The optimized fault diagnosis model is shown in table 4, the fault diagnosis result of the multi-unmanned aerial vehicle system is shown in fig. 8, the optimized fault diagnosis rate is 97.00%, and the accuracy of the fault diagnosis model constructed by the invention is proved.
Table 1 reference level and reference value of error between node 1 and node 2
Table 2 reference levels and reference values between node 2 and node 3
TABLE 3 initial fault diagnostic model
TABLE 4 optimized fault diagnosis model
The invention also provides a multi-agent distributed fault diagnosis system under the influence of communication noise, which comprises:
the intelligent agent related model building module is used for building an intelligent agent related model; the intelligent agent related model comprises an intelligent agent system model, a measurement output information model and an information communication model;
the consistency protocol establishing module is used for establishing a consistency protocol based on the relevant model of the intelligent agent to obtain the self output of the intelligent agent and the measurement output of the neighbor of the intelligent agent under the action of the consistency protocol;
the fault characteristic determining module is used for determining fault characteristics based on the output of the intelligent agent and the measurement output of the neighbor of the intelligent agent;
the fault diagnosis model construction module is used for determining fault diagnosis model parameters based on the fault characteristics and constructing a fault diagnosis model;
and the fault diagnosis module is used for performing multi-agent distributed fault diagnosis through the fault diagnosis model.
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 system 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 principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to assist in understanding the methods of the present invention and the core ideas thereof; also, it is within the scope of the present invention to be modified by those of ordinary skill in the art in light of the present teachings. In view of the foregoing, this description should not be construed as limiting the invention.
Claims (2)
1. The utility model provides a multi-agent distributed fault diagnosis method under communication noise influence, which is characterized in that the method comprises the following steps:
establishing an agent related model; the intelligent agent related model comprises an intelligent agent system model, a measurement output information model and an information communication model;
establishing a consistency protocol based on the relevant model of the intelligent agent to obtain the self output of the intelligent agent and the measurement output of the neighbor of the intelligent agent under the action of the consistency protocol; the establishing a consistency protocol based on the agent correlation model specifically comprises the following steps: according to the information communication model, establishing a consistency protocol by using the self output information of the intelligent agent in the intelligent agent system model and the measurement output information of the neighbor intelligent agent in the measurement output information model;
determining fault characteristics based on the output of the agent itself and the measured output of its neighbors;
determining fault diagnosis model parameters based on the fault characteristics, and constructing a fault diagnosis model; the method specifically comprises the following steps: calculating the matching degree of the fault characteristics relative to each rule in the BRB-R based on the fault characteristics; calculating the activation weight of each rule based on the matching degree;
performing multi-agent distributed fault diagnosis through the fault diagnosis model; the method specifically comprises the following steps: calculating an output fault feature vector of the fault diagnosis model according to the activation weight of each rule; performing fault diagnosis on the intelligent agent according to the output fault characteristic vector;
the calculation formula of the matching degree of the fault characteristics relative to each rule in the BRB-R is as follows:
wherein,matching degree in the j rule after the conversion of the ith fault feature of the ith intelligent agent, R lk And R is l(k+1) Reference class in the k and k+1 rules for agent first failure feature, +.>For the first fault feature at time t, L i A rule number for the ith agent;
the calculation formula of the activation weight of each rule is as follows:
wherein,activation weight, alpha, for the ith agent at the kth rule k 、α l The matching degree of the fault characteristics relative to the kth rule and the first rule respectively, theta k 、θ l The weights of the kth rule and the first rule are respectively;
the calculation formula of the output fault characteristic vector of the fault diagnosis model is as follows:
wherein,output fault feature vector of the ith agent output for fault diagnosis model,/for the fault diagnosis model>Activation weight, beta, for the ith agent at the kth rule n,k 、β j,k The nth and the jth output fault feature vectors in the kth rule, L i And H is the type of fault diagnosis results, wherein H is the rule number of the ith agent.
2. A multi-agent distributed fault diagnosis system under the influence of communication noise, comprising:
the intelligent agent related model building module is used for building an intelligent agent related model; the intelligent agent related model comprises an intelligent agent system model, a measurement output information model and an information communication model;
the consistency protocol establishing module is used for establishing a consistency protocol based on the relevant model of the intelligent agent to obtain the self output of the intelligent agent and the measurement output of the neighbor of the intelligent agent under the action of the consistency protocol; the establishing a consistency protocol based on the agent correlation model specifically comprises the following steps: according to the information communication model, establishing a consistency protocol by using the self output information of the intelligent agent in the intelligent agent system model and the measurement output information of the neighbor intelligent agent in the measurement output information model;
the fault characteristic determining module is used for determining fault characteristics based on the output of the intelligent agent and the measurement output of the neighbor of the intelligent agent;
the fault diagnosis model construction module is used for determining fault diagnosis model parameters based on the fault characteristics and constructing a fault diagnosis model; the method specifically comprises the following steps: calculating the matching degree of the fault characteristics relative to each rule in the BRB-R based on the fault characteristics; calculating the activation weight of each rule based on the matching degree;
the fault diagnosis module is used for performing multi-agent distributed fault diagnosis through the fault diagnosis model; the method specifically comprises the following steps: calculating an output fault feature vector of the fault diagnosis model according to the activation weight of each rule; performing fault diagnosis on the intelligent agent according to the output fault characteristic vector;
the calculation formula of the matching degree of the fault characteristics relative to each rule in the BRB-R is as follows:
wherein,matching degree in the j rule after the conversion of the ith fault feature of the ith intelligent agent, R lk And R is l(k+1) Reference class in the k and k+1 rules for agent first failure feature, +.>For the first fault feature at time t, L i A rule number for the ith agent;
the calculation formula of the activation weight of each rule is as follows:
wherein,activation weight, alpha, for the ith agent at the kth rule k 、α l The matching degree of the fault characteristics relative to the kth rule and the first rule respectively, theta k 、θ l The weights of the kth rule and the first rule are respectively;
the calculation formula of the output fault characteristic vector of the fault diagnosis model is as follows:
wherein,output fault feature vector of the ith agent output for fault diagnosis model,/for the fault diagnosis model>Activation weight, beta, for the ith agent at the kth rule n,k 、β j,k The nth and the jth output fault feature vectors in the kth rule, L i And H is the type of fault diagnosis results, wherein H is the rule number of the ith agent.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111570502.4A CN114237207B (en) | 2021-12-21 | 2021-12-21 | Multi-agent distributed fault diagnosis method under influence of communication noise |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111570502.4A CN114237207B (en) | 2021-12-21 | 2021-12-21 | Multi-agent distributed fault diagnosis method under influence of communication noise |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114237207A CN114237207A (en) | 2022-03-25 |
| CN114237207B true CN114237207B (en) | 2024-02-02 |
Family
ID=80760152
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111570502.4A Active CN114237207B (en) | 2021-12-21 | 2021-12-21 | Multi-agent distributed fault diagnosis method under influence of communication noise |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114237207B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105763396A (en) * | 2016-04-12 | 2016-07-13 | 北京理工大学 | Distributed multi-intelligent-agent real time fault detection method based on neighbor associated condition |
| CN108092833A (en) * | 2018-02-28 | 2018-05-29 | 哈尔滨理工大学 | A kind of multiple agent uniformity method for diagnosing faults |
| CN111290277A (en) * | 2020-02-26 | 2020-06-16 | 鹏城实验室 | Distributed multi-agent cooperative fault detection method, storage medium and equipment |
| CN113325708A (en) * | 2021-05-13 | 2021-08-31 | 南京航空航天大学 | Fault estimation method of multi-unmanned aerial vehicle system based on heterogeneous multi-agent |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7260501B2 (en) * | 2004-04-21 | 2007-08-21 | University Of Connecticut | Intelligent model-based diagnostics for system monitoring, diagnosis and maintenance |
| US8543280B2 (en) * | 2011-04-29 | 2013-09-24 | Toyota Motor Engineering & Manufacturing North America, Inc. | Collaborative multi-agent vehicle fault diagnostic system and associated methodology |
-
2021
- 2021-12-21 CN CN202111570502.4A patent/CN114237207B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105763396A (en) * | 2016-04-12 | 2016-07-13 | 北京理工大学 | Distributed multi-intelligent-agent real time fault detection method based on neighbor associated condition |
| CN108092833A (en) * | 2018-02-28 | 2018-05-29 | 哈尔滨理工大学 | A kind of multiple agent uniformity method for diagnosing faults |
| CN111290277A (en) * | 2020-02-26 | 2020-06-16 | 鹏城实验室 | Distributed multi-agent cooperative fault detection method, storage medium and equipment |
| CN113325708A (en) * | 2021-05-13 | 2021-08-31 | 南京航空航天大学 | Fault estimation method of multi-unmanned aerial vehicle system based on heterogeneous multi-agent |
Non-Patent Citations (2)
| Title |
|---|
| 可重构制造系统故障诊断多Agent自学习模型;贺鹏;王庆林;;计算机工程与设计(08);全文 * |
| 智能故障诊断技术的现状及展望;张卫峰;惠俊军;;工业仪表与自动化装置(05);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114237207A (en) | 2022-03-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108594788B (en) | Airplane actuator fault detection and diagnosis method based on deep random forest algorithm | |
| CN110058519B (en) | Active formation fault-tolerant control method based on rapid self-adaptive technology | |
| CN110348752B (en) | Large industrial system structure safety assessment method considering environmental interference | |
| CN114578792B (en) | Multi-agent fault diagnosis method and system | |
| CN114564000B (en) | Active fault tolerance method and system based on intelligent aircraft actuator fault diagnosis | |
| CN111415010B (en) | Bayesian neural network-based wind turbine generator parameter identification method | |
| Norouzi et al. | Real time estimation of impaired aircraft flight envelope using feedforward neural networks | |
| CN112947385B (en) | Aircraft fault diagnosis method and system based on improved Transformer model | |
| CN108427400A (en) | A kind of aircraft airspeed pipe method for diagnosing faults based on neural network Analysis design | |
| Aldebrez et al. | Dynamic modelling of a twin rotor system in hovering position | |
| CN117390498B (en) | Flight capability assessment method of fixed wing cluster unmanned aerial vehicle based on Transformer model | |
| Han et al. | Quadratic-Kalman-filter-based sensor fault detection approach for unmanned aerial vehicles | |
| CN114237207B (en) | Multi-agent distributed fault diagnosis method under influence of communication noise | |
| CN106168760B (en) | Uncertain time-delayed systems determination of stability method based on convex polyhedron fault model | |
| Fan et al. | UAV engine fault and diagnosis with parameter models based on telemetry data | |
| CN116540665A (en) | Multi-unmanned aerial vehicle system safety control method based on unknown input observer | |
| CN102788955A (en) | Remaining lifetime prediction method of ESN (echo state network) turbine generator classification submodel based on Kalman filtering | |
| CN115993077A (en) | Optimal decision method and optimal decision system for inertial navigation system under complex road condition transportation condition | |
| CN115660039A (en) | Intelligent diagnosis method, system, equipment and medium for fault of shipboard aircraft | |
| CN110275514B (en) | Asymptotic fault diagnosis method for formation flight control system with time-varying sensor fault | |
| Omkar et al. | Identification of helicopter dynamics based on flight data using nature inspired techniques | |
| Emami et al. | Identification of nonlinear time‐varying systems using wavelet neural networks | |
| CN114611377B (en) | Method for reconstructing reentry track of gliding missile in real time based on kernel extreme learning machine | |
| CN111680356B (en) | Unmanned aerial vehicle actuator real-time safety performance evaluation method based on two-stage observer | |
| Liu et al. | State Estimation of UAV under Network Attack |
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 |