Preset time fault fusion estimation method for intelligent networked automobile queue

Abstract

The invention relates to preset time fault estimation for an intelligent connected automobile queue. The invention discloses a preset time fault estimation method for an intelligent networked automobile queue. The method comprises an intelligent networked automobile queue dynamics model under multi-source faults such as an actuator fault, a vehicle-road communication fault, an inter-vehicle communication fault and external interference, a distributed preset time fault fusion estimation mechanism and a multi-target optimization mechanism based on dynamic planning. For a multi-source fault, a fault fusion estimation mechanism is designed in combination with a distributed architecture and a preset time theory. Based on dynamic planning and a strategy game theory, a multi-objective optimization mechanism is provided, an information flow topology reconstruction and queue spacing reconstruction strategy is constructed, and the driving safety and coordination of a motorcade are optimized. And finally, the optimal solution is fed back to a fault estimation mechanism, and online optimization adjustment is carried out. According to the method, the fault estimation and optimization problem of the intelligent network connection automobile queue under the complex fault is solved, and the safety and reliability of the system are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of fault estimation for intelligent connected vehicle platoons, and more specifically, to a preset time fault fusion estimation method for intelligent connected vehicle platoons. Background Technology

[0002] Driven by technological innovations in intelligent transportation and intelligent connected vehicles, intelligent connected vehicle platoons offer significant advantages in improving road traffic efficiency and achieving intelligent traffic management due to their collaborative driving capabilities between vehicles and between vehicles and infrastructure. While intelligent connected vehicle platoons rely on in-vehicle self-organizing network communication technology, achieving information exchange and collaborative control through inter-vehicle and vehicle-to-infrastructure (V2I) communication, in actual operation, the occurrence of multiple sources of faults, such as actuator failures, inter-vehicle communication failures, V2I communication failures, and external interference, can lead to a decline in the safety and stability of platoon control, seriously threatening driving safety.

[0003] To address the challenges of multi-source fault estimation and safety control in intelligent connected vehicle platoons, numerous research methods have been proposed. The literature [Z. Chen, J. Zhang, and L. Wang, "Prescribed-Time Fault Estimation for Nonlinear Systems with Applications to Vehicle Platoons," IEEE Transactions on Intelligent Transportation Systems, 2022, vol. 23, no. 8, pp. 6543-6554] proposes a prescribed-time fault estimation method for intelligent connected vehicle platoon systems, providing a technical path for rapid fault estimation. The literature [X. Li, Y. Wang, and H. Zhang, "Distributed Fault Estimation for Multi-Agent Systems with Uncertainties," IEEE Transactions on Neural Networks and Learning Systems, 2021, vol. 32, no. 5, pp. 2010-2023] focuses on distributed fault estimation for multi-agent systems, and its distributed architecture provides a reference for fault estimation in intelligent connected vehicle platoons.

[0004] The aforementioned research has laid the foundation for fault estimation in intelligent systems, but there are still shortcomings in the distributed pre-set time fault fusion estimation of intelligent connected vehicle platoons under multi-source faults: first, the coupling characteristics of multi-source faults are not sufficiently characterized, making it difficult to accurately describe the propagation impact of faults in the platoon; second, there is a lack of multi-objective optimization design combined with dynamic programming, resulting in deficiencies in dynamic adjustment of information flow topology and real-time estimation optimization and control. As a typical scenario of multi-agent collaboration, intelligent connected vehicle platoons are widely used in urban transportation, logistics, and other fields, and multi-source faults directly restrict their safe and stable operation. Therefore, a distributed pre-set time fault fusion estimation method for intelligent connected vehicle platoons is a key technical problem that urgently needs to be solved to ensure the reliable operation of intelligent connected vehicle platoons. Summary of the Invention

[0005] The technical problem to be solved by the invention is to address the issue of untimely and inaccurate fault estimation that may occur in intelligent connected vehicle queues under multi-source faults, and to propose a preset time fault fusion estimation method for intelligent connected vehicle queues.

[0006] To achieve the above objectives, according to one aspect of a specific embodiment of the present invention, a system is provided that includes intelligent connected vehicle queue dynamics modeling under multi-source faults, a distributed preset time fault fusion estimation mechanism, and a multi-objective optimization mechanism based on dynamic programming.

[0007] This paper constructs a dynamic model of intelligent connected vehicle (ICV) queuing under multi-source faults, considering actuator faults, vehicle-to-vehicle communication faults, vehicle-to-infrastructure communication faults, and external interference characteristics. Combining a distributed architecture and preset time design, a distributed preset time fault fusion estimation mechanism is used to achieve online fusion estimation of vehicle state and multi-source faults within a preset time. Based on dynamic programming and strategy game theory, and incorporating driving state and multi-source fault fusion estimation information, a multi-objective optimization mechanism based on dynamic programming is designed. A dynamic information flow topology reconstruction evolution strategy and a queue spacing strategy reconstruction mechanism based on strategy game theory are also constructed to achieve online optimization and adjustment, ensuring the accuracy of fault estimation and the safety and stability of system control.

[0008] Specifically, the intelligent connected vehicle platoon dynamics model under multi-source faults consists of j intelligent connected vehicles that achieve inter-vehicle communication and vehicle-to-infrastructure communication based on an onboard ad hoc network, denoted as q = {q1, q2, ..., q j}, j=1,2,…N. Consider the existence of multi-source faults during the operation of a convoy of intelligent connected vehicles, including actuator faults F. act Workshop communication failure F v2v Vehicle-to-infrastructure communication fault F v2i And external interference D, multi-source fault data are collected through online detection and identification technology to form a multi-source fault model: F={F act,F v2v ,F v2i ,D}。 Integrating the longitudinal following model M l Lane Changing Driving Model M c Yaw angle following model M y A dynamic model M of intelligent connected vehicle queues under multi-source faults is formed. dyn ={M l M c M y}

[0009] Specifically, a distributed, pre-set time-based fault fusion estimation mechanism is constructed. First, based on the intelligent connected vehicle queue dynamics model M under multi-source faults... dyn Parameter information, actual driving status information A = {p i ,v i ,a i ,…}(p i The position, v, represents the actual driving state. i The speed represented by a indicates the actual speed during driving. i Using the acceleration (representing the actual driving state) and multi-source fault information F, combined with a distributed architecture and preset time theory, a distributed preset time fault fusion estimator ε(M) is constructed. dyn ,A,F,e i ,m i ,T), where e i To estimate the error, m i Given the intermediate fault quantity and T as the preset time, output the state and fault fusion estimation information of the intelligent connected vehicle queue. Provides information on vehicle driving status estimation. This provides information for multi-source fault estimation.

[0010] Specifically, a multi-objective optimization mechanism based on dynamic programming is designed. First, for state and fault estimation information divided according to information dimensions... By utilizing the evolutionary laws of information flow topology dynamics and strategy dynamics, a set of information flow topology adjustment strategies S = {S1,…,S} is formulated. i}, constructing a policy scheduling set {(C) using information flow topology stability mapping. V C N C c This allows us to obtain the execution solution for information flow topology reconstruction. Where R Vi (·), R Ni (·), R ci (·) represent the information flow topology strategy distribution methods, respectively. and This represents the state and fault estimation information. Then, it is combined with the information flow topology optimization benefit set G = {g1, g2, ..., g...} m The set of fault threat strategies R = {r1, r2, ..., r} n} and the set of information flow topology adjustment strategies S = {S1, ..., S} i}, construct the payoff function U(s) of the strategy game model k ,g p ,r q )=αg p -βr q -γc(S k Solve the game equilibrium. Obtain the optimal topology reconstruction strategy S * This drives the topology reconstruction of the information flow. Where g... p For the performance gains of the topology, r q The threat level of the fault to the topology is represented by α, β, and γ, which are weighting coefficients, and c(S) is the weighting coefficient. k Strategy S k The execution cost. And further, based on the information flow, the topology information is reconstructed. Considering error propagation stability theory and traffic flow stability theory, we define vehicle q i With the car in front q i-1 longitudinal spacing d x,i (t)=d0+v i (t)τ+k v (v i (t)-v i-1 (t) and lateral spacing d y,i (t)=w+k u (u i (t)-u i-1 (t)), where d0 is the stationary safety distance, v i (t) represents vehicle q i The actual longitudinal velocity, τ is the reaction time, and k v Let w be the longitudinal coupling coefficient, w be the vehicle width, and k be the [missing value]. u For the lateral coupling coefficient, u i (t) represents the actual lateral velocity of the vehicle, and a two-dimensional spacing d is constructed. i (t)=[d x,i (t),d y,i (t)] T And establish dynamic coupling adjustment equations Where K x K y These are the longitudinal and lateral feedback gain vectors, respectively. , are the longitudinal and lateral position estimates of the vehicle, respectively, and H is the topology-state coupling matrix. Vehicle driving state estimation information is obtained. Finally, based on the reconstructed information flow topology information, two-dimensional spacing strategy, and state and fault estimation information, the optimal solution K is obtained by solving the multi-objective performance function optimization problem of the fault fusion estimator combination using dynamic programming. * Feedback is sent to the fusion estimator for online optimization and adjustment of the gain value. Where K * To obtain the optimal gain through dynamic programming, The current gain value of the fusion estimator, α is the gain adjustment step size, This provides an online update value for the fusion estimator gain, improving the accuracy of the fusion estimator in estimating states and faults. Ultimately, the optimized estimator outputs more reliable fusion estimation information for the queue controller U = {u1, u2, ..., u...}. j It provides accurate input to enable intelligent connected vehicle platoons to drive safely and stably under multi-source faults, while ensuring error propagation stability and internal stability, and enhancing the system's autonomous coordination capabilities.

[0011] The beneficial effects of this invention are that it can effectively solve the problems of untimely and inaccurate fault fusion estimation of intelligent connected vehicle queues under multi-source faults, thereby improving the stability and security of intelligent connected vehicle queues under multi-source faults, enhancing the ability of intelligent connected vehicle queues to resist multi-source faults, and increasing the overall performance of the system.

[0012] The invention will be further explained below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of a specific embodiment of the present invention;

[0014] Figure 2 This is a schematic diagram of the information flow topology reconstruction principle under strategic game theory. Detailed Implementation

[0015] The following description, using a multi-source fault intelligent connected vehicle queue as an example, along with accompanying drawings, provides a detailed, clear, and complete description of the technical solution of this invention, so as to facilitate a better understanding of this invention by those skilled in the art.

[0016] Example

[0017] Figure 1 This is a schematic diagram illustrating the specific implementation of the preset time fault fusion estimation method for the intelligent connected vehicle queue in this example. Figure 1 As shown, the specific steps for the preset time fault fusion estimation of the intelligent connected vehicle queue in this example include: intelligent connected vehicle queue dynamic modeling under multi-source faults, a distributed preset time fault fusion estimation mechanism, and a multi-objective optimization mechanism based on dynamic programming.

[0018] This example demonstrates the dynamics of a queuing of intelligent connected vehicles under multi-source faults. The queuing consists of j intelligent connected vehicles, denoted as q = {q1, q2, ..., q...}, which implement inter-vehicle and vehicle-to-infrastructure (V2I) communication based on an onboard ad hoc network. j}, j=1,2,…N. Consider the existence of multi-source faults during the operation of a convoy of intelligent connected vehicles, including actuator faults F. act Workshop communication failure F v2v Vehicle-to-infrastructure communication fault F v2i And external interference D, multi-source fault data are collected through online detection and identification technology to form a multi-source fault model: F={F act ,F v2v ,F v2i ,D}。 Integrating the longitudinal following model M l Lane Changing Driving Model M c Yaw angle following model M y A dynamic model M of intelligent connected vehicle queues under multi-source faults is formed. dyn ={M l M c M y}

[0019] This example demonstrates a distributed, pre-defined time-based fault fusion estimation mechanism based on the intelligent connected vehicle queue dynamics model M under multi-source faults. dyn Parameter information, actual driving status information A = {p i ,v i ,a i ,…}(p i The position, v, represents the actual driving state. i The speed represented by a indicates the actual speed during driving. i Using the acceleration (representing the actual driving state) and multi-source fault information F, combined with a distributed architecture and preset time theory, a distributed preset time fault fusion estimator ε(M) is constructed. dyn ,A,F,e i ,m i ,T), where e i To estimate the error, m i Given the intermediate fault quantity and T as the preset time, output the state and fault fusion estimation information of the intelligent connected vehicle queue. Provides information on vehicle driving status estimation. This provides information for multi-source fault estimation.

[0020] First, the state and fault fusion estimation information output by the distributed preset time fault fusion estimator are divided according to the dimension of information usage. Divide into three categories of information Basic driving state estimation information for all vehicles. This is data for estimating the cooperative state between vehicles. Fault estimation data for all vehicles. Segmented by information usage dimension, three categories of status and fault estimation information are provided. They provide clearly categorized inputs for subsequent information flow topology adjustment strategy generation, strategy scheduling set construction, and multi-objective performance optimization, ultimately supporting the stable operation of intelligent connected vehicle queues under multi-source faults.

[0021] This example uses a multi-objective optimization mechanism based on dynamic programming to address the state and fault estimation information divided by information dimensions. By utilizing the evolutionary laws of information flow topology dynamics and strategy dynamics, a set of information flow topology adjustment strategies S = {S1,…,S} is formulated. i}, constructing a policy scheduling set {(C) using information flow topology stability mapping. V C N C c This allows us to obtain the execution solution for information flow topology reconstruction. Where R Vi (·), R Ni (·), R ci (·) represent the information flow topology strategy distribution methods, respectively. and Representing state and fault estimation information; the principle diagram of information flow topology reconstruction under strategic game is as follows: Figure 2 As shown. Then, combined with the information flow topology optimization benefit set G={g1,g2,…,g m The set of fault threat strategies R = {r1, r2, ..., r} n} and the set of information flow topology adjustment strategies S = {S1, ..., S} i}, construct the payoff function U(s) of the strategy game model k ,g p ,r q )=αg p -βr q -γc(S k Solve the game equilibrium. Obtain the optimal topology reconstruction strategy S * This drives the topology reconstruction of the information flow. Where g... p For the performance gains of the topology, r q The threat level of the fault to the topology is represented by α, β, and γ, which are weighting coefficients, and c(S) is the weighting coefficient. k Strategy S k The execution cost. And further, based on the information flow, the topology information is reconstructed. Considering error propagation stability theory and traffic flow stability theory, we define vehicle q i With the car in front qi-1 longitudinal spacing d x,i (t)=d0+v i (t)τ+k v (v i (t)-v i-1 (t) and lateral spacing d y,i (t)=w+k u (u i (t)-u i-1 (t)), where d0 is the stationary safety distance, v i (t) represents vehicle q i The actual longitudinal velocity, τ is the reaction time, and k v Let w be the longitudinal coupling coefficient, w be the vehicle width, and k be the [missing value]. u For the lateral coupling coefficient, u i (t) represents the actual lateral velocity of the vehicle, and a two-dimensional spacing d is constructed. i (t)=[d x,i (t),d y,i (t)] T And establish dynamic coupling adjustment equations Where K x K y These are the longitudinal and lateral feedback gain vectors, respectively. , are the longitudinal and lateral position estimates of the vehicle, respectively, and H is the topology-state coupling matrix. Vehicle driving state estimation information is obtained. Finally, based on the reconstructed information flow topology information, two-dimensional spacing strategy, and state and fault estimation information, the optimal solution K is obtained by solving the multi-objective performance function optimization problem of the fault fusion estimator combination using dynamic programming. * Feedback is sent to the fusion estimator for online optimization and adjustment of the gain value. Where K * To obtain the optimal gain through dynamic programming, The current gain value of the fusion estimator, α is the gain adjustment step size, This provides an online update value for the fusion estimator gain, improving the accuracy of the fusion estimator in estimating states and faults. Ultimately, the optimized estimator outputs more reliable fusion estimation information for the queue controller U = {u1, u2, ..., u...}. j It provides accurate input to support the stable operation of the entire intelligent connected vehicle fleet under multi-source faults, while ensuring the stability of error propagation and internal stability, and enhancing the system's autonomous coordination capabilities.

Claims

1. A preset time fault fusion estimation method for intelligent connected vehicle platoons, characterized in that, The intelligent connected vehicle queue dynamics model, distributed preset time fault fusion estimation mechanism, and dynamic programming-based multi-objective optimization mechanism under multiple source faults such as actuator failure, vehicle-to-infrastructure communication failure, vehicle-to-vehicle communication failure, and external interference include the following steps: Step 1: Combine online detection and identification technologies to collect multi-source fault data such as actuator faults, vehicle-to-infrastructure communication faults, inter-vehicle communication faults, and external interference, and construct a dynamic model of an intelligent connected vehicle queue consisting of j intelligent connected vehicles; Step 2: Combining the parameters of the intelligent connected vehicle queue dynamic model and driving status information, as well as the characteristics of multi-source fault data, a distributed architecture and preset time theory are introduced to design a distributed preset time fault fusion estimation mechanism to realize online fusion estimation of the driving status and multi-source faults of the intelligent connected vehicle queue within a preset time. Step 3: Combining driving status and multi-source fault fusion estimation information, based on dynamic programming and strategy game theory, design a multi-objective optimization mechanism based on dynamic programming, construct a dynamic information flow topology reconstruction evolution strategy and a queue spacing strategy reconstruction mechanism based on strategy game theory; feed the obtained optimal solution back to the distributed preset time fault estimation mechanism to achieve online optimization and adjustment.

2. The preset time fault fusion estimation method for an intelligent connected vehicle queue according to claim 1, characterized in that, The aforementioned design of the intelligent connected vehicle queue dynamics model under multi-source faults: It consists of j intelligent connected vehicles, denoted as q = {q1, q2, ..., q...}, which are vehicles that achieve inter-vehicle and vehicle-to-infrastructure communication based on an onboard ad hoc network. j }, j=1,2,…N;Consider that during the operation of a convoy of intelligent connected vehicles, there are multiple sources of faults, including actuator faults F act Workshop communication failure F v2v Vehicle-to-infrastructure communication fault F v2i And external interference D, multi-source fault data are collected through online detection and identification technology to form a multi-source fault model: F={F act ,F v2v ,F v2i ,D};Integrated longitudinal following model M l Lane Changing Driving Model M c Yaw angle following model M y A dynamic model M of intelligent connected vehicle queues under multi-source faults is formed. dyn ={M l M c M y } 3. The preset time fault fusion estimation method for an intelligent connected vehicle queue according to claim 1, characterized in that, The distributed preset time fault fusion estimation mechanism is designed as follows: First, based on the intelligent connected vehicle queue dynamics model M under multi-source faults. dyn Parameter information, actual driving status information A = {p i ,v i ,a i ,…}, where p i The position indicating the actual driving state, v i Speed ​​representing actual driving conditions, a i Representing the acceleration under actual driving conditions, and multi-source fault information F, a distributed preset-time fault fusion estimator ε(M) is constructed by combining a distributed architecture and preset-time theory. dyn ,A,F,e i ,m i ,T), where e i To estimate the error, m i Given the intermediate fault quantity and T as the preset time, output the state and fault fusion estimation information of the intelligent connected vehicle queue. Provides information on vehicle driving status estimation. This provides information for multi-source fault estimation.

4. The preset time fault fusion estimation method for an intelligent connected vehicle queue according to claim 1, characterized in that, The proposed multi-objective optimization mechanism based on dynamic programming involves: using reconstructed information flow topology information, two-dimensional spacing strategy, and state and fault estimation information, dynamically programming is used to solve the multi-objective performance function optimization problem of fault fusion estimator combination, and the obtained optimal solution K is then used. * Feedback is sent to the fusion estimator for online optimization and adjustment of the gain value. Where K * To obtain the optimal gain through dynamic programming, The current gain value of the fusion estimator, α is the gain adjustment step size, This serves as an online update value for the fusion estimator gain, improving the accuracy of the fusion estimator in estimating states and faults. Ultimately, the optimized estimator outputs more reliable fusion estimation information for the queue controller U = {u1, u2, ..., u...}. j It provides accurate input to enable intelligent connected vehicle platoons to drive safely and stably under multi-source faults, while ensuring error propagation stability and internal stability, and enhancing the system's autonomous coordination capabilities.

5. The dynamic information flow topology reconstruction and evolution strategy mechanism for constructing strategic games according to claim 4, characterized in that, First, regarding the status and fault estimation information categorized by information dimension... By utilizing the evolutionary laws of information flow topology dynamics and strategy dynamics, a set of information flow topology adjustment strategies S = {S1,…,S} is formulated. i }, constructing a policy scheduling set {(C) using information flow topology stability mapping. V C N C c This allows us to obtain the execution solution for information flow topology reconstruction. Where R Vi (·), R Ni (·), R ci (·) represent the information flow topology strategy distribution methods, respectively. and This represents the state and fault estimation information; then, it is combined with the information flow topology optimization benefit set G = {g1, g2, ..., g...} m The set of fault threat strategies R = {r1, r2, ..., r} n } and the set of information flow topology adjustment strategies S = {S1, ..., S} i }, construct the payoff function U(s) of the strategy game model k ,g p ,r q )=αg p -βr q -γc(S k Solve the game equilibrium. Obtain the optimal topology reconstruction strategy S * This drives the topology reconstruction of the information flow. Where g... p For the performance gains of the topology, r q The threat level of the fault to the topology is represented by α, β, and γ, which are weighting coefficients, and c(S) is the weighting coefficient. k Strategy S k The execution cost.

6. The queue spacing strategy reconstruction mechanism according to claim 4, characterized in that, Further information based on information flow to reconstruct the topology Considering error propagation stability theory and traffic flow stability theory, we define vehicle q i With the car in front q i-1 longitudinal spacing d x,i (t)=d0+v i (t)τ+k v (v i (t)-v i-1 (t) and lateral spacing d y,i (t)=w+k u (u i (t)-u i-1 (t)), where d0 is the stationary safety distance, v i (t) represents vehicle q i The actual longitudinal velocity, τ is the reaction time, and k v Let w be the longitudinal coupling coefficient, w be the vehicle width, and k be the [missing value]. u For the lateral coupling coefficient, u i (t) represents the actual lateral velocity of the vehicle, and a two-dimensional spacing d is constructed. i (t)=[d x,i (t),d y,i (t)] T And establish dynamic coupling adjustment equations Where K x K y These are the longitudinal and lateral feedback gain vectors, respectively. , are the longitudinal and lateral position estimates of the vehicle, respectively, and H is the topology-state coupling matrix. Provides information on vehicle driving status estimation.

Images