High-speed train fault-tolerant safety control method based on iterative learning strategy

Abstract

The invention discloses a high-speed train fault-tolerant safety control method based on an iterative learning strategy, and relates to the technical field of high-speed train safety control, and the method comprises the following steps: 1, dynamic modeling: building a dynamic model considering train repeated operation characteristics, resistance and fault influence; step 2, fault estimation: designing an estimator based on an iterative learning strategy to estimate train operation fault influence; and step 3, fault-tolerant safety control: designing a train fault-tolerant safety controller based on a finite convergence iterative learning control framework. According to the high-speed train fault-tolerant safety control method based on the iterative learning strategy, the accuracy of fault estimation is improved by using historical data generated in the repeated operation of the train; according to the method, a finite convergence iterative learning controller is designed, the number of times of repeated operation is remarkably reduced through an updating mechanism of an iterative learning gain matrix, and the requirements for real-time performance and high efficiency of operation control of a high-speed train system are effectively met.

Description

Technical Field

[0001] The invention relates to the technical field of high-speed train safety control, and in particular to a high-speed train fault-tolerant safety control method based on an iterative learning strategy. Background Art

[0002] In the field of safety control of high-speed trains, existing research and technology have explored a variety of methods to improve safety and reliability. Traditional control strategies such as PID controllers and adaptive control have been widely used, but they have limitations in dealing with nonlinear systems, parameter changes, and uncertainties. In recent years, with the growing demand for intelligent transportation systems, learning-based methods have gradually attracted attention, among which iterative learning control has shown potential due to its unique characteristics. However, current applications based on iterative learning control are mainly concentrated in areas such as robot arm motion control in laboratory environments. In actual industrial applications, especially in complex and high-risk environments such as high-speed rail, the application of ILC is still in its early stages. Some studies that attempt to apply ILC to rail transportation mainly focus on how to improve the smoothness and energy efficiency of trains during acceleration / deceleration, but there are few designs for fault-tolerant mechanisms. In addition, in practical applications, especially in real-time systems such as high-speed trains, the limitations of computing time and number of iterations must be considered to ensure rapid convergence to satisfactory control performance, which is also a consideration lacking in existing technologies. Summary of the invention

[0003] The purpose of the present invention is to provide a high-speed train fault-tolerant safety control method based on an iterative learning strategy, to design an iterative learning controller with finite convergence, and to effectively meet the real-time and high-efficiency requirements of the operation control of the high-speed train system through the updating mechanism of the iterative learning gain matrix.

[0004] To achieve the above object, the present invention provides a high-speed train fault-tolerant safety control method based on an iterative learning strategy, comprising the following steps:

[0005] Step 1: Dynamic modeling: Establish a dynamic model that takes into account the train's repeated operation characteristics, resistance, and fault effects;

[0006] Step 2: Fault estimation: design an estimator based on iterative learning strategy to estimate the impact of train operation faults

[0007] Step 3: Fault-tolerant safety control: design a train fault-tolerant safety controller based on a finite-time convergent iterative learning control framework.

[0008] Preferably, in step 1, based on the high-speed train dynamics equation, the following train repetitive model is established:

[0009]

[0010] in, is the first-order derivative of the system state variable with respect to time during the k-th operation; k represents the number of repeated operations of the train; x k is the system state variable at the kth run, and x k =[s k v k ] T ,s k and v k They represent the running distance and running speed of the train during the kth operation respectively; u k represents the system control input during the k-th run; represents the disturbance caused by external resistance during the kth operation; f is the fault impact.

[0011] Preferably, step 2 comprises the following steps:

[0012] S21, construct Luenberger observer;

[0013]

[0014] in, It is the first-order derivative of the observed value of the system state variable with respect to time during the k-th run; is the observed value of the system state variables during the k-th run, represents the observed value of the disturbance caused by the external resistance during the k-th run, is the virtual fault during the k-th operation; L is the observer gain matrix;

[0015] S22. Design a fault estimator based on an open-closed-loop PD-type iterative learning strategy:

[0016]

[0017] Among them, α1 is the open-loop proportional term gain, α2 is the closed-loop proportional term gain, β1 is the open-loop proportional term gain, and β2 is the closed-loop proportional term gain.

[0018] Preferably, in step 2, the high-speed train system state variable observation residual is defined as:

[0019]

[0020] Preferably, step three includes the following steps:

[0021] S31, high-speed train performs periodic repetitive operation tasks, and its task target tracking is x d (t), the operating interval is t∈[0,T];

[0022] S32, the finite convergence iterative learning controller is:

[0023] u k+1 (t) = u k (t)+Q(t)e k (t);

[0024] Among them, e k is the state tracking error during the k-th run, Q(t) is the variable learning gain matrix;

[0025] S33, the fault-tolerant safety control law with feedforward is:

[0026]

[0027] Among them, u k * (t) represents the final fault-tolerant safety control input, C = [0M].

[0028] Preferably, in S32, to achieve finite convergence of the iterative learning controller, the variable learning gain matrix Q(t) follows the following update mechanism:

[0029] Q(t)=KP(t)[P(t)+R] -1 ;

[0030] Among them, K is a positive definite diagonal matrix used to adjust the learning rate; P(t) is the solution of the Riccati equation, reflecting the state covariance of the system; R is the positive definite weight matrix.

[0031] Therefore, the present invention adopts the above-mentioned high-speed train fault-tolerant safety control method based on iterative learning strategy, which has the following effects:

[0032] (1) Considering the impact of faults on high-speed trains, a fault estimator based on an open-closed-loop PD-type iterative learning strategy is proposed. The historical data generated during the repeated operation of the train can be used to achieve rapid and effective estimation of the fault. The accuracy of fault estimation is continuously improved with the repeated operation of the train, which is conducive to further fault-tolerant safety control.

[0033] (2) By utilizing the historical data generated during repeated train operation, an iterative learning controller and an iterative learning gain matrix update mechanism are designed. This can achieve good state tracking results after a limited number of repeated operations. Compared with other similar methods, the stable value of the tracking error is significantly reduced, and the number of repeated operations required to converge to a lower stable value also has a clear advantage.

[0034] (3) Based on accurate fault estimation results and a finite-time convergence iterative learning controller, combined with the use of feedforward compensation, a fault-tolerant safety control scheme based on an iterative learning strategy is formed, which can effectively suppress the impact of faults on train operation and achieve accurate operation status tracking.

[0035] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a flow chart of the method of the present invention;

[0037] Figure 2 This is a comparison diagram of the fault estimation effect of the present invention;

[0038] Figure 3 This is a comparison chart of the fault-tolerant safety control effect of the present invention. DETAILED DESCRIPTION

[0039] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0040] Unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0041] Example

[0042] See also Figure 1-3 The present invention provides a high-speed train fault-tolerant safety control method based on an iterative learning strategy, comprising the following steps:

[0043] Step 1: Dynamic modeling of high-speed trains: Analyze the dynamic characteristics of high-speed trains, combine their operating environment and periodic driving characteristics, and establish a dynamic model of train repeated operation that takes into account faults, disturbances and uncertainties.

[0044] According to Newton's laws of mechanics, the dynamic equation of a high-speed train considering the effects of running resistance and faults is as follows:

[0045]

[0046] in, and They are respectively the first-order derivative of the train displacement with respect to time and the first-order derivative of the train speed with respect to time; s is the train displacement, v is the train speed, M represents the total mass of the train, u is the control input, ξ is the combination of resistance, fault and other effects on the system, a, b, c are the basic resistance coefficients, and f is the effect of the fault factor.

[0047] Based on the high-speed train dynamics equation, the following train repeated operation model is further established:

[0048]

[0049] in, is the first-order derivative of the system state variable with respect to time during the k-th operation; k represents the number of repeated operations of the train; x k is the system state variable at the kth run, and x k =[s k v k ] T , where s k and v k They represent the running distance and running speed of the train during the kth operation respectively; u k represents the system control input during the k-th run; represents the disturbance caused by the external resistance during the k-th run, and f is the fault impact.

[0050] Step 2: High-speed train fault estimation based on iterative learning strategy; construct a train operation status observer to online estimate the actual operation status of the train, including key parameters such as speed, position, acceleration, etc.; based on the observation residuals, design a lumped fault estimator based on iterative learning strategy to gradually improve the accuracy of fault estimation with repeated operation.

[0051] After establishing the high-speed train repeated operation dynamic model, the following Luenberger observer is constructed:

[0052]

[0053] in, It is the first-order derivative of the observed value of the system state variable with respect to time during the k-th run; is the observed value of the system state variables during the k-th run, represents the observed value of the disturbance caused by the external resistance during the k-th run, is the virtual fault during the k-th operation; L is the observer gain matrix.

[0054] The high-speed train system state variable observation residual is defined as follows:

[0055]

[0056] Based on the above observation residuals, the following fault estimator based on open-closed-loop PD-type iterative learning strategy is designed:

[0057]

[0058] Among them, α1 is the open-loop proportional term gain, α2 is the closed-loop proportional term gain, β1 is the open-loop proportional term gain, and β2 is the closed-loop proportional term gain. Through the use of iterative learning mechanism, the historical data generated by high-speed trains during repeated periodic operation is fully mined, the data from the previous operation is used to accelerate the convergence speed of fault estimation, and the dynamic performance of fault estimation is further guaranteed by combining the data generated during the current operation.

[0059] Comparing this fault estimation method with other estimation methods, it can be shown that the proposed method is effective for fault estimation, and the estimation accuracy is significantly improved with repeated operation. Figure 2 shown.

[0060] Step 3. Design a train fault-tolerant safety controller based on a finite-time convergent iterative learning control framework; design a finite-time convergent iterative learning controller that can complete a complete iterative process in a short time; combine the fault estimation results to design a feedforward control module to compensate for the negative impact of the fault in advance; combine the finite-time iterative learning controller with the feedforward control based on fault estimation to form a new high-speed train fault-tolerant safety control strategy.

[0061] First, consider the high-speed train performing periodic repetitive operation tasks, and its task target tracking is x d (t), the operating interval is t∈[0,T].

[0062] The state tracking error is defined as

[0063] e k (t) = x d (t)-x k (t);

[0064] Design a finite convergence iterative learning controller as

[0065] u k+1 (t) = u k (t)+Q(t)e k (t);

[0066] Among them, ek is the state tracking error during the kth run, and Q(t) is the variable learning gain matrix.

[0067] In order to achieve finite convergence of the iterative learning controller, the variable learning gain matrix Q(t) follows the following update mechanism:

[0068] Q(t)=KP(t)[P(t)+R] -1 ;

[0069] Among them, K is a positive definite diagonal matrix used to adjust the learning rate; P(t) is the solution of the Riccat i equation, reflecting the state covariance of the system; R is the positive definite weight matrix.

[0070] Based on the results of the finite convergence iterative learning controller and fault estimation, the following fault-tolerant safety control law with feedforward is designed:

[0071]

[0072] Among them, u k * (t) represents the final fault-tolerant safety control input, C = [0M].

[0073] like Figure 3 As shown, compared with other iterative learning controllers, the fault-tolerant safety control method of the present invention can achieve convergence in a finite number of iterations, and the tracking error index is much lower than that of other existing iterative learning methods; based on accurate fault estimation results, the use of feedforward compensation can well compensate for the impact of the fault on the train operation performance.

[0074] Therefore, the present invention adopts the above-mentioned high-speed train fault-tolerant safety control method based on iterative learning strategy, which can utilize the historical data generated during the repeated operation of the train to improve the accuracy of fault estimation; an iterative learning controller with finite convergence is designed, and the number of required repeated operations is significantly reduced through the update mechanism of the iterative learning gain matrix, effectively meeting the real-time and high-efficiency requirements of the operation control of the high-speed train system.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A fault-tolerant safety control method for high-speed trains based on iterative learning strategy, characterized in that: The following steps are involved: Step 1: Dynamic modeling: Establish a dynamic model that takes into account the train's repeated operation characteristics, resistance, and fault effects; Step 2: Fault estimation: design an estimator based on iterative learning strategy to estimate the impact of train operation faults Step 3: Fault-tolerant safety control: design a train fault-tolerant safety controller based on a finite-time convergent iterative learning control framework.

2. The method for fault-tolerant safety control of high-speed trains based on iterative learning strategy according to claim 1 is characterized in that: Step 1: Based on the high-speed train dynamics equation, the following train repetitive model is established: in, is the first-order derivative of the system state variable with respect to time during the k-th operation; k represents the number of repeated operations of the train; x k is the system state variable at the kth run, and x k =[s k v k ] T ,s k and v k They represent the running distance and running speed of the train during the kth operation respectively; u k represents the system control input during the k-th run; represents the disturbance caused by external resistance during the kth operation; f is the fault impact.

3. A high-speed train fault-tolerant safety control method based on iterative learning strategy according to claim 2, characterized in that: Step 2 includes the following steps: S21, construct Luenberger observer; in, It is the first-order derivative of the observed value of the system state variable with respect to time during the k-th run; is the observed value of the system state variables during the k-th run, represents the observed value of the disturbance caused by the external resistance during the k-th run, is the virtual fault during the k-th operation; L is the observer gain matrix; S22. Design a fault estimator based on an open-closed-loop PD-type iterative learning strategy: Among them, α1 is the open-loop proportional term gain, α2 is the closed-loop proportional term gain, β1 is the open-loop proportional term gain, and β2 is the closed-loop proportional term gain.

4. The method for fault-tolerant safety control of high-speed trains based on iterative learning strategy according to claim 3 is characterized in that: In step 2, the observed residual of the high-speed train system state variables is defined as:

5. A high-speed train fault-tolerant safety control method based on iterative learning strategy according to claim 4, characterized in that: Step three includes the following steps: S31, high-speed train performs periodic repetitive operation tasks, and its task target tracking is x d (t), the operating interval is t∈[0,T]; S32, the finite convergence iterative learning controller is: u k+1 (t)=u k (t)+Q(t)e k (t); Among them, e k is the state tracking error during the k-th run, Q(t) is the variable learning gain matrix; S33, the fault-tolerant safety control law with feedforward is: Among them, u k * (t) represents the final fault-tolerant safety control input, C = [0M].

6. The method for fault-tolerant safety control of high-speed trains based on iterative learning strategy according to claim 5 is characterized in that: In S32, in order to achieve finite convergence of the iterative learning controller, the variable learning gain matrix Q(t) follows the following update mechanism: Q(t)=KP(t)[P(t)+R] -1 ; Among them, K is a positive definite diagonal matrix used to adjust the learning rate; P(t) is the solution of the Riccati equation, reflecting the state covariance of the system; R is the positive definite weight matrix.

Images