Model-free control method and system for train pantograph with actuator failure

Abstract

The present invention discloses a model-free control method and system for a train pantograph with an actuator fault. The method comprises collecting dynamic data of a train pantograph operation process; estimating an actuator fault state variable of the train pantograph; establishing a train pantograph data model with an actuator fault; optimizing and updating pseudo partial derivative information in the train pantograph data model; designing a model-free compensation control scheme for a train pantograph with an actuator fault; optimizing and updating the error control gain and the compensation control gain in the control scheme; using the model-free compensation control scheme to calculate a control input signal, and controlling the train pantograph until the control task is completed. The control method and system provided by the present invention are not only applicable to a train pantograph with an actuator fault, but also solve the actuator fault problem, reduce the difficulty of modeling, and reduce the consumption of computing resources, and also have the application prospect of being extended to other single-input and single-output systems.

Description

Technical Field

[0001] The present invention belongs to the field of train pantograph control, and in particular relates to a model-free control method and system for a train pantograph with actuator failure. Background Art

[0002] The pantograph on the roof of the train is the only way for the train to obtain current from the contact network, and its current collection quality is directly related to the operational safety of high-speed railways. Contact force is an important indicator for measuring the current collection quality. When the contact force between the pantograph and the contact network is too large, it will cause excessive wear of the contact wire and the pantograph; when the contact force is too small, it may cause an offline arc between the pantograph and the contact network, and even affect the safety of train operation.

[0003] Active pantograph control is one of the effective means to solve the problem of contact force stability. With the advancement of data acquisition technology, data-driven active pantograph control methods have become a research hotspot. Patents CN115635854A and CN116577991A propose contact force adaptive control methods based on machine learning. It should be noted that machine learning requires a lot of computing resources and time for offline training, which may be limited by the computing power of the controller on the pantograph. In addition, the pantograph operates outdoors for a long time, and its actuator may fail due to long-term action or environmental influences. Once a failure occurs, it will directly affect the current collection quality and thus affect driving safety. The existing technology has not fully considered this problem. Therefore, there is an urgent need for a data-driven online control method for train pantographs that can solve the problem of pantograph actuator failure and have low computing resource consumption to improve train safety and reliability and reduce maintenance costs. Summary of the invention

[0004] In order to solve the problems existing in the background technology, the object of the present invention is to provide a model-free control method for a train pantograph with an actuator failure, the method comprising the following steps:

[0005] Step (1): Collect dynamic data of the train pantograph operation process;

[0006] Step (2): Estimate the actuator fault state variables of the train pantograph;

[0007] Step (3): establishing a train pantograph data model with an actuator failure, wherein the data model includes pseudo partial derivative information of the train pantograph;

[0008] Step (4): constructing a cost function and solving the cost function using a function extremum method, optimizing and updating the pseudo partial derivative information of the train pantograph in step (3);

[0009] Step (5): designing a model-free compensation control scheme for a train pantograph with an actuator failure, wherein the control scheme includes an error control gain and a compensation control gain;

[0010] Step (6): constructing a performance optimization objective function and solving the objective function using a gradient descent method, optimizing and updating the error control gain and the compensation control gain in step (5);

[0011] Step (7): Based on the error control gain and the compensation control gain optimized and updated in step (6), the control input signal is calculated using the model-free compensation control scheme in step (5) to control the train pantograph;

[0012] Repeat the above steps (1) to (7) until the control task is completed.

[0013] According to the model-free control method for a train pantograph with an actuator failure provided by the present invention, the step (1) collects dynamic data of the train pantograph operation process, and the dynamic data at least includes a set value of the contact force between the pantograph and the contact network, and a real-time value of the contact force between the pantograph and the contact network.

[0014] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, in step (2), the estimation algorithm for estimating the actuator fault state variable of the train pantograph is:

[0015] ;

[0016] in, is the sampling time, for The estimated value of the actuator fault state variable of the train pantograph at the sampling moment; is a constant; for Real-time value of contact force between pantograph and catenary at sampling time; for The contact force setting value between the pantograph and the overhead line at the sampling moment.

[0017] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, the actuator fault in step (3) is described as follows:

[0018] ;

[0019] in, and represent multiplication and addition actuator failures, respectively; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time; The control input signal actually received by the pantograph of the train with actuator failure;

[0020] In step (3), a train pantograph data model with an actuator failure is established, and the data model is established as follows:

[0021] ;

[0022] in, is the sampling time, for The real-time value of the contact force between the pantograph and the contact network at the sampling moment, ; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time is: ; for The estimated value of the fault state variable of the train pantograph actuator at the sampling time, ; , for Pseudo partial derivative information of the train pantograph at the sampling moment.

[0023] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, the cost function is constructed in step (4) and the cost function is solved by using the function extreme value method, and the pseudo partial derivative information of the train pantograph in step (3) is optimized and updated, including the following steps:

[0024] Step (4.1): Pseudo partial derivative information Construct the cost function:

[0025] ;

[0026] in, is the sampling time, Represents a function; , for Pseudo partial derivative information of train pantograph at sampling time; for The real-time value of the contact force between the pantograph and the contact network at the sampling moment, ; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time is: ; for The estimated value of the fault state variable of the train pantograph actuator at the sampling time, ; is the first weight coefficient; ;

[0027] Step (4.2): Use the function extreme value method to solve the cost function described in step (4.1), and substitute the estimated value of the actuator fault state variable of the train pantograph obtained in step (2) , get the pseudo partial derivative information of the train pantograph The update law of:

[0028] ;

[0029] in, is the first step length coefficient;

[0030] Step (4.3): Pseudo partial derivative information Construct the cost function:

[0031] ;

[0032] in, is the second weight coefficient; ;

[0033] Step (4.4): Use the function extreme value method to solve the cost function described in step (4.3), and substitute the estimated value of the actuator fault state variable of the train pantograph obtained in step (2) , get the pseudo partial derivative information of the train pantograph The update law of:

[0034] ;

[0035] in, is the second step length coefficient.

[0036] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, the model-free compensation control scheme for a train pantograph with an actuator fault is designed in step (5), and the mathematical expression of the control scheme is:

[0037] ;

[0038] in, is the sampling time; for The train pantograph control input signal calculated at the sampling time; for The error control gain at the sampling time, for The compensation control gain at the sampling time; for The contact force error between the pantograph and the contact network at the sampling time, , for The contact force setting value between the pantograph and the overhead line at the sampling time, for Real-time value of contact force between pantograph and catenary at sampling time; ; for The estimated value of the actuator fault state variable of the train pantograph at the sampling moment; ; , is the controller pseudo order, is an integer.

[0039] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, the step (6) of constructing a performance optimization objective function and solving the objective function using a gradient descent method, and optimizing and updating the error control gain and the compensation control gain in the step (5) include the following steps:

[0040] Step (6.1): Construct performance optimization objective function:

[0041] ;

[0042] in, is the sampling time; for The contact force setting value between the pantograph and the overhead line at the sampling moment; for Real-time value of contact force between pantograph and catenary at sampling time; for The train pantograph control input signal calculated at the sampling time; is the penalty factor;

[0043] Step (6.2): ​​Solve the objective function described in step (6.1) and optimize the updated error control gain :

[0044] ;

[0045] in, for Error control gain at the sampling moment; is the first learning rate; for Pseudo partial derivative information of train pantograph at sampling time; , ;

[0046] Step (6.3): Solve the objective function described in step (6.1) and optimize and update the compensation control gain :

[0047] ;

[0048] in, for The compensation gain matrix at the sampling time; is the second learning rate; , .

[0049] The present invention also provides a model-free control system for a train pantograph with an actuator failure, comprising:

[0050] Data acquisition module, used to collect dynamic data of the train pantograph operation process;

[0051] A controller module, including a fault state estimation submodule, a data model establishment submodule, a pseudo partial derivative update submodule, a control scheme calculation submodule, and a control scheme update submodule;

[0052] The fault state estimation submodule is used to estimate the fault state variable of the actuator of the train pantograph;

[0053] The data model establishment submodule is used to establish a train pantograph data model with actuator failure;

[0054] The pseudo partial derivative updating submodule is used to construct a cost function and solve the cost function using a function extremum method, so as to optimize and update the pseudo partial derivative information of the train pantograph;

[0055] The control scheme calculation submodule is used to design a model-free compensation control scheme for a train pantograph with an actuator failure, wherein the control scheme includes an error control gain and a compensation control gain, which are used to calculate a train pantograph control input signal;

[0056] The control scheme updating submodule is used to construct a performance optimization objective function and solve the objective function using a gradient descent method, and optimize and update the error control gain and compensation control gain of the control scheme calculation module;

[0057] Actuator module, including air supply source, precision pressure regulating valve, and air bag;

[0058] The air supply source is used to provide air pressure to the airbag;

[0059] The precision pressure regulating valve is connected to the air supply source and the airbag, and is used to execute the train pantograph control input signal calculated by the controller module, regulate the gas entering the airbag from the air supply source, and then regulate the air pressure of the airbag;

[0060] The airbag is used to raise and lower the train pantograph according to the different air pressures of the airbag;

[0061] The controlled object module includes a train pantograph, which is used to receive control input signals from the actuator module.

[0062] Furthermore, the present invention adopts the following technical solutions:

[0063] A non-transitory computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned model-free control method for a train pantograph with an actuator failure.

[0064] Furthermore, the present invention adopts the following technical solutions:

[0065] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the model-free control method for a train pantograph with an actuator failure as described above is implemented.

[0066] The beneficial technical effects of the present invention are:

[0067] (1) Solving the problem of actuator failure: The present invention can effectively solve the problem of pantograph actuator failure due to long-term operation or environmental influence, greatly improving the reliability of the pantograph system;

[0068] (2) Reduce modeling difficulty: Adopting a pure data-driven control method, there is no need to establish an accurate dynamic model, which simplifies the control system design and reduces R&D costs;

[0069] (3) Reduce computing resource consumption: It adopts fully online computing, which has lower computing resource consumption and is easier to apply in practice. It is especially suitable for scenarios where the computing power of the pantograph controller is limited.

[0070] (4) Broadening the scope of application: The control method and system proposed in the present invention are not only applicable to train pantographs with actuator failures, but can also be extended to other single-input single-output control systems with actuator failures, and have broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is the calculated mass model of the train pantograph;

[0072] Figure 2A schematic flow chart of a model-free control method for a train pantograph with an actuator failure provided by the present invention;

[0073] Figure 3 A schematic diagram of the connection modules of the model-free control system for a train pantograph with an actuator failure provided by the present invention. DETAILED DESCRIPTION

[0074] The present invention discloses a model-free control method and system for a train pantograph with an actuator failure; the control method includes collecting dynamic data of the train pantograph operation process; estimating the actuator failure state variable of the train pantograph; establishing a train pantograph data model with an actuator failure; optimizing and updating the pseudo-partial derivative information in the train pantograph data model; designing a model-free compensation control scheme for a train pantograph with an actuator failure; optimizing and updating the error control gain and compensation control gain in the control scheme; using the model-free compensation control scheme to calculate the control input signal, and controlling the train pantograph until the control task is completed. The control method and system provided by the present invention are not only applicable to train pantographs with actuator failures, solving the actuator failure problem, reducing the difficulty of modeling, and reducing the consumption of computing resources, but also have the application prospect of being extended to other single-input and single-output systems.

[0075] The following is a further clear and complete description of the model-free control method and system for a train pantograph with an actuator failure provided by the present invention in conjunction with the accompanying drawings:

[0076] Figure 1 The calculated mass model of the train pantograph is given. The train pantograph includes the pantograph head, upper frame and lower frame. m 1. m 2. m 3 are the masses of the pantograph head, upper frame and lower frame respectively; k 1. k 2. k 3 are the stiffness of the pantograph head, upper frame and lower frame respectively; c 1. c 2. c 3 are the damping of the pantograph head, upper frame and lower frame respectively. Figure 2 A flow chart of a model-free control method for a train pantograph with an actuator fault provided by the present invention is given; the present invention provides a model-free control method for a train pantograph with an actuator fault, the method comprising the following steps:

[0077] Step (1): Collect dynamic data of the train pantograph operation process;

[0078] Step (2): Estimate the actuator fault state variables of the train pantograph;

[0079] Step (3): establishing a train pantograph data model with an actuator failure, wherein the data model includes pseudo partial derivative information of the train pantograph;

[0080] Step (4): constructing a cost function and solving the cost function using a function extremum method, optimizing and updating the pseudo partial derivative information of the train pantograph in step (3);

[0081] Step (5): designing a model-free compensation control scheme for a train pantograph with an actuator failure, wherein the control scheme includes an error control gain and a compensation control gain;

[0082] Step (6): constructing a performance optimization objective function and solving the objective function using a gradient descent method, optimizing and updating the error control gain and the compensation control gain in step (5);

[0083] Step (7): Based on the error control gain and the compensation control gain optimized and updated in step (6), the control input signal is calculated using the model-free compensation control scheme in step (5) to control the train pantograph;

[0084] Repeat the above steps (1) to (7) until the control task is completed.

[0085] According to the model-free control method for a train pantograph with an actuator failure provided by the present invention, the step (1) collects dynamic data of the train pantograph operation process, and the dynamic data at least includes a set value of the contact force between the pantograph and the contact network, and a real-time value of the contact force between the pantograph and the contact network.

[0086] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, in step (2), the estimation algorithm for estimating the actuator fault state variable of the train pantograph is:

[0087] ;

[0088] in, is the sampling time, for The estimated value of the actuator fault state variable of the train pantograph at the sampling moment; is a constant; for Real-time value of contact force between pantograph and catenary at sampling time; for The contact force setting value between the pantograph and the overhead line at the sampling moment.

[0089] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, the actuator fault in step (3) is described as follows:

[0090] ;

[0091] in, and represent multiplication and addition actuator failures, respectively; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time; The control input signal actually received by the pantograph of the train with actuator failure;

[0092] In step (3), a train pantograph data model with an actuator failure is established, and the data model is established as follows:

[0093] ;

[0094] in, is the sampling time, for The real-time value of the contact force between the pantograph and the contact network at the sampling moment, ; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time is: ; for The estimated value of the fault state variable of the train pantograph actuator at the sampling time, ; , for Pseudo partial derivative information of the train pantograph at the sampling moment.

[0095] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, the cost function is constructed in step (4) and the cost function is solved by using the function extreme value method, and the pseudo partial derivative information of the train pantograph in step (3) is optimized and updated, including the following steps:

[0096] Step (4.1): Pseudo partial derivative information Construct the cost function:

[0097] ;

[0098] in, is the sampling time, Represents a function; , for Pseudo partial derivative information of train pantograph at sampling time; for The real-time value of the contact force between the pantograph and the contact network at the sampling moment, ; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time is: ; for The estimated value of the fault state variable of the train pantograph actuator at the sampling time, ; is the first weight coefficient; ;

[0099] Step (4.2): Use the function extreme value method to solve the cost function described in step (4.1), and substitute the estimated value of the actuator fault state variable of the train pantograph obtained in step (2) , get the pseudo partial derivative information of the train pantograph The update law of:

[0100] ;

[0101] in, is the first step length coefficient;

[0102] Step (4.3): Pseudo partial derivative information Construct the cost function:

[0103] ;

[0104] in, is the second weight coefficient; ;

[0105] Step (4.4): Use the function extreme value method to solve the cost function described in step (4.3), and substitute the estimated value of the actuator fault state variable of the train pantograph obtained in step (2) , get the pseudo partial derivative information of the train pantograph The update law of:

[0106] ;

[0107] in, is the second step length coefficient.

[0108] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, the model-free compensation control scheme for a train pantograph with an actuator fault is designed in step (5), and the mathematical expression of the control scheme is:

[0109] ;

[0110] in, is the sampling time; for The train pantograph control input signal calculated at the sampling time; for The error control gain at the sampling time, for The compensation control gain at the sampling time; for The contact force error between the pantograph and the contact network at the sampling time, , for The contact force setting value between the pantograph and the overhead line at the sampling time, for Real-time value of contact force between pantograph and catenary at sampling time; ; for The estimated value of the actuator fault state variable of the train pantograph at the sampling moment; ; , is the controller pseudo order, is an integer.

[0111] According to the model-free control method for a train pantograph with an actuator fault provided by the present invention, the step (6) of constructing a performance optimization objective function and solving the objective function using a gradient descent method, and optimizing and updating the error control gain and the compensation control gain in the step (5) include the following steps:

[0112] Step (6.1): Construct performance optimization objective function:

[0113] ;

[0114] in, is the sampling time; for The contact force setting value between the pantograph and the overhead line at the sampling moment; for Real-time value of contact force between pantograph and catenary at sampling time; for The train pantograph control input signal calculated at the sampling time; is the penalty factor;

[0115] Step (6.2): ​​Solve the objective function described in step (6.1) and optimize the updated error control gain :

[0116] ;

[0117] in, for Error control gain at the sampling moment; is the first learning rate; for Pseudo partial derivative information of train pantograph at sampling time; , ;

[0118] Step (6.3): Solve the objective function described in step (6.1) and optimize and update the compensation control gain :

[0119] ;

[0120] in, for The compensation gain matrix at the sampling time; is the second learning rate; , .

[0121] The present invention also provides a model-free control system for a train pantograph with an actuator failure, comprising:

[0122] Data acquisition module, used to collect dynamic data of the train pantograph operation process;

[0123] A controller module, including a fault state estimation submodule, a data model establishment submodule, a pseudo partial derivative update submodule, a control scheme calculation submodule, and a control scheme update submodule;

[0124] The fault state estimation submodule is used to estimate the fault state variable of the actuator of the train pantograph;

[0125] The data model establishment submodule is used to establish a train pantograph data model with actuator failure;

[0126] The pseudo partial derivative updating submodule is used to construct a cost function and solve the cost function using a function extremum method, so as to optimize and update the pseudo partial derivative information of the train pantograph;

[0127] The control scheme calculation submodule is used to design a model-free compensation control scheme for a train pantograph with an actuator failure, wherein the control scheme includes an error control gain and a compensation control gain, which are used to calculate a train pantograph control input signal;

[0128] The control scheme updating submodule is used to construct a performance optimization objective function and solve the objective function using a gradient descent method, and optimize and update the error control gain and compensation control gain of the control scheme calculation module;

[0129] Actuator module, including air supply source, precision pressure regulating valve, and air bag;

[0130] The air supply source is used to provide air pressure to the airbag;

[0131] The precision pressure regulating valve is connected to the air supply source and the airbag, and is used to execute the train pantograph control input signal calculated by the controller module, regulate the gas entering the airbag from the air supply source, and then regulate the air pressure of the airbag;

[0132] The airbag is used to raise and lower the train pantograph according to the different air pressures of the airbag;

[0133] The controlled object module includes a train pantograph, which is used to receive control input signals from the actuator module.

[0134] Furthermore, the present invention adopts the following technical solutions:

[0135] A non-transitory computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned model-free control method for a train pantograph with an actuator failure.

[0136] Furthermore, the present invention adopts the following technical solutions:

[0137] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the model-free control method for a train pantograph with an actuator failure as described above is implemented.

[0138] Through the description of the above embodiments, it can be clearly understood by those skilled in the art that the facilities of the present invention can be implemented by means of software plus the necessary general hardware platform. The embodiments of the present invention can be implemented using existing processors, or by a dedicated processor used for this purpose or other purposes for an appropriate system, or by a hard-wired system. The embodiments of the present invention also include non-transitory computer-readable storage media, which include machine-readable media for carrying or having machine-executable instructions or data structures stored thereon; such machine-readable media can be any available medium that can be accessed by a general or special-purpose computer or other machine with a processor. For example, such machine-readable media can include RAM, ROM, EPROM, EEPROM, CD-ROM or other optical disk storage, disk storage or other magnetic storage devices, or any other medium that can be used to carry or store the required program code in the form of machine-executable instructions or data structures, and can be accessed by a general or special-purpose computer or other machine with a processor. When information is transmitted or provided to a machine via a network or other communication connection (hard-wired, or wireless, or a combination of hard-wired and wireless), the connection is also considered a machine-readable medium.

[0139] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A model-free control method for a train pantograph with an actuator failure, characterized in that: The method comprises the following steps: Step (1): Collect dynamic data of the train pantograph operation process; Step (2): Estimate the actuator fault state variables of the train pantograph; Step (3): establishing a train pantograph data model with an actuator failure, wherein the data model includes pseudo partial derivative information of the train pantograph; Step (4): constructing a cost function and solving the cost function using a function extremum method, optimizing and updating the pseudo partial derivative information of the train pantograph in step (3); Step (5): designing a model-free compensation control scheme for a train pantograph with an actuator failure, wherein the control scheme includes an error control gain and a compensation control gain; Step (6): constructing a performance optimization objective function and solving the objective function using a gradient descent method, optimizing and updating the error control gain and the compensation control gain in step (5); Step (7): Based on the error control gain and the compensation control gain optimized and updated in step (6), the control input signal is calculated using the model-free compensation control scheme in step (5) to control the train pantograph; Repeat the above steps (1) to (7) until the control task is completed; The actuator fault in step (3) is described as: ； in, and represent multiplication and addition actuator failures, respectively; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time; The control input signal actually received by the pantograph of the train with actuator failure; In step (3), a train pantograph data model with actuator failure is established, and the data model is established as follows: ； in, is the sampling time, for The real-time value of the contact force between the pantograph and the contact network at the sampling moment, ; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time is: ; for The estimated value of the fault state variable of the train pantograph actuator at the sampling time, ; , for Pseudo partial derivative information of train pantograph at sampling time; The design of the model-free compensation control scheme for the pantograph of a train with actuator failure in step (5) has a mathematical expression as follows: ； in, is the sampling time; for The train pantograph control input signal calculated at the sampling time; for The error control gain at the sampling time, for The compensation control gain at the sampling time; for The contact force error between the pantograph and the contact network at the sampling time, , for The contact force setting value between the pantograph and the overhead line at the sampling time, for Real-time value of contact force between pantograph and catenary at sampling time; ; for The estimated value of the actuator fault state variable of the train pantograph at the sampling moment; ; , is the controller pseudo order, is an integer.

2. The model-free control method for a train pantograph with an actuator failure according to claim 1, characterized in that: The step (1) collects dynamic data of the pantograph operation process of the train, and the dynamic data at least includes a set value of the contact force between the pantograph and the contact network and a real-time value of the contact force between the pantograph and the contact network.

3. The model-free control method for a train pantograph with an actuator failure according to claim 1, characterized in that: In step (2), the estimation algorithm for estimating the fault state variable of the actuator of the train pantograph is: ； in, is the sampling time, for The estimated value of the actuator fault state variable of the train pantograph at the sampling moment; is a constant; for Real-time value of contact force between pantograph and catenary at sampling time; for The contact force setting value between the pantograph and the overhead line at the sampling moment.

4. The model-free control method for a train pantograph with an actuator failure according to claim 1, characterized in that: The cost function is constructed in step (4) and the cost function is solved by using the function extreme value method to optimize and update the pseudo partial derivative information of the train pantograph in step (3), including the following steps: Step (4.1): Pseudo partial derivative information Construct the cost function: ； in, is the sampling time, Represents a function; , for Pseudo partial derivative information of train pantograph at sampling time; for The real-time value of the contact force between the pantograph and the contact network at the sampling moment, ; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time is: ; for The estimated value of the fault state variable of the train pantograph actuator at the sampling time, ; is the first weight coefficient; ; Step (4.2): Use the function extreme value method to solve the cost function described in step (4.1), and substitute the estimated value of the actuator fault state variable of the train pantograph obtained in step (2) , get the pseudo partial derivative information of the train pantograph The update law of: ； in, is the first step length coefficient; Step (4.3): Pseudo partial derivative information Construct the cost function: ； in, is the second weight coefficient; ; Step (4.4): Use the function extreme value method to solve the cost function described in step (4.3), and substitute the estimated value of the actuator fault state variable of the train pantograph obtained in step (2) , get the pseudo partial derivative information of the train pantograph The update law of: ； in, is the second step length coefficient.

5. The model-free control method for a train pantograph with an actuator failure according to claim 1, characterized in that: The step (6) of constructing a performance optimization objective function and solving the objective function using a gradient descent method, and optimizing and updating the error control gain and the compensation control gain in the step (5), comprises the following steps: Step (6.1): Construct performance optimization objective function: ； in, is the sampling time; for The contact force setting value between the pantograph and the overhead line at the sampling moment; for Real-time value of contact force between pantograph and catenary at sampling time; for The train pantograph control input signal calculated at the sampling time; is the penalty factor; Step (6.2): ​​Solve the objective function described in step (6.1) and optimize the updated error control gain : ； in, for Error control gain at the sampling moment; is the first learning rate; for Pseudo partial derivative information of train pantograph at sampling time; , ; , for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time; Step (6.3): Solve the objective function described in step (6.1) and optimize and update the compensation control gain : ； in, for The compensation gain matrix at the sampling time; is the second learning rate; , .

6. A model-free control system for a train pantograph with actuator failure, characterized in that: include: Data acquisition module, used to collect dynamic data of the train pantograph operation process; A controller module, including a fault state estimation submodule, a data model establishment submodule, a pseudo partial derivative update submodule, a control scheme calculation submodule, and a control scheme update submodule; The fault state estimation submodule is used to estimate the fault state variable of the actuator of the train pantograph; The data model establishment submodule is used to establish a train pantograph data model with actuator failure; The pseudo partial derivative updating submodule is used to construct a cost function and solve the cost function using a function extremum method, so as to optimize and update the pseudo partial derivative information of the train pantograph; The control scheme calculation submodule is used to design a model-free compensation control scheme for a train pantograph with an actuator failure, wherein the control scheme includes an error control gain and a compensation control gain, which are used to calculate a train pantograph control input signal; The control scheme updating submodule is used to construct a performance optimization objective function and solve the objective function using a gradient descent method, and optimize and update the error control gain and compensation control gain of the control scheme calculation module; Actuator module, including air supply source, precision pressure regulating valve, and air bag; The air supply source is used to provide air pressure to the airbag; The precision pressure regulating valve is connected to the air supply source and the airbag, and is used to execute the train pantograph control input signal calculated by the controller module, regulate the gas entering the airbag from the air supply source, and then regulate the air pressure of the airbag; The airbag is used to raise and lower the train pantograph according to the different air pressures of the airbag; The controlled object module includes a train pantograph, which is used to receive control input signals from the actuator module; The actuator fault description is: ； in, and represent multiplication and addition actuator failures, respectively; for The train pantograph control input signal calculated by the model-free compensation control scheme described in step (5) at the sampling time; The control input signal actually received by the pantograph of the train with actuator failure; The train pantograph data model with actuator failure is established, and the data model is established as follows: ； in, is the sampling time, for The real-time value of the contact force between the pantograph and the contact network at the sampling moment, ; for The train pantograph control input signal calculated by the model-free compensation control scheme at the sampling time, ; for The estimated value of the fault state variable of the train pantograph actuator at the sampling time, ; , for Pseudo partial derivative information of train pantograph at sampling time; The design is a model-free compensation control scheme for a train pantograph with an actuator failure, and the mathematical expression of the control scheme is: ； in, is the sampling time; for The train pantograph control input signal calculated at the sampling time; for The error control gain at the sampling time, for The compensation control gain at the sampling time; for The contact force error between the pantograph and the contact network at the sampling time, , for The contact force setting value between the pantograph and the overhead line at the sampling time, for Real-time value of contact force between pantograph and catenary at sampling time; ; for The estimated value of the actuator fault state variable of the train pantograph at the sampling moment; ; , is the controller pseudo order, is an integer.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the model-free control method for a train pantograph with an actuator failure as described in any one of claims 1 to 5 is implemented.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the model-free control method for a train pantograph with an actuator failure as described in any one of claims 1 to 5 is implemented.

Images