Heavy haul train control method and system based on vehicle dynamic response identification

Abstract

The application discloses a heavy haul train control method and system based on vehicle dynamic response identification, and relates to the technical field of train control.The method comprises the following steps: dividing the carriages of the heavy haul train into multiple blocks according to the differences in the dynamic responses of different carriages of the heavy haul train; taking the tractive force or braking force of each block as input and the train speed as output to establish a vehicle operation model for each block; using a Smith predictor to compensate for the time delay characteristics of the speed of the heavy haul train under complex line conditions; when the speed variation of the heavy haul train or the acceleration variation under stepless speed regulation exceeds the set range, using a fuzzy PID-Smith time delay controller to control the output speed of the vehicle operation model; and when the speed tracking accuracy of the heavy haul train exceeds the set range, using an improved particle swarm algorithm to dynamically adjust the parameters of the fuzzy PID-Smith time delay controller.The application improves the parameter setting accuracy and the reliability of the control under the time delay characteristics of the heavy haul train.

Description

Technical Field

[0001] This invention relates to the field of train control technology, and in particular to a control method and system for heavy-haul trains based on vehicle dynamic response identification. Background Technology

[0002] Heavy-haul railway transportation, due to its advantages such as large capacity, low cost, and environmental friendliness, has become the main mode of bulk and medium-to-long-distance freight transportation worldwide, both now and in the future. However, the longitudinal impact load on trains and the safety of locomotives and rolling stock under pressure remain bottlenecks affecting safe transportation. Compared to passenger trains, heavy-haul trains exhibit significant differences in longitudinal dynamic characteristics and dynamic response. In terms of vehicle control, heavy-haul trains have significant time delays (driver reaction time, control command transmission time, and vehicle response time), resulting in substantial longitudinal impacts and coupler forces near the train's variable operating point, which can lead to safety accidents such as coupler separation and breakage, severely hindering the healthy development of heavy-haul railways.

[0003] Due to the lack of a reliable control model, the traditional multi-mass model is too computationally complex to meet the requirements of real-time computing by onboard computers. The automation capability of heavy-haul train operation is insufficient, and the safe operation control process mainly relies on the driver and crew. This is too dependent on the skills and experience of the driver and crew, resulting in high labor intensity and high operational safety risks. Summary of the Invention

[0004] The purpose of this invention is to provide a control method and system for heavy-haul trains based on vehicle dynamic response identification, thereby improving the control accuracy of heavy-haul trains.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] A heavy-haul train control method based on vehicle dynamic response identification includes:

[0007] Based on the differences in dynamic response of different vehicles in a heavy-haul train, the carriages of the heavy-haul train are divided into multiple blocks.

[0008] Based on the common input and output of the mechanistic characteristics and dynamic response of each block, the vehicle operation model of each block is established with the traction force or braking force of each block as input and the train speed as output. That is, the first-order time delay model is used to equivalently describe the vehicle dynamic response. The mechanistic characteristics are the longitudinal kinematic equations of the vehicle; the dynamic response is the variation law of the vehicle interface signal.

[0009] A first-order time-delay model and a Schmis predictor are integrated to compensate for the train speed time-delay characteristics under complex track conditions for heavy-haul trains. The control signals of the corresponding blocks are then adjusted based on the compensated train speed. The complex track is defined as an operational scenario where the algebraic difference in track gradient length exceeds 30‰ and the coupler force exceeds 2250KN, posing a safety hazard of coupler breakage. The control signal is either the traction force or the braking force, and the vehicle operation model output is the train speed.

[0010] Under complex track conditions for heavy-haul trains, when the speed change of the heavy-haul train per unit time exceeds a first threshold, or the acceleration change under stepless speed regulation exceeds a second threshold, a fuzzy PID-Smith time-delay controller is designed based on the vehicle operation model for speed tracking control. The fuzzy PID-Smith time-delay controller is constructed based on different speed control strategies corresponding to different handle levels. The input to the fuzzy PID-Smith time-delay controller is the traction force or braking force, and the output is the speed. During the speed tracking control process under complex track conditions for heavy-haul trains, when the speed tracking accuracy exceeds a third threshold, an improved particle swarm optimization algorithm is used to dynamically adjust the parameters of the fuzzy PID-Smith time-delay controller. The first threshold is an overshoot of no more than 4% under a square wave control signal; the second threshold is an acceleration change of no more than 0.47 m / s² within two sampling periods under emergency braking. -2 The third threshold is a constant speed control error of less than 0.1 km / h.

[0011] Optionally, the vehicle operation model is represented as:

[0012] y(k)=-a1y(k-1)-a2y(k-1-τ)+b0u(k-1)+ξ(k);

[0013] Where y(k) represents the train speed at time k, y(k-1) represents the train speed at time k-1, y(k-1-τ) represents the train speed at time k-1-τ, τ represents the delay constant, u(k-1) represents the traction force at time k-1, ξ(k) represents the noise sequence at time k, a1 represents the first parameter, a2 represents the second parameter, and b0 represents the third parameter.

[0014] Optionally, it also includes:

[0015] The first parameter, the second parameter, and the third parameter in the vehicle operation model are identified using a recursive least squares algorithm with a forgetting factor.

[0016] Optionally, the particle update strategy of the improved particle swarm optimization algorithm is expressed as:

[0017]

[0018] Where, x t x represents the position of the particle at time t. t+1 V represents the position of the particle at time t+1. t v represents the velocity of the particle at time t. t+1 Let P represent the velocity of the particle at time t+1, w be the inertial weight, c1 and c2 be learning factors, r1 and r2 be random numbers between [0, 1], and P be the velocity of the particle at time t+1. t For the optimal position of a single particle at time t, G t Let t be the optimal position of the entire particle swarm at time t.

[0019] Optionally, the difference in dynamic response is specifically a difference in velocity response.

[0020] This invention discloses a heavy-haul train control system based on vehicle dynamic response identification, comprising:

[0021] The block division module is used to divide the carriages of a heavy-haul train into multiple blocks based on the differences in the dynamic response of different carriages.

[0022] The vehicle operation model building module is used to build vehicle operation models for each block based on the common input and output of the mechanism characteristics and dynamic response of each block, with the traction force or braking force of each block as input and the train speed as output. That is, the first-order time delay model is used to equivalently describe the vehicle dynamic response.

[0023] The Schmis predictor feedback adjustment module is used to compensate for the train speed time delay characteristics under complex track conditions of heavy-haul trains by integrating a first-order time delay model and a Schmis predictor, and to provide feedback adjustment to the control signals of the corresponding blocks based on the compensated train speed. The complex track is an operational scenario where the algebraic difference of the track slope length exceeds 30‰ within every 400 meters and the coupler force exceeds 2250KN, posing a safety hazard of coupler breakage. The control signal is the traction force or the braking force, and the vehicle operation model output is the train speed.

[0024] A fuzzy PID-Smith time-delay controller module is used for speed tracking control under complex heavy-load train conditions. When the speed change of the heavy-load train per unit time exceeds a first threshold, or the acceleration change under stepless speed regulation exceeds a second threshold, a fuzzy PID-Smith time-delay controller is designed based on the vehicle operation model for speed tracking control. The fuzzy PID-Smith time-delay controller is constructed based on different speed control strategies corresponding to different handle levels. The input of the fuzzy PID-Smith time-delay controller is the traction force or braking force, and the output is the speed. During speed tracking control under complex heavy-load train conditions, when the speed tracking accuracy exceeds a third threshold, an improved particle swarm optimization algorithm is used to dynamically adjust the parameters of the fuzzy PID-Smith time-delay controller. The first threshold is an overshoot of no more than 4% under a square wave control signal; the second threshold is an acceleration change of no more than 0.47 m / s² within two sampling periods under emergency braking. -2 The third threshold is a constant speed control error of less than 0.1 km / h.

[0025] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0026] This invention addresses the optimization control problem of heavy-haul trains with large time delay characteristics and discloses a control method for heavy-haul trains based on vehicle dynamic response identification. According to the differences in dynamic response of different carriages of the heavy-haul train, the carriages of the heavy-haul train are divided into multiple blocks.

[0027] Based on the common input and output of the mechanistic characteristics and dynamic response of each block, the vehicle operation model of each block is established with the traction force or braking force of each block as input and the train speed as output. That is, the first-order time delay model is used to equivalently describe the vehicle dynamic response. The mechanistic characteristics are the longitudinal kinematic equations of the vehicle; the dynamic response is the variation law of the vehicle interface signal.

[0028] A first-order time-delay model and a Schmis predictor are integrated to compensate for the train speed time-delay characteristics under complex track conditions for heavy-haul trains. The control signals of the corresponding blocks are then adjusted based on the compensated train speed. The complex track is defined as an operational scenario where the algebraic difference in track gradient length exceeds 30‰ and the coupler force exceeds 2250KN, posing a safety hazard of coupler breakage. The control signal is either the traction force or the braking force, and the vehicle operation model output is the train speed.

[0029] Under complex track conditions for heavy-haul trains, when the speed change of the heavy-haul train per unit time exceeds a first threshold, or the acceleration change under stepless speed regulation exceeds a second threshold, a fuzzy PID-Smith time-delay controller is designed based on the vehicle operation model for speed tracking control. The fuzzy PID-Smith time-delay controller is constructed based on different speed control strategies corresponding to different handle levels. The input to the fuzzy PID-Smith time-delay controller is the traction force or braking force, and the output is the speed. During the speed tracking control process under complex track conditions for heavy-haul trains, when the speed tracking accuracy exceeds a third threshold, an improved particle swarm optimization algorithm is used to dynamically adjust the parameters of the fuzzy PID-Smith time-delay controller. The first threshold is an overshoot of no more than 4% under a square wave control signal; the second threshold is an acceleration change of no more than 0.47 m / s² within two sampling periods under emergency braking. -2 The third threshold is a constant speed control error of less than 0.1 km / h. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 A schematic flowchart of a heavy-haul train control method based on vehicle dynamic response identification provided in an embodiment of the present invention;

[0032] Figure 2 A schematic diagram of a blockchain-based heavy-haul train structure provided in an embodiment of the present invention;

[0033] Figure 3 This is a schematic diagram of the least squares modeling error analysis with forgetting factor provided in an embodiment of the present invention;

[0034] Figure 4 A schematic diagram of a fuzzy PID-Smith control system model based on an improved PSO algorithm provided in an embodiment of the present invention;

[0035] Figure 5 This is a schematic diagram of the PSO-fuzzy PID-Smith controller principle provided in an embodiment of the present invention;

[0036] Figure 6 A schematic diagram illustrating the iterative change of the optimal individual fitness value based on the improved PSO algorithm, provided for an embodiment of the present invention;

[0037] Figure 7 This is a schematic diagram of the PSO-fuzzy PID-Smith control effect curve provided in an embodiment of the present invention;

[0038] Figure 8 This is a comparison curve of disturbance recovery capability provided in an embodiment of the present invention;

[0039] Figure 9 A schematic diagram showing the comparison of speed control errors provided in an embodiment of the present invention;

[0040] Figure 10 The speed tracking curve diagram of PSO-fuzzy PID-Smith control provided in the embodiments of the present invention;

[0041] Figure 11 This is a schematic diagram of a heavy-haul train control system based on vehicle dynamic response identification, provided as an embodiment of the present invention. Detailed Implementation

[0042] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] The purpose of this invention is to provide a control method and system for heavy-haul trains based on vehicle dynamic response identification, thereby improving the control accuracy of heavy-haul trains.

[0044] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0045] Example 1

[0046] like Figure 1 As shown in the figure, this embodiment provides a heavy-load train control method based on vehicle dynamic response identification, which includes the following steps.

[0047] S101: Based on the differences in dynamic response of different carriages of a heavy-haul train, the carriages of a heavy-haul train are divided into multiple blocks.

[0048] S102: Based on the common input and output of the mechanism characteristics and dynamic response of each block, the vehicle operation model of each block is established with the traction force or braking force of each block as input and the train speed as output. That is, the first-order time delay model is used to equivalently describe the vehicle dynamic response.

[0049] Among them, the mechanistic characteristics are the longitudinal kinematic equations of the vehicle; the dynamic response is the variation law of the vehicle interface signal.

[0050] This embodiment first introduces vehicle dynamic response identification technology. Based on the differences in the dynamic response (speed, acceleration, and coupler force) of different vehicles in long-formation heavy-haul trains under complex track conditions, the longitudinal dynamic model of the heavy-haul train is equivalent to a system with different blocks coupled together. The speed, coupler force, gradient, etc. during the train operation are tracked and recorded. Combining the time delay response characteristics of the block model and the available field data (locomotive data, train operation data, and track data), a reliable data-driven vehicle operation model is established.

[0051] This embodiment analyzes the actual operating data (traction / braking force and speed for locomotives; coupler force and speed for trailers) in the blocks to uncover the least squares structure common to the vehicle dynamic response identification model and the dynamics model. It uses a recursive least squares algorithm with a forgetting factor to identify the model parameters, realize the difference in dynamic response in different blocks, and thus verify the hierarchical phenomenon of vehicle dynamics of heavy-haul trains.

[0052] As a specific implementation method, the difference in dynamic response is specifically the difference in velocity response.

[0053] For the vehicle dynamic response of the block, a second-order system is used to describe its characteristics, which can be equivalently represented by a first-order model with time delay in control. Using a first-order model with time delay to approximate the vehicle dynamic response fully meets the requirements of practical applications. The relationship between the traction transfer function F(s) and the velocity transfer function V(s) is as follows:

[0054]

[0055] Where K is the proportionality coefficient, T is the inertial time constant, τ is the delay constant, and s represents the complex variable.

[0056] By identifying the least squares structure shared by the vehicle dynamic response identification model and the dynamics model, the discretized structure of equation (1) can be used to describe the train traction process using a first-order autoregressive model with time delay. The autoregressive model (vehicle operation model) is as follows:

[0057]

[0058] in, Let θ = [a1 a2 b0] be the data vector, y(k) be the train speed at time k, y(k-1) be the train speed at time k-1, y(k-1-τ) be the train speed at time k-1-τ, τ be the delay constant, u(k-1) be the traction force at time k-1, ξ(k) be the noise sequence at time k, a1 be the first parameter (the first parameter to be identified), a2 be the second parameter (the second parameter to be identified), and b0 be the third parameter (the third parameter to be identified).

[0059] This embodiment of a heavy-haul train control method based on vehicle dynamic response identification also includes:

[0060] The recursive least squares algorithm with a forgetting factor is used to identify the first, second, and third parameters in the vehicle operation model, i.e., the unknown parameters in equation (2), as follows:

[0061]

[0062] Where λ is the forgetting factor, and λ is usually chosen in the range of (0.9~1.0), P(0)=(10 4 ~10 10 I, where I is the identity matrix, and the vector θ can be estimated through recursive iteration. T denotes the transpose, P(k) denotes the gain matrix of the iteration, and K(k) denotes the gain matrix associated with the observation vector. This represents the estimated value of the unknown parameter; k is a natural number greater than or equal to 1.

[0063] S103: The train speed time delay characteristics under complex track conditions of heavy-load trains are compensated by integrating a first-order time delay model and a Schmis predictor, and the control signal of the corresponding block is adjusted by feedback according to the compensated train speed.

[0064] Among them, the complex lines for heavy-haul trains are operation scenarios where the algebraic difference of the slope length within every 400 meters exceeds 30‰ and the coupler force exceeds 2250KN, resulting in the safety hazard of coupler breakage; the control signal is traction force or braking force, and the output of the vehicle operation model is the train speed.

[0065] S104: Under the complex conditions of heavy-haul trains, when the speed change of a heavy-haul train per unit time exceeds the first threshold, or the acceleration change under stepless speed regulation exceeds the second threshold, a fuzzy PID-Smith time-delay controller is designed based on the vehicle operation model to perform speed tracking control. The fuzzy PID-Smith time-delay controller is constructed based on different speed control strategies corresponding to different handle levels. The input of the fuzzy PID-Smith time-delay controller is traction force or braking force, and the output of the fuzzy PID-Smith time-delay controller is speed. During the speed tracking control process under the complex conditions of heavy-haul trains, when the speed tracking accuracy exceeds the third threshold, an improved particle swarm optimization algorithm is used to dynamically adjust the parameters of the fuzzy PID-Smith time-delay controller.

[0066] The first threshold is that the overshoot under a square wave control signal does not exceed 4%; the second threshold is that the change in acceleration within two sampling periods under emergency braking does not exceed 0.47 m / s². -2 The third threshold is a constant speed control error of less than 0.1 km / h.

[0067] Different handle levels correspond to different speed control strategies, such as: when the handle level is positive 10, the traction corresponds to the maximum traction power; when the handle level is positive 5, the traction corresponds to the preset constant speed traction power; and when the handle level is negative 10, the traction corresponds to the maximum braking power.

[0068] This embodiment takes a block involving both locomotives and trailers as an example, and uses a Schmith predictor to compensate for the time delay characteristics of heavy-load trains. To address the difficulty of the Schmith predictor in handling the impact of system disturbances and parameter changes on train operation control (when the train speed change per unit time exceeds the first threshold, or the acceleration change under stepless speed regulation exceeds the second threshold), a speed prediction fuzzy PID-Smith time delay controller is designed by integrating the driver's operating experience and the identified vehicle operation model. This results in the system's speed tracking control having small overshoot (overshoot not exceeding 4% under square wave control signal), high tracking accuracy (constant speed control error less than 0.1 km / h), and fast response speed (steady-state response time not exceeding two sampling periods).

[0069] If the change in train speed of a heavily loaded train exceeds the first threshold per unit time, it may be due to a malfunction of the driver's controller.

[0070] This embodiment models the experience and knowledge accumulated by drivers over a long period of practice using a fuzzy rule base to establish an offline query fuzzy matrix table; during online operation, a fuzzy PID-Smith time-delay controller for speed prediction is designed using fuzzy inference. By processing the results of fuzzy logic rules, looking up tables, and performing calculations, the impact of uncertainties on speed tracking control during train operation is reduced.

[0071] This embodiment of a heavy-haul train control method based on vehicle dynamic response identification also includes adjusting the parameters of the fuzzy PID-Smith time-delay controller using a fuzzy rule base during the speed tracking control process under complex track conditions of heavy-haul trains.

[0072] In this embodiment, the fuzzy PID-Smith time-delay controller utilizes the fuzzy control principle to adjust the PID parameters online in real time, enabling the controlled object to exhibit good adaptability and control performance. The fuzzy PID-Smith time-delay controller is designed with the heavy-load train speed tracking error e and error change rate ec as inputs, ΔK... p ΔK i ΔK d The change in parameters is the output. A suitable fuzzy rule base should be established, and its fuzzy control rules should follow these principles:

[0073] ① When e and ec are small or equal to 0, K p The value of K is relatively large. i The value of K is relatively large. d The value is of medium magnitude. That is, when e is greater than or equal to 0 and less than the first threshold, and ec is greater than or equal to 0 and less than the second threshold, K... p K takes values ​​within the first range i K takes values ​​within the second range d The value is within the third range.

[0074] ② When e and ec are large, K p The value of K is relatively large. i The value is zero, K d The value is relatively small. That is, when e is greater than or equal to the third threshold and ec is greater than or equal to the fourth threshold, K... p K takes values ​​within the first range i The value is zero, K d The value is within the fourth range.

[0075] ③ When e and ec are of equal size, K p The value of K is relatively small. i The value of K is moderate. d The value is moderate. That is, when e is greater than or equal to the first threshold and less than the third threshold, and ec is greater than or equal to the second threshold and less than the fourth threshold, K... p K takes values ​​within the fifth range. i K takes values ​​within the range of the sixth. d The value is within the third range.

[0076] The minimum value in the first range is greater than the maximum value in the fifth range, the minimum value in the second range is greater than the maximum value in the sixth range, and the minimum value in the third range is greater than the maximum value in the fourth range.

[0077] K p K i K d These represent the proportional, integral, and derivative coefficients in a fuzzy PID-Smith controller, respectively. ΔK p ΔK i ΔK d K p K i K d The change in quantity.

[0078] For the problem of significant time delays (driver reaction time, control command transmission time, and vehicle response time) during the operation of heavy-haul trains, this invention uses an integrated approach of a first-order time delay model and a Schmis predictor to compensate for the train speed time delay characteristics under complex track conditions. The improved transfer function expression is as follows:

[0079]

[0080] Where G(s) represents the transfer function of the Smith predictor compensator, and S represents the transfer function operator.

[0081] Optionally, during the speed tracking control process under complex track conditions of heavy-load trains, an improved particle swarm optimization algorithm is used to dynamically adjust the time delay characteristic parameters of the fuzzy PID-Smith time delay controller.

[0082] Due to the limitations and blind spots of driver experience, as well as the excessive uncertainties in fuzzy control, the control performance of heavy-haul trains under complex operating conditions is poor. This embodiment considers the need for stable operation of heavy-haul trains under complex conditions. Using a fixed-parameter fuzzy PID-Smith time-delay controller would worsen train performance or even cause instability. Therefore, an improved Particle Swarm Optimization (PSO) algorithm is designed to optimize the parameters of the fuzzy PID-Smith time-delay controller for heavy-haul trains, improving the smoothness and robustness of operation under different conditions. More specifically, the improved PSO algorithm in this embodiment possesses excellent global optimization capabilities, enabling dynamic adjustment of the fuzzy PID-Smith controller parameters and optimization of the quantization factor and proportional factor based on control performance indicators. This achieves intelligent optimization of the fuzzy PID-Smith time-delay controller, further improving the dynamic performance of the control system.

[0083] In the speed tracking control process under complex track conditions of heavy-haul trains, when the speed tracking accuracy is greater than the third threshold, which is the constant speed control error of less than 0.1 km / h, an improved particle swarm algorithm is used to dynamically adjust the time delay characteristic parameters of the fuzzy PID-Smith controller.

[0084] This embodiment utilizes an improved PSO algorithm to dynamically adjust the parameters of the fuzzy PID-Smith time-delay controller, enabling it to intelligently optimize the control parameter K of the PID controller. p1 K i1 K d1 Initial values ​​are used to improve the dynamic performance of the control system. The absolute value of error time integral performance index is used as the fitness function of the PSO algorithm to optimize the system's control performance. The particle is considered as a solution vector in an N-dimensional space, let x... t and v t Let i represent the spatial position and velocity of particle i. Then, at time t, the particle's update strategy can be expressed as:

[0085]

[0086] Where, x t x represents the position of the particle at time t. t+1 V represents the position of the particle at time t+1. t v represents the velocity of the particle at time t. t+1 Let P represent the velocity of the particle at time t+1, w be the inertial weight, c1 and c2 be learning factors, r1 and r2 be random numbers between [0, 1], and P be the velocity of the particle at time t+1. t For the optimal position of a single particle at time t, G t Let t be the optimal position of the entire particle swarm at time t.

[0087] This invention addresses the optimization control problem of heavy-haul trains with large time delays. Based on the differences in the dynamic responses of different vehicles in a heavy-haul train, the vehicles are divided into multiple blocks. A vehicle operation model for each block is established using the traction or braking force of each block as input and the train speed as output. A Schmith predictor is used to compensate for the time delay characteristics of the speed under complex track conditions. When the speed change or acceleration change under stepless speed regulation exceeds a set range, a fuzzy PID-Smith time-delay controller is used to control the output speed of the vehicle operation model. When the speed tracking accuracy of the heavy-haul train exceeds a set range, an improved particle swarm optimization algorithm is used to dynamically adjust the time delay parameters of the fuzzy PID-Smith time-delay controller. This invention improves the parameter tuning accuracy and control reliability under the time delay characteristics of heavy-haul trains.

[0088] Example 2

[0089] This embodiment provides a heavy-load train control method based on vehicle dynamic response identification.

[0090] This embodiment focuses on the speed control of heavy-haul trains. Figure 2As shown, based on vehicle dynamic response identification technology, the vehicle control system can be divided into blocks involving only trailers and blocks involving both locomotives and trailers. The number of blocks is N, where N is a positive integer. Track data, Automatic Train Protection (ATP) data, and train operation experience are transmitted to the consortium blockchain via roadside units. Train blockchain data sharing is performed on these two types of blocks, dividing the vehicle control system into a data input end (ATP) and a data receiving end (train). Specifically applied to the longitudinal dynamics modeling of heavy-haul trains, based on the differences in dynamics (speed, acceleration, and coupler force) responses of different vehicles under complex track conditions in long-formation heavy-haul trains, the longitudinal dynamics model of the heavy-haul train is equivalent to a system with different coupled blocks. Vehicle operation models are established for each block, and through data sharing, coupler forces at each node and driver operation suggestions are derived. This embodiment proposes a fuzzy PID-Smith control method based on an improved PSO algorithm for the first block involving both locomotives and trailers. This method not only solves the accuracy problem of speed control for heavy-haul trains but also eliminates the adverse effects of model delay on the control system.

[0091] This implementation uses a single-input, single-output first-order autoregressive model to describe the block vehicle model. The autoregressive model is as follows:

[0092] A(z -1 )y(k)=B(z -1 u(k-1)+ξ(k) (6)

[0093] Where ξ(k) is the noise sequence, A(z) -1 B(z) -1 The expansion is as follows:

[0094]

[0095] Where, n a n b These are the orders of the input and output in formula (7), respectively. As analyzed above, the dynamic response model of the heavy-load train within the block can be equivalent to a first-order system, i.e., n... b =1, n a =0, then equation (6) can be described by a first-order autoregressive model with a single input and a single output:

[0096] y(k)=-a1y(k-1)+b0u(k-1)+ξ(k) (8)

[0097] in, θ is the data vector; θ = [a1, b0] is the parameter vector to be estimated.

[0098] By analyzing the data, such as Figure 3The prediction model shown in this embodiment lags behind the actual data by one unit sampling period, and the established mathematical model (vehicle motion model) is a first-order pure time-delay transfer function model.

[0099] To address the significant time delay issue during heavy-haul train operation, this implementation employs an integrated approach combining a first-order time delay model and a Schmis predictor to compensate for the train speed time delay characteristics under complex track conditions. For example... Figure 4 As shown, this embodiment provides a design method for a fuzzy PID-Smith controller based on dynamic response identification of heavy-haul trains:

[0100] Based on the train speed error e and the error change rate ec, the PID control parameter ΔK is adjusted in real time according to fuzzy rules. p ΔK i ΔK d To find the optimal solution, and then adjust the PID control parameter K p K i K d Adjust online. The adjustment strategy is shown in the following formula:

[0101]

[0102] Among them, K p1 K i1 K d1 These are the initial values ​​for the PID control parameters.

[0103] like Figure 5 As shown, this embodiment provides a method for optimizing the parameters of a fuzzy PID-Smith time-delay controller based on an improved PSO algorithm:

[0104] The improved PSO algorithm parameter setting process is as follows:

[0105] ① Initialization: Set the particle population size to SwarmSize = 100, dimension Dim = 3, inertia weights to random inertia weights, learning factors c1 = c2 = 2, and the maximum search space V. max =1, minimum value V min =-1, the maximum number of iterations is MaxIter=20, and the position and velocity of the particles are initialized at the same time.

[0106] ② To obtain the optimal solution, the absolute value of error time integral performance index is used as the fitness function of the PSO algorithm, with the following criteria:

[0107]

[0108] In the formula, F is the fitness value, t is time, and e(t) is the systematic error.

[0109] ③ Generate a particle swarm, and the particle swarm affects K. p1 K i1 K d1 The fitness values ​​of each particle are assigned according to formula (10), and then compared and selected to find the best individual and the best global.

[0110] ④ Update the position and velocity of the particles. If the termination condition is met, output the optimal solution; otherwise, perform a particle update operation to generate a new particle swarm for the next cycle.

[0111] Figure 5 In this context, r(t) represents the projected velocity, y(t) represents the actual velocity, and u(t) represents the traction force.

[0112] The PID control parameters K are obtained after optimization using an improved particle swarm optimization algorithm. p K i K d ,Depend on Figure 6 The optimization results of the fitness value show that the particle swarm can reach the optimal fitness value when it reaches 11 iterations.

[0113] To determine the first threshold for the change in speed of heavily loaded trains per unit time. Figure 7 The speed response effect under square wave control signal is presented. Compared with the traditional fuzzy PID control strategy, the overshoot in this embodiment is smaller (close to 2%), which meets the requirement that the overshoot does not exceed 4%.

[0114] To determine the second threshold for the change in acceleration under continuously variable speed (CVT) conditions. Figure 8 The acceleration response under emergency braking is presented. Comparative analysis of the figures shows that the acceleration change in this embodiment is 0.462 m / s² over two sampling periods. -2 The change in acceleration must not exceed 0.47 m / s². -2 Traditional PID controllers have become unstable, with the fuzzy PID controller providing an acceleration change of 0.56 m / s². -2 The threshold setting requirements are not met.

[0115] To determine the third threshold for velocity tracking accuracy, Figure 9 The effects of different control strategies on speed tracking error are presented. When using traditional fuzzy-Smith PID control, the constant speed control error is approximately 0.3 km / h, and the error is larger during sudden speed changes. When using the control strategy implemented in this paper, the constant speed control error does not exceed 0.1 km / h, and the error is smaller during sudden speed changes, which improves the stability and safety of heavy-haul trains to a certain extent.

[0116] from Figure 10In terms of speed tracking control performance, the fuzzy PID controller presented in this embodiment can eliminate the adverse effects of time delay characteristics and parameter changes in heavy-load trains, accurately reflect the actual speed change pattern, and has good robustness, meeting the speed control requirements of heavy-load trains (constant speed control error not exceeding 0.1 km / h).

[0117] Example 3

[0118] like Figure 11 As shown, this embodiment also provides a heavy-haul train control system based on vehicle dynamic response identification, the system comprising:

[0119] The block division module 201 is used to divide the carriages of a heavy-haul train into multiple blocks based on the differences in the dynamic response of different carriages.

[0120] The vehicle operation model establishment module 202 is used to establish the vehicle operation model of each block based on the common input and output of the mechanism characteristics and dynamic response of each block, with the traction force or braking force of each block as the input and the train speed as the output. That is, the first-order time delay model is used to equivalently describe the vehicle dynamic response.

[0121] The Schmis predictor feedback adjustment module 203 is used to compensate for the train speed time delay characteristics under complex track conditions of heavy-haul trains by integrating a first-order time delay model and a Schmis predictor, and to adjust the control signal of the corresponding block according to the compensated train speed. The complex track is an operation scenario in which the algebraic difference of the track slope length exceeds 30‰ within every 400 meters and the coupler force exceeds 2250KN, resulting in the safety hazard of coupler breakage. The control signal is traction force or braking force, and the output of the vehicle operation model is the train speed.

[0122] The fuzzy PID-Smith time-delay controller module 204 is used for speed tracking control under complex track conditions of heavy-load trains. When the speed change of the heavy-load train per unit time exceeds a first threshold, or the acceleration change under stepless speed regulation exceeds a second threshold, the fuzzy PID-Smith time-delay controller is designed based on the vehicle operation model to perform speed tracking control. The fuzzy PID-Smith time-delay controller is constructed based on different speed control strategies corresponding to different handle levels. The input of the fuzzy PID-Smith time-delay controller is traction force or braking force, and the output of the fuzzy PID-Smith time-delay controller is speed. During speed tracking control under complex track conditions of heavy-load trains, when the speed tracking accuracy exceeds a third threshold, an improved particle swarm optimization algorithm is used to dynamically adjust the parameters of the fuzzy PID-Smith time-delay controller. The first threshold is that the overshoot under square wave control signal does not exceed 4%; the second threshold is that the acceleration change within two sampling periods under emergency braking does not exceed 0.47 m / s². -2The third threshold is a constant speed control error of less than 0.1 km / h.

[0123] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0124] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A control method for heavy-haul trains based on vehicle dynamic response identification, characterized in that, include: Based on the differences in dynamic response of different carriages in a heavy-haul train, the carriages of the heavy-haul train are divided into multiple blocks; Based on the common input and output characteristics and dynamic responses of each block, a vehicle operation model for each block is established, using the traction or braking force of each block as input and the train speed as output. This model uses a first-order time-delay model to equivalently describe the vehicle's dynamic response. The vehicle operation model is expressed as follows: ； in, This represents the train speed at time k. This represents the train speed at time k-1. Let τ represent the train speed at time k-1-τ, and τ represent the time delay constant. This represents the traction force at time k-1. Represents the noise sequence at time k. Indicates the first parameter. b0 represents the second parameter, and b0 represents the third parameter; a recursive least squares algorithm with a forgetting factor is used to identify the first parameter, the second parameter, and the third parameter in the vehicle operation model; A first-order time-delay model and a Schmis predictor are integrated to compensate for the train speed time-delay characteristics under complex track conditions for heavy-haul trains. The control signals of the corresponding blocks are then adjusted based on the compensated train speed. The complex track is defined as an operational scenario where the algebraic difference in track gradient length exceeds 30‰ and the coupler force exceeds 2250KN, posing a safety hazard of coupler breakage. The control signal is either the traction force or the braking force, and the vehicle operation model output is the train speed. Under complex track conditions for heavy-haul trains, when the speed change of the heavy-haul train per unit time exceeds a first threshold, or the acceleration change under stepless speed regulation exceeds a second threshold, a fuzzy PID-Smith time-delay controller is designed based on the vehicle operation model for speed tracking control. The fuzzy PID-Smith time-delay controller is constructed based on different speed control strategies corresponding to different handle levels. The input to the fuzzy PID-Smith time-delay controller is the traction force or the braking force, and the output is the speed. During the speed tracking control process under complex track conditions for heavy-haul trains, when the speed tracking accuracy exceeds a third threshold, an improved particle swarm optimization algorithm is used to dynamically adjust the parameters of the fuzzy PID-Smith time-delay controller. The first threshold is an overshoot of no more than 4% under a square wave control signal; the second threshold is an acceleration change of no more than 0.47 m / s² within two sampling periods under emergency braking. -2 The third threshold is a constant speed control error of less than 0.1 km / h.

2. The heavy-haul train control method based on vehicle dynamic response identification according to claim 1, characterized in that, The particle update strategy of the improved particle swarm optimization algorithm is expressed as follows: ； in, This indicates the position of the particle at time t. This indicates the position of the particle at time t+1. This represents the velocity of the particle at time t. This represents the velocity of the particle at time t+1. For inertial weights, and All are learning factors. All are random numbers between [0, 1] Let be the optimal position of a single particle at time t. Let t be the optimal position of the entire particle swarm at time t.

3. The heavy-haul train control method based on vehicle dynamic response identification according to claim 1, characterized in that, The difference in dynamic response specifically refers to the difference in velocity response.

4. A heavy-haul train control system based on vehicle dynamic response identification, characterized in that, include: The block division module is used to divide the carriages of a heavy-haul train into multiple blocks based on the differences in the dynamic response of different carriages. The vehicle operation model establishment module is used to establish a vehicle operation model for each block based on the common input and output of the mechanistic characteristics and dynamic response of each block, using the traction force or braking force of each block as input and the train speed as output. That is, a first-order time-delay model is used to equivalently describe the vehicle's dynamic response; wherein, the vehicle operation model is expressed as: ； in, This represents the train speed at time k. This represents the train speed at time k-1. Let τ represent the train speed at time k-1-τ, and τ represent the time delay constant. This represents the traction force at time k-1. Represents the noise sequence at time k. Indicates the first parameter. b0 represents the second parameter, and b0 represents the third parameter; a recursive least squares algorithm with a forgetting factor is used to identify the first parameter, the second parameter, and the third parameter in the vehicle operation model; The Schmis predictor feedback adjustment module is used to compensate for the train speed time delay characteristics under complex track conditions of heavy-haul trains by integrating a first-order time delay model and a Schmis predictor, and to provide feedback adjustment to the control signals of the corresponding blocks based on the compensated train speed. The complex track is an operational scenario where the algebraic difference of the track slope length exceeds 30‰ within every 400 meters and the coupler force exceeds 2250KN, posing a safety hazard of coupler breakage. The control signal is the traction force or the braking force, and the vehicle operation model output is the train speed. A fuzzy PID-Smith time-delay controller module is used to perform speed tracking control under complex heavy-load train conditions. When the speed change of the heavy-load train per unit time exceeds a first threshold, or the acceleration change under stepless speed regulation exceeds a second threshold, a fuzzy PID-Smith time-delay controller is designed based on the vehicle operation model for speed tracking control. The fuzzy PID-Smith time-delay controller is constructed based on different speed control strategies corresponding to different handle levels. The input of the fuzzy PID-Smith time-delay controller is the traction force or the braking force, and the output is the speed. During speed tracking control under complex heavy-load train conditions, when the speed tracking accuracy exceeds a third threshold, an improved particle swarm optimization algorithm is used to dynamically adjust the parameters of the fuzzy PID-Smith time-delay controller. The first threshold is an overshoot of no more than 4% under a square wave control signal; the second threshold is an acceleration change of no more than 0.47 m / s² within two sampling periods under emergency braking. -2 The third threshold is a constant speed control error of less than 0.1 km / h.

Images