A vehicle lateral stability active control simulation method and system

Abstract

The present invention provides a vehicle lateral stability active control simulation method and system, designs a railway vehicle secondary lateral full active suspension layout, identifies the track excitation state and the vehicle vibration state, and designs an active control method that takes into account multiple vibration modes of the vehicle body. The present invention takes into account the suppression of multiple vibration modes of the vehicle body, can significantly improve the lateral stability of railway vehicles under broadband excitation, and provides a solution to the optimization design of high-speed EMU dynamic performance and the optimization of existing EMU operation and maintenance strategies.

Description

Technical Field

[0001] The invention belongs to the technical field of railway vehicle system dynamics and control, and in particular relates to a vehicle lateral stability active control simulation method and system for suppressing broadband vibration. Background Art

[0002] my country's high-speed railways are characterized by "high speed, wide area, and multiple lines". The long-term service dynamics of high-speed trains directly affects the operational safety, efficiency, and cost of high-speed railways. Due to the weak lateral constraints of the wheel and rail, the rich lateral unevenness of the track, and the human body's greater sensitivity to lateral medium and low-frequency vibrations, the problem of abnormal lateral vibration of vehicles in actual service is more prominent than that of vertical vibration. However, my country's current high-speed trains all use passive suspension, and their adaptability to changes in vehicle operating boundary conditions is limited. The dynamics performance can only be guaranteed by shortening the wheel turning and repair cycle, frequently grinding rails, regularly replacing shock absorbers, or reducing speed. This seriously affects the operating efficiency of high-speed railways and increases the cost and difficulty of vehicle and line operation and maintenance. In the future, my country's high-speed trains will be accelerated to more than 300km / h on a large scale and cross-line operation between multiple high-speed railway lines, and these problems will become more prominent. Due to certain differences in the construction and maintenance status of high-speed railway lines, the amplitude and wavelength of track unevenness fluctuate greatly, and the low-frequency and medium-high-frequency wheel-rail excitations of trains running on different lines or at different speeds are significantly different. However, low-frequency and mid-high frequency uneven excitations affect the running smoothness, and passive suspension is difficult to take into account wide-band vibration suppression.

[0003] Therefore, passive suspension cannot effectively solve the new problems caused by speed increase and cross-line operation. Semi-active suspension can adjust the damping of the shock absorber, but it is only effective in the low frequency band and has limited capabilities. The output force of the fully active suspension is controllable in real time, with strong control capabilities and a wide frequency band, which can solve the above bottlenecks. Compared with the first-order active suspension, the second-order lateral active suspension has the advantages of excellent performance, high safety and reliability, good technical maturity, etc., and has a very broad prospect for engineering application. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method and system for simulating active control of vehicle lateral stability for suppressing broadband vibrations. The present invention takes failure safety and redundant design into consideration and adopts an active-passive parallel structure, and takes into account the lateral displacement, rolling, head shaking and lateral bending modal vibrations of the vehicle body. It can achieve forced vibration suppression under broadband excitation of track unevenness, and provides an important reference for the optimization design of vehicle dynamics performance.

[0005] In order to achieve the above objectives, the technical solution adopted by the present invention is as follows.

[0006] A vehicle lateral stability active control simulation method comprises the following steps:

[0007] S1. Design the secondary lateral active suspension structure of the vehicle, and replace one of the two lateral passive shock absorbers in the secondary suspension with an active actuator;

[0008] S2. Establishing a dynamics and control simulation model of a vehicle active suspension system, including: establishing a three-dimensional nonlinear dynamics model of the vehicle based on SIMPACK, taking into account the wheel-rail relationship and the nonlinearity of suspension parameters; establishing an active suspension controller model based on MATLAB / Simulink; establishing a dynamics model of an active actuator based on MATLAB / Simulink; data exchange between SIMPACK and Simulink is achieved through a Simat interface; the dynamics and control simulation model of the vehicle active suspension system also includes a detection system, which detects vibration state feedback quantities at multiple measuring points on the floor surface of the vehicle body, i.e., vibration acceleration, and inputs the detected vibration state feedback quantities into the active suspension controller model;

[0009] S3, simulate the random excitation input of the track under any running speed condition, apply broadband forced vibration to the vehicle through the wheel-rail action, and transmit it to the vehicle body through the two-system suspension system;

[0010] S4, identifying the vehicle running speed, track excitation state and vehicle lateral stability state according to the current vibration state feedback, judging whether it is necessary to turn on the active suspension controller, and if so, performing the following steps, otherwise turning off the active suspension controller, and the active actuator works in the passive control mode;

[0011] S5. High-pass filtering and integrating the vibration state feedback of the vehicle body floor surface, i.e., the vibration acceleration, to obtain the vehicle body vibration velocity and displacement, and then identifying the vibration state quantity in the modal space through modal decomposition; or high-pass filtering the vibration acceleration of the vehicle body floor surface, identifying the vibration state quantity in the modal space, i.e., the modal vibration acceleration, through modal decomposition, and then integrating the modal vibration acceleration to obtain the modal vibration velocity and modal vibration displacement;

[0012] S6. Determine the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space by using a ceiling hang damping control method, a ceiling hang inertia container control method, a ceiling hang stiffness control method, or a linear combination control method and / or a cubic combination control method thereof;

[0013] S7. According to the ideal active control force, the active actuator actually outputs the active control force F 实际 Act on the vehicle.

[0014] Furthermore, the present invention also provides a vehicle lateral stability active control simulation system,

[0015] It includes a three-dimensional nonlinear dynamic model of the vehicle based on SIMPACK, in which one of the two lateral passive shock absorbers in the secondary suspension is replaced with an active actuator;

[0016] It also includes a detection system, which detects vibration state feedback quantities at multiple measurement points on the vehicle body floor surface, and inputs the detected vibration state feedback quantities into the active suspension controller model;

[0017] The invention also includes an active suspension controller model and an active actuator dynamics model established based on MATLAB / Simulink, wherein the active suspension controller model includes a high-pass filter, an integrator, a modal decomposition module and a skyhook controller, wherein the high-pass filter is used to perform high-pass filtering on the vibration state feedback quantity, the integrator is used to integrate the vibration state feedback quantity after high-pass filtering to obtain the vibration velocity and displacement of the vehicle body, the modal decomposition module is used to identify the vibration state quantity in the modal space, and the skyhook controller adopts a skyhook damping control method or a skyhook inertia container control method or a skyhook stiffness control method, or a linear combination control method and / or a cubic term combination control method of these methods to determine the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space, and inputs the ideal active control force instruction into the active actuator dynamics model; the active actuator in the three-dimensional nonlinear dynamics model of the vehicle actually outputs the active control force F 实际 Act on the vehicle;

[0018] The system also includes a train running state identification module, which identifies the vehicle running speed, track excitation state and vehicle lateral stability state according to the current vibration state feedback, and judges whether it is necessary to turn on the active suspension controller. If not, the active suspension controller is turned off, and the active actuator works in a passive control mode; if necessary, the active suspension controller is turned on, and the active suspension controller performs high-pass filtering, integration and modal decomposition on the vibration state feedback of the vehicle body floor surface to obtain the vibration state quantity in the modal space, and the ceiling controller determines the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space; the active actuator follows the ideal active control force output by the active suspension controller, and actually outputs the active control force F 实际 Act on the vehicle;

[0019] Data exchange between SIMPACK and MATLAB / Simulink is achieved through the Simat interface.

[0020] Beneficial effects of the present invention:

[0021] The present invention provides a method and system for actively controlling the lateral stability of a vehicle, designs a second-system lateral fully active suspension arrangement for a railway vehicle, identifies the track excitation state and the vehicle vibration state, and designs an active control method that takes into account multiple vibration modes of the vehicle body. The present invention takes into account the suppression of multiple vibration modes of the vehicle body, can significantly improve the lateral stability of railway vehicles under broadband excitation, and provides a solution for the optimization design of the dynamic performance of high-speed EMUs and the optimization of the operation and maintenance strategies of existing EMUs. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a layout diagram of the secondary transverse active suspension structure for railway vehicles.

[0023] Figure 2 It is a simulation model for the dynamics and control of the active suspension system of railway vehicles.

[0024] Figure 3 This is the working principle diagram of the secondary lateral active suspension of railway vehicles.

[0025] Figure 4 Schematic diagram of acceleration measurement points for vehicle body stability indicators. DETAILED DESCRIPTION

[0026] A vehicle lateral stability active control simulation method comprises the following steps:

[0027] S1. Design the secondary lateral active suspension structure of the vehicle, and replace one of the two lateral passive shock absorbers in the secondary suspension with an active actuator;

[0028] S2. Establishing a dynamics and control simulation model of a vehicle active suspension system, including: establishing a three-dimensional nonlinear dynamics model of the vehicle based on SIMPACK, taking into account the wheel-rail relationship and the nonlinearity of suspension parameters; establishing an active suspension controller model based on MATLAB / Simulink; establishing a dynamics model of an active actuator based on MATLAB / Simulink; data exchange between SIMPACK and Simulink is achieved through a Simat interface; the dynamics and control simulation model of the vehicle active suspension system also includes a detection system, which detects vibration state feedback quantities at multiple measuring points on the floor surface of the vehicle body, i.e., vibration acceleration, and inputs the detected vibration state feedback quantities into the active suspension controller model;

[0029] S3, simulate the random excitation input of the track under any running speed condition, apply broadband forced vibration to the vehicle through the wheel-rail action, and transmit it to the vehicle body through the two-system suspension system;

[0030] S4, identifying the vehicle running speed, track excitation state and vehicle lateral stability state according to the current vibration state feedback, judging whether it is necessary to turn on the active suspension controller, and if so, performing the following steps, otherwise turning off the active suspension controller, and the active actuator works in the passive control mode;

[0031] S5. High-pass filtering and integrating the vibration state feedback of the vehicle body floor surface to obtain the vehicle body vibration velocity and displacement, and then identifying the vibration state quantity in the modal space, i.e., the modal vibration acceleration, the modal vibration velocity, and the modal vibration displacement, through modal decomposition; or high-pass filtering the vibration state feedback of the vehicle body floor surface, i.e., the vibration acceleration, identifying the vibration state quantity in the modal space, i.e., the modal vibration acceleration, through modal decomposition, and then integrating the modal vibration acceleration to obtain the modal vibration velocity and the modal vibration displacement;

[0032] S6. Determine the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space by using a ceiling hang damping control method, a ceiling hang inertia container control method, a ceiling hang stiffness control method, or a linear combination control method and / or a cubic combination control method thereof;

[0033] S7. According to the ideal active control force, the active actuator actually outputs the active control force F 实际 Act on the vehicle.

[0034] Furthermore, the secondary lateral active suspension structure in step S1 is as follows: Figure 1 As shown: In order to achieve fail-safe and redundant design, an active and passive parallel structure is adopted. Two lateral passive shock absorbers are generally configured in the secondary suspension of the vehicle, and only one of them is replaced by the active actuator. The active actuator adopts an integrated electro-hydraulic actuator (EHA), which is mainly composed of a brushless motor, a hydraulic pump, an accumulator, an actuator cylinder and a servo controller. It has the technical characteristics of high system integration, small installation space and integration of active and passive control modes.

[0035] Furthermore, in step S2, a vehicle active suspension system dynamics and control simulation model is established, such as Figure 2As shown. A three-dimensional nonlinear dynamic model of the vehicle is established based on SIMPACK, taking into account the wheel-rail relationship and the nonlinearity of the suspension parameters. The detection system detects the vibration state feedback at multiple measuring points on the vehicle floor, that is, the vibration acceleration of the vehicle floor, and inputs the detected vibration state feedback into the active suspension controller model. An active suspension controller model is established based on MATLAB / Simulink, the vibration state quantity in the modal space is identified, and the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space is determined. A dynamic model of the active actuator is established based on MATLAB / Simulink, taking into account the dynamic response characteristics of the active actuator. Data exchange between SIMPACK and Simulink is achieved through the Simat interface, including vibration state feedback and active control force, to achieve joint simulation modeling of the dynamics and control of the vehicle active suspension system.

[0036] like Figure 3 As shown, the vehicle is simulated to be subjected to random track excitation input under any operating speed condition, and broadband forced vibration is applied to the vehicle through the wheel-rail action, and is transmitted to the vehicle body through the two-system suspension system. First, the train running state identification module identifies the vehicle running speed, track excitation state and running stability state according to the vibration state feedback, and judges whether the active suspension controller needs to be turned on. If necessary, the following steps are performed, otherwise the active suspension controller is turned off, and the active actuator works in the passive control mode. When the train running state identification module judges that the active suspension controller needs to be turned on, the active suspension controller performs high-pass filtering and integration on the vibration acceleration at the measuring point of the vehicle body floor surface to obtain the vehicle body vibration velocity and displacement, and then identifies the vibration state quantity in the modal space through modal decomposition, namely, the modal vibration displacement, modal vibration velocity and modal vibration acceleration; or the active suspension controller performs high-pass filtering on the vibration acceleration of the vehicle body floor surface, identifies the vibration state quantity in the modal space, namely, the modal vibration acceleration, through modal decomposition, and then integrates the modal vibration acceleration to obtain the modal vibration velocity and modal vibration displacement. Then, based on the skyhook control principle, the mapping relationship between the ideal active control force and the vibration state quantity in the modal space is designed. The mapping relationship can be linear or nonlinear. Secondly, taking the lateral stability of the vehicle as the control target, the influence of the vehicle body lateral displacement, head shaking, rolling and lateral bending vibration modes on the lateral stability index is analyzed, and the vehicle body vibration mode vibration type that has the main influence on lateral stability is determined. The control gain with a relatively large influence on lateral stability is optimized using genetic algorithms or univariate analysis methods to determine the optimal range of the control gain. Finally, the active actuator tries to follow the ideal active control force command of the active suspension controller, but there may be phase delays, such as time lags, and the actual output active control force F 实际 Act on the vehicle.

[0037] Furthermore, the principle of identifying the vibration state quantity in the modal space in step S5 is as follows. Figure 4 , a1 and a2 are the lateral stability measurement points of locomotive vehicles specified in the GB / T 5599-2019 standard, that is, the vibration acceleration at the left and right distance w (e.g. 1m) from the center line of the car body floor surface above the front and rear bogies, including lateral and vertical accelerations. According to the theory of multi-body system dynamics, the lateral and vertical accelerations of the front right measurement point a1, the rear left measurement point a2, and the rear right measurement point a3 of the floor can be expressed as the superposition of the lateral displacement, head shaking, side rolling, floating, nodding and lateral bending modal vibrations of the car body, that is:

[0038]

[0039] Where: a 1y and a 1z Indicates the lateral and vertical acceleration of the measuring point a1; a 2y and a 2z Indicates the lateral and vertical acceleration of the measuring point a2; a 3y and a 3z Indicates the lateral and vertical accelerations of the measuring point a3; are the vibration accelerations of the vehicle body in the following modes: lateral displacement, head shaking, rolling, floating, nodding and lateral bending; Y 1y (x i ) is the vibration mode function of the vehicle body transverse bending mode. For example, the first-order vibration mode function of the Euler Bernoulli beam with free ends is Where β1 = 3π / 2L, where x i Indicates measuring point a i The longitudinal distance from the rear end of the car body, i = 1, 2, 3; w represents the lateral distance between the measuring points a1, a2 and a3 and the center line of the car body floor surface, for example, the value is 1m; l1 is the longitudinal distance between the measuring point a3 and the center of the rear bogie, and it is required that 0<l1<2L c ; x1 and x2 are the longitudinal distances from the measuring points a1 and a2 to the rear end of the vehicle body, respectively, and satisfy x1=L / 2+L c and x2=L / 2-L c , x3 is the longitudinal distance from the measuring point a3 to the rear end of the vehicle body, and x3=x2+l1; L is the length of the vehicle body; L c It represents half of the vehicle spacing (the longitudinal distance between the centers of the front and rear bogies); h1 represents the vertical height of the center of mass of the vehicle body from the floor surface.

[0040] According to the last three equations of the above formula, the amplitudes of the vehicle body's floating, nodding and rolling modes can be solved, and then substituted into the first three equations to solve the vibration acceleration of the vehicle body's shaking, lateral movement and lateral bending modes:

[0041]

[0042] The vibration acceleration of the vehicle floor is subjected to high-pass filtering and integration to obtain the vibration velocity and displacement of the vehicle body, and then the vibration state quantity in the modal space, namely the modal vibration velocity and modal vibration displacement, is identified through modal decomposition. Alternatively, the vibration acceleration of the vehicle floor is subjected to high-pass filtering, the modal vibration acceleration is identified through modal decomposition, and then the modal vibration acceleration is integrated to obtain the modal vibration velocity and modal vibration displacement.

[0043] Furthermore, the ideal active control force of the front and rear ends of the vehicle in step S6 is determined by the following method. The vibration state quantities in the roll, head shake, lateral displacement and lateral bending modal space of the vehicle body are known, and the mapping relationship between the ideal active control force and the vibration state quantity in the modal space is determined by using the ceiling damping control method, ceiling inertia control method or ceiling stiffness control method, or the linear combination control and / or cubic term combination control method of these methods. Secondly, taking the lateral stability of the vehicle as the control target, the influence of the lateral displacement, head shake, roll and lateral bending vibration modes of the vehicle body on the lateral stability index is analyzed, and the vibration modal vibration shape of the vehicle body that has a major influence on the lateral stability is determined. Furthermore, the control gain that has a relatively large influence on the lateral stability is optimized by using a genetic algorithm or a univariate analysis method to determine the optimal range of the control gain.

[0044] Among them, the ceiling damping control method is a linear control based on modal vibration velocity feedback. The vibration acceleration at the measuring point of the vehicle floor is high-pass filtered and integrated to obtain the vehicle body vibration velocity, and then the vibration state quantity in the modal space, that is, the modal vibration velocity, is identified through modal decomposition; or the vibration state feedback quantity of the vehicle floor, that is, the vibration acceleration, is high-pass filtered, and the vibration state quantity in the modal space, that is, the modal vibration acceleration, is identified through modal decomposition, and then the modal vibration acceleration is integrated to obtain the modal vibration velocity. Determine the ideal lateral active control force F of the second system of the whole vehicle 理想1 The general expression is:

[0045]

[0046] In the formula, is the vibration state quantity in the modal space, including the lateral modal vibration velocity Shaking head mode vibration speed Rolling mode vibration velocity and transverse-bending mode vibration velocity A is the control gain coefficient matrix, A 11 etc. are control gain coefficients, and the optimal range of control gain coefficients is determined by genetic algorithm or univariate analysis; F 理想1 The two element values ​​are the ideal active control forces of the two actuators.

[0047] The ceiling inertia control method is a linear control based on modal vibration acceleration feedback. The vibration acceleration at the measuring point on the vehicle floor is high-pass filtered, and then the vibration state quantity in the modal space, i.e., the modal vibration acceleration, is identified through modal decomposition. Determine the ideal lateral active control force F of the whole vehicle 理想2 The general expression is:

[0048]

[0049] In the formula, is the vibration state quantity in the modal space, including the lateral modal vibration acceleration Shaking modal vibration acceleration Rolling mode vibration acceleration and transverse bending mode vibration acceleration B is the control gain coefficient matrix, B 11 etc. are control gain coefficients, and the optimal range of control gain coefficients is determined by genetic algorithm or univariate analysis; F 理想2 The two element values ​​are the ideal active control forces of the two actuators.

[0050] The ceiling stiffness control method is a linear control based on modal vibration displacement feedback. The vibration acceleration at the measuring point of the vehicle floor is high-pass filtered and integrated twice to obtain the vehicle body vibration displacement, and then the vibration state quantity in the modal space, that is, the modal vibration displacement, is identified through modal decomposition; or the vibration state feedback quantity of the vehicle floor, that is, the vibration acceleration, is high-pass filtered, and the vibration state quantity in the modal space, that is, the modal vibration acceleration, is identified through modal decomposition, and then the modal vibration acceleration is integrated twice to obtain the modal vibration displacement. Determine the ideal lateral active control force F of the whole vehicle 理想3 The general expression is:

[0051]

[0052] Where u is the vibration state quantity in the modal space, including the lateral modal vibration displacement y c 、Shaking modal vibration displacement ψ c , Rolling mode vibration displacement φ c and transverse bending mode vibration displacement q 1y ; C is the control gain coefficient matrix, C 11 etc. are control gain coefficients, and the optimal range of control gain coefficients is determined by genetic algorithm or univariate analysis; F 理想3 The two element values ​​are the ideal active control forces of the two actuators.

[0053] Furthermore, the linear combination control method is a linear combination of the ceiling hang damping control method, the ceiling hang inertia container control method and the ceiling hang stiffness control method, that is,

[0054]

[0055] In the formula, A 11 etc. are control gain coefficients, and the optimal range of all control gain coefficients is determined by genetic algorithm or univariate analysis; F 理想4 The two element values ​​are the ideal active control forces of the two actuators.

[0056] Similarly, the cubic combination control method is similar to the linear combination control method, except that the state feedback is a cubic term, that is,

[0057]

[0058] In the formula, A 11 ' etc. are control gain coefficients, and the optimal range of all control gain coefficients is determined by genetic algorithm or univariate analysis; F 理想5 The two element values ​​are the ideal active control forces of the two actuators.

[0059] Furthermore, the linear combination control and cubic combination control can be further combined to obtain a generalized combination control method, that is, the ideal active control force F 理想6 It is a combination of linear combination control and cubic combination control, and the general expression of the ideal lateral active control force of the whole vehicle is determined as F 理想6 =F 理想4 +F 理想5 .

Claims

1. A vehicle lateral stability active control simulation method, comprising the following steps: S1. Design the secondary lateral active suspension structure of the vehicle, and replace one of the two lateral passive shock absorbers in the secondary suspension with an active actuator; S2. Establishing a dynamics and control simulation model of a vehicle active suspension system, including: establishing a three-dimensional nonlinear dynamics model of the vehicle based on SIMPACK, taking into account the wheel-rail relationship and the nonlinearity of suspension parameters; establishing an active suspension controller model based on MATLAB / Simulink; establishing a dynamics model of an active actuator based on MATLAB / Simulink; data exchange between SIMPACK and MATLAB / Simulink is achieved through a Simat interface; the dynamics and control simulation model of the vehicle active suspension system also includes a detection system, which detects vibration state feedback quantities at multiple measuring points on the floor surface of the vehicle body, and inputs the detected vibration state feedback quantities into the active suspension controller model; S3, simulate the random excitation input of the track under any running speed condition, apply broadband forced vibration to the vehicle through the wheel-rail action, and transmit it to the vehicle body through the two-system suspension system; S4, identifying the vehicle running speed, track excitation state and vehicle lateral stability state according to the current vibration state feedback, judging whether it is necessary to turn on the active suspension controller, and if so, performing the following steps, otherwise turning off the active suspension controller, and the active actuator works in the passive control mode; S5. High-pass filtering, integration and modal decomposition are performed on the vibration state feedback of the vehicle body floor surface to identify the vibration state quantity in the modal space, where the vibration state quantity in the modal space includes modal vibration acceleration, modal vibration velocity and modal vibration displacement; S6. Determine the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space by using a ceiling hang damping control method, a ceiling hang inertia container control method, or a ceiling hang stiffness control method, or a linear combination control method and / or a cubic combination control method of these methods; S7. According to the ideal active control force, the active actuator actually outputs the active control force F 实际 Act on the vehicle.

2. The vehicle lateral stability active control simulation method according to claim 1, characterized in that: The active actuator adopts an integrated electric hydrostatic actuator with both active and passive control modes.

3. The vehicle lateral stability active control simulation method according to claim 1, characterized in that: The vibration state feedback quantity is the vibration acceleration at the front right measuring point a1, the rear left measuring point a2, and the rear right measuring point a3 located at a lateral distance w from the center line of the vehicle body floor surface, including lateral and vertical accelerations.

4. The vehicle lateral stability active control simulation method according to claim 3 is characterized by: In step S5, the modal vibration acceleration is identified using the following formula: In the formula, a 1y and a 1z Indicates the lateral and vertical acceleration of the measuring point a1; a 2y and a 2z Indicates the lateral and vertical acceleration of the measuring point a2; a 3y and a 3z Indicates the lateral and vertical accelerations of the measuring point a3; are the vibration accelerations of the vehicle body in lateral displacement, sway, roll and lateral bending modes; Y 1y (x i ) is the vibration mode function of the vehicle body lateral bending mode, Where β1 = 3π / 2L, where x i Indicates measuring point a i The longitudinal distance from the rear end of the car body, i = 1, 2, 3; w represents the lateral distance between the measuring points a1, a2 and a3 and the center line of the car body floor surface; l1 is the longitudinal distance between the measuring point a3 and the center of the bogie near it, and 0 < l1 < 2L is required c ; x1 and x2 are the longitudinal distances from the measuring points a1 and a2 to the rear end of the vehicle body, respectively, and satisfy x1=L / 2+L c and x2=L / 2-L c ; x3 is the longitudinal distance from the measuring point a3 to the rear end of the vehicle body, and satisfies x3=x2+l1; L is the length of the vehicle body; L c It represents half of the vehicle's fixed distance; h1 represents the vertical height from the center of mass of the vehicle to the floor surface; The vibration velocity and displacement of the vehicle body are obtained by high-pass filtering and integrating the vibration acceleration at the measuring point of the vehicle body floor surface, and then the modal vibration velocity and modal vibration displacement are obtained through modal decomposition; or the vibration acceleration of the vehicle body floor surface is high-pass filtered, and the vibration state quantity in the modal space, namely the modal vibration acceleration, is identified through modal decomposition, and then the modal vibration acceleration is integrated to obtain the modal vibration velocity and modal vibration displacement.

5. The vehicle lateral stability active control simulation method according to claim 4, characterized in that: The step S6 of using the skyhook damping control method to determine the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space specifically includes: determining the ideal active control force F of the second system lateral direction of the whole vehicle 理想1 The general expression is: In the formula, is the vibration state quantity in the modal space, including the lateral modal vibration velocity Shaking head mode vibration speed Rolling mode vibration velocity and transverse-bending mode vibration velocity A is the control gain coefficient matrix, A 11 -A 24 is the control gain coefficient, and the vehicle lateral stability is taken as the control target. The optimal range of all control gain coefficients is determined by genetic algorithm or univariate analysis method; F 理想1 The two element values ​​are the ideal active control forces of the two actuators.

6. The vehicle lateral stability active control simulation method according to claim 4, characterized in that: The step S6 in which the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space is determined by using the ceiling inertia container control method specifically includes: determining the ideal active control force F of the second system of the whole vehicle in the lateral direction 理想2 The general expression is: In the formula, is the vibration state quantity in the modal space, including the lateral modal vibration acceleration Shaking modal vibration acceleration Rolling mode vibration acceleration and transverse bending mode vibration acceleration B is the control gain coefficient matrix, B 11 -B 24 is the control gain coefficient, and the vehicle lateral stability is taken as the control target. The optimal range of all control gain coefficients is determined by genetic algorithm or univariate analysis method; F 理想2 The two element values ​​are the ideal active control forces of the two actuators.

7. The vehicle lateral stability active control simulation method according to claim 4, characterized in that: The step S6 in which the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space is determined by using the skyhook stiffness control method specifically includes: determining the ideal active control force F of the second system lateral direction of the whole vehicle 理想3 The general expression is: Where u is the vibration state quantity in the modal space, including the lateral modal vibration displacement y c 、Shaking modal vibration displacement ψ c , Rolling mode vibration displacement φ c and transverse bending mode vibration displacement q 1y ; C is the control gain coefficient matrix, C 11 -C 24 is the control gain coefficient, and the vehicle lateral stability is taken as the control target. The optimal range of all control gain coefficients is determined by genetic algorithm or univariate analysis method; F 理想3 The two element values ​​are the ideal active control forces of the two actuators.

8. The vehicle lateral stability active control simulation method according to claim 4, characterized in that: The optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space determined by the linear combination control method in step S6 specifically includes: determining the ideal active control force F of the second system lateral direction of the whole vehicle 理想4 The general expression is: The step S6 in which the cubic term combination control method is used to determine the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space specifically includes: determining the ideal lateral active control force F of the whole vehicle 理想5 The general expression is: In step S6, linear combination control and cubic combination control are used to determine the ideal lateral active control force F of the entire vehicle. 理想6 The general expression is: F 理想6 =F 理想4 +F 理想5 ; In the formula, A, B, C, A', B', C' are control gain coefficient matrices, A 11 -A 24 , B 11 -B 24 , C 11 -C 24 , A 11 '-A 24 ', B 11 '-B 24 '、C 11 '-C 24 ' is the control gain coefficient, and the vehicle lateral stability is taken as the control target. The optimal range of all control gain coefficients is determined by genetic algorithm or univariate analysis method; They are the vibration velocity of the lateral displacement mode, the vibration velocity of the swaying mode, the vibration velocity of the rolling mode, and the vibration velocity of the lateral bending mode, y c , c ,φ c ,q 1y They are the lateral displacement, the swaying vibration displacement, the rolling vibration displacement, and the lateral bending vibration displacement; F 理想1 、F 理想2 、F 理想3 、F 理想4 、F 理想5 、F 理想6 The two element values ​​are the ideal active control forces of the two actuators.

9. A vehicle lateral stability active control simulation system, characterized in that: It includes a three-dimensional nonlinear dynamic model of the vehicle based on SIMPACK, in which one of the two lateral passive shock absorbers in the secondary suspension is replaced with an active actuator; It also includes a detection system, which detects vibration state feedback quantities at multiple measurement points on the vehicle body floor surface, and inputs the detected vibration state feedback quantities into the active suspension controller model; It also includes an active suspension controller model and an active actuator dynamics model established based on MATLAB / Simulink, wherein the active suspension controller model includes a high-pass filter, an integrator, a modal decomposition module and a skyhook controller, the high-pass filter is used to perform high-pass filtering on the vibration state feedback quantity, the modal decomposition module is used to identify the vibration state quantity in the modal space, the integrator is used to integrate the vibration state feedback quantity after high-pass filtering or the vibration state quantity in the modal space, the skyhook controller adopts a skyhook damping control method or a skyhook inertia container control method or a skyhook stiffness control method, or a linear combination control method and / or a cubic term combination control method of these methods to determine the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space, and inputs the ideal active control force instruction into the active actuator dynamics model; The actual output active control force F of the active actuator in the three-dimensional nonlinear dynamic model of the vehicle 实际 Act on the vehicle; The system also includes a train running state identification module, which identifies the vehicle running speed, track excitation state and vehicle lateral stability state according to the current vibration state feedback, and judges whether it is necessary to turn on the active suspension controller. If not, the active suspension controller is turned off, and the active actuator works in a passive control mode; if necessary, the active suspension controller is turned on, and the active suspension controller performs high-pass filtering, integration and modal decomposition on the vibration state feedback of the vehicle body floor surface to obtain the vibration state quantity in the modal space, and the ceiling controller determines the optimal mapping relationship between the ideal active control force and the vibration state quantity in the modal space; the active actuator follows the ideal active control force output by the active suspension controller, and actually outputs the active control force F 实际 Act on the vehicle; Data exchange between SIMPACK and MATLAB / Simulink is achieved through the Simat interface.

Images