Semi-active suspension control strategy based on IIR filter frequency division

Abstract

The invention discloses a semi-active suspension control strategy based on IIR (Infinite Impulse Response) filter frequency division, which is characterized in that an acceleration sensor is used as input, a low-pass IIR filter and a high-pass IIR filter are respectively used for processing the acceleration of a cab, and appropriate damping is output according to a result so as to suppress the resonance phenomenon of a vehicle body during low-frequency vibration; and the vibration isolation coefficient during high-frequency vibration is improved. According to the method, the number of required sensors is small, the complexity is low, the computing power requirement for hardware is small, the cost of the semi-active suspension can be effectively reduced, and the method has a certain engineering practical application value and is high in generalizability. Test results show that compared with passive suspension, the control strategy can effectively reduce vibration of a vehicle cab in the x direction and the z direction, so that the comfort index of the long-distance logistics vehicle is improved, x-direction optimization is 21.3% at most, and z-direction optimization is 11.7% at most.

Description

Technical Field

[0001] The invention relates to the field of vehicle semi-active suspension control, in particular to a semi-active suspension control strategy based on IIR filter frequency division. Background Art

[0002] With the further development of new energy vehicles, the research on intelligent chassis has attracted the attention of more and more car companies. People have higher and higher requirements for comfort during vehicle driving. Suspension, as an important part of intelligent chassis, has become one of the research hotspots. According to the control method, suspension can be divided into passive suspension, active suspension and semi-active suspension. The stiffness and damping characteristics of passive suspension cannot be adjusted in real time, and it is difficult to adapt to the comfort requirements of different driving conditions. Although active suspension has good comfort, it is difficult to achieve large-scale promotion and application in low-end and mid-end models due to its complex structure, high energy consumption and high cost. Semi-active suspension first appeared in 1973. As an alternative to the high-cost active suspension system, semi-active suspension is close to active suspension in terms of comfort, but has the advantages of relatively simple structure, low energy consumption, low cost and high reliability. At the same time, the rapid development of new energy vehicles in China has provided a wide range of application scenarios for the application of semi-active suspension. This also makes the intelligent suspension system with semi-active suspension as the core a current research and industrial application hotspot.

[0003] At present, many algorithms use vehicle status information such as absolute speed or relative speed. In actual vehicles, sensors that directly detect vehicle vibration speed are rarely equipped. In semi-active suspension systems, there are generally two solutions to obtain vibration speed. One is to use a vehicle height sensor to obtain the relative displacement of the suspension, and then differentiate the relative displacement to obtain the relative speed. The advantage of this solution is high accuracy, but it cannot obtain absolute speed. At the same time, the vehicle height sensor has high cost and complexity, and has high requirements for installation location and space. Another solution is to use an acceleration sensor to obtain vibration acceleration, and then numerically integrate the acceleration to obtain the speed. This solution has low hardware requirements, but there will be low-frequency trend items in the numerical integration. For semi-active suspension systems, the frequency range of low-frequency trend items is close to the frequency range of vibration, and the commonly used detrending term algorithm has poor effect.

[0004] Therefore, it is necessary to develop a semi-active suspension control strategy that does not require absolute or relative body vibration velocity variables to overcome the above problems, improve driving comfort, and avoid the instability caused by the introduction of large error speed signals. In general, this control strategy can significantly reduce the application cost of semi-active suspension in vehicle systems while improving driving comfort. Summary of the invention

[0005] The purpose of the present invention is to provide a semi-active suspension control strategy based on IIR filter frequency division, which improves driving comfort while improving the problem of deterioration of control effect after introducing a vibration velocity signal with large error without large calculation amount.

[0006] To implement the above method, the present invention adopts the following technical solution: a semi-active suspension control strategy based on IIR filter frequency division, the steps are as follows:

[0007] Step S1, establish a vehicle dynamics model and a random road model, preliminarily calculate the resonant frequency of the vehicle dynamics system, and then proceed to step S2.

[0008] Step S2, respectively design a low-pass IIR filter and a high-pass IIR filter, combine the resonant frequency ω0 of the vehicle dynamics system, preliminarily determine the orders and cutoff frequency points of the two IIR filters, design a filter frequency division controller, preliminarily determine the sampling time length and the size of the high and low frequency judgment thresholds, and then proceed to step S3.

[0009] Step S3, taking the comfort index of the sprung mass as the optimization target, using the fmincon optimization algorithm to determine the optimal solutions for various parameters in the low-pass IIR filter, the high-pass IIR filter and the filter frequency division controller, so as to optimize the comprehensive performance of the semi-active suspension.

[0010] Compared with the prior art, the present invention has the following beneficial effects:

[0011] (1) A semi-active suspension control strategy based on IIR filter frequency division has better driving comfort compared with passive suspension.

[0012] (2) This control strategy requires a small number of sensors, has low complexity, and requires little computing power for the hardware. It can effectively reduce the cost of the semi-active suspension, has certain practical engineering application value, and is highly scalable.

[0013] (3) The control strategy has good effect and can effectively reduce the vibration of the vehicle cab in the x-direction and z-direction, with the highest optimization of 21.3% in the x-direction and 11.7% in the z-direction. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 The present invention discloses a flow chart of a semi-active suspension control strategy based on IIR filter frequency division.

[0015] Figure 2 It is a framework diagram of a semi-active suspension control strategy based on IIR filter frequency division proposed by the present invention.

[0016] Figure 31 is a comparison diagram of the sprung mass acceleration and passive suspension effect of the present invention, wherein (a) is a comparison diagram of the frequency domain diagram at 0 to 5 Hz, and (b) is a comparison diagram of the frequency domain diagram at 9 to 14 Hz. DETAILED DESCRIPTION

[0017] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] Semi-active control strategies are closely related to ride comfort and maneuverability. Currently, many algorithms use vehicle status information such as absolute speed or relative speed. However, in actual vehicles, there are generally few sensors that can directly detect the vehicle vibration speed. The speed information obtained by integration has a large error, making it difficult to apply in actual vehicles.

[0019] Aiming at the phenomenon that the vehicle body vibration speed cannot be obtained quickly and accurately in actual vehicles, which leads to the deterioration of the effect of most algorithms, this paper proposes a semi-active suspension control strategy based on IIR filter frequency division. The IIR filter can be used to judge the high and low frequency components of the vehicle body acceleration, and output the appropriate damping size according to the result, so as to suppress the resonance of the vehicle body during low frequency vibration and improve the vibration isolation coefficient during high frequency vibration. Figure 1 As shown, the following steps are included:

[0020] Step S1: Establish a vehicle dynamics model and a random road surface model, and preliminarily calculate the resonant frequency of the vehicle dynamics system, as follows:

[0021] The vehicle dynamics model is a 12-degree-of-freedom commercial vehicle model:

[0022] Refer to the actual vehicle dynamics parameters, use vehicle dynamics principles and physical laws to simulate the dynamic behavior of the entire vehicle system.

[0023] Establish a random road surface model: select the white noise time domain road surface as the input of the vehicle dynamics model, combine it with the vehicle dynamics model that can simulate the real situation, and simulate and calculate the resonant frequency of the vehicle dynamics system.

[0024] The vehicle dynamics model includes three vibration differential equations and a kinematic relationship.

[0025] The three vibration differential equations include the vertical vibration differential equations of the six tire parts, the three-degree-of-freedom vibration differential equations of the frame part, and the three-degree-of-freedom vibration differential equations of the cab part, as follows:

[0026] Differential equations for vertical vibration of the six tire sections:

[0027]

[0028] In the formula, m wi is the mass of the i-th tire; z wi is the vertical displacement of the center of the ith tire; is the vertical velocity of the center of the ith tire; is the vertical acceleration of the center of the ith tire; k wi is the stiffness of the ith tire; q i is the input displacement at the ith tire; k si is the stiffness of the i-th tire suspension; z fi is the displacement of the frame suspension spring corresponding to the i-th tire; is the frame suspension spring speed corresponding to the i-th tire; is the acceleration of the frame suspension spring corresponding to the i-th tire; c si is the frame suspension damping coefficient corresponding to the i-th tire; where i = 1, 2, …, 6.

[0029] The three-degree-of-freedom vibration differential equation of the frame:

[0030]

[0031] In the formula, m f is the frame mass; z f is the vertical displacement of the frame; is the vertical acceleration of the frame; z bj is the unsprung displacement corresponding to the jth cab suspension; z cj is the sprung displacement corresponding to the jth cab suspension; f j is the controllable damping force of the jth cab suspension; I fxx ,I fyy is the moment of inertia of the frame about the center of mass; (P fxi ,P fyi ) is the coordinate of the projection point of the center of the ith tire in the coordinate system with the center of mass of the frame as the origin, (P bxj ,P byj ) is the coordinate of the jth cab suspension in the coordinate system with the center of mass of the frame as the origin; k cj is the stiffness of the jth cab suspension; α f is the angular displacement of the frame roll, γ f is the angular displacement of the frame pitch; is the angular acceleration of the frame roll, is the angular acceleration of the frame pitch; where j = 1, 2, …, 4.

[0032] The three-degree-of-freedom vibration differential equation of the cab part:

[0033]

[0034] In the formula, m c is the cab mass; c is the vertical displacement of the cab; is the vertical acceleration of the cab; I cxx ,I cyy is the moment of inertia of the cab about its center of mass; (P cxj ,P cyj ) is the coordinate of the cab suspension in the coordinate system with the cab center of mass as the origin; α c is the angular displacement of the cab roll, γ c is the angular displacement of the cab pitch; is the angular acceleration of the cab roll, is the angular acceleration of the cab pitch.

[0035] The kinematic relationship between the three degrees of freedom (roll, pitch and vertical) motion of the frame and cab and each suspension is as follows:

[0036]

[0037] The random road model is selected as white noise road:

[0038] According to ISO / TC108 / SC2N67 and GB7031-86 "Vehicle Vibration Input-Road Flatness Representation Method Standard", the random road model is established using the filtered white noise method, as follows:

[0039]

[0040] Where n 00 is the lower cutoff spatial frequency, u is the vehicle speed, z r (t) is the random road elevation displacement, is the vertical vibration velocity of the random road surface, n0 is the reference spatial frequency, G q (n0) is the road roughness coefficient, and W(t) is Gaussian white noise with a mean of 0.

[0041] Since the present invention only involves the suspension damping control of the cab part, the resonant frequency of the vehicle dynamics system, that is, the vertical resonant frequency ω0 of the cab part is:

[0042]

[0043] k cj is the stiffness of the jth cab suspension.

[0044] Step S2, respectively design a low-pass IIR filter and a high-pass IIR filter, combine the vehicle dynamics system resonance frequency ω0, preliminarily determine the two IIR filter orders and cutoff frequency points, design a filter frequency division controller, and preliminarily determine the sampling time length and the high and low frequency judgment threshold values, as follows:

[0045] Step S2.1, design a low-pass IIR filter and a high-pass IIR filter respectively:

[0046] The transfer function expression of the first-order low-pass IIR filter is:

[0047]

[0048] Among them, ω L is the cutoff frequency of the low-pass IIR filter, s represents the complex variable in the Laplace transform, and Y L represents the complex domain low-pass IIR filter output, X L Represents the low-pass IIR filter input in the complex domain.

[0049] Discretize the above formula, that is, let s = (1-z -1 ) / T s , we get the difference equation:

[0050]

[0051] Among them, y L,j (n) is the filtering value of the sprung low-pass IIR filter corresponding to the nth and jth cab suspension, y L,j (n-1) is the filtering value of the sprung low-pass IIR filter corresponding to the j-th cab suspension at the n-1th order, is the sprung acceleration sampling value corresponding to the jth cab suspension at the nth time, T s is the sampling time, and z represents the complex variable in the Z transform.

[0052] The transfer function expression of the first-order high-pass IIR filter is:

[0053]

[0054] Among them, ω H is the cutoff frequency of the high-pass IIR filter, Y H represents the complex domain high-pass IIR filter output, X H Represents the high-pass IIR filter input in the complex domain.

[0055] Discretize the above equation and get the difference equation:

[0056]

[0057] Among them, yH,j (n) is the filtering value of the sprung high-pass IIR filter corresponding to the nth and jth cab suspension, y H,j (n-1) is the filtering value of the sprung high-pass IIR filter corresponding to the j-th cab suspension at the n-1th time. is the sprung acceleration sampling value corresponding to the j-th cab suspension at the n-1th time.

[0058] Set the low-pass IIR filter cutoff frequency ω L and high-pass IIR filter cutoff frequency ω H , which is initially set to the vehicle dynamics system resonant frequency ω0 calculated in step S1.

[0059] Step S2.2, design the filter frequency division controller:

[0060]

[0061] In the formula, f j is the controllable damping force of the jth cab suspension, is the unsprung speed corresponding to the jth cab suspension; is the sprung velocity corresponding to the jth cab suspension; c min is the minimum controllable damping coefficient of the semi-active shock absorber; c max is the maximum controllable damping coefficient of the semi-active shock absorber; Th is the damping output judgment threshold; k L k is the number of sampling points for low-frequency signals; H is the number of sampling points for high frequency signals; L,j (m) is the filtering value of the sprung low-pass IIR filter corresponding to the mth jth cab suspension; y H,j (l) is the filtering value of the sprung high-pass IIR filter corresponding to the lth jth cab suspension.

[0062] Step S3: Taking the comfort index of the sprung mass as the optimization target, the fmincon optimization algorithm is used to determine the optimal solutions of various parameters in the low-pass IIR filter, the high-pass IIR filter and the filter frequency division controller, so as to optimize the comprehensive performance of the semi-active suspension:

[0063] Select the low-pass IIR filter cutoff frequency ω L , high-pass IIR filter cutoff frequency ω H , and the number of sampling points k of the low-frequency signal of the filter frequency division controller L , high frequency signal sampling points k H And the threshold Th is optimized, and the objective function J is as follows:

[0064]

[0065] Where T is the simulation running time length, is the vertical acceleration vibration signal of the cab part, and t is the time.

[0066] According to the objective function J, the cutoff frequency ω of the low-pass IIR filter is L , high-pass IIR filter cutoff frequency ω H , and the number of sampling points k of the low-frequency signal of the filter frequency division controller L , high frequency signal sampling points k H The optimization is performed based on the threshold Th to make the objective function reach the global minimum value to ensure the smoothness of vehicle driving.

[0067] Example 1

[0068] Combination Figure 1 and Figure 2 , a semi-active suspension control strategy based on IIR filter frequency division, the steps are as follows:

[0069] Step S1: Establish a vehicle dynamics model and a random road surface model, and preliminarily calculate the resonant frequency of the vehicle dynamics system, as follows:

[0070] The vehicle dynamics model is a 12-degree-of-freedom commercial vehicle model, including three vibration differential equations and a kinematic relationship.

[0071] The three vibration differential equations include the vertical vibration differential equations of the six tire parts, the three-degree-of-freedom vibration differential equations of the frame part, and the three-degree-of-freedom vibration differential equations of the cab part, as follows:

[0072] Differential equations for vertical vibration of the six tire sections:

[0073]

[0074] In the formula, m wi is the mass of the i-th tire; z wi is the vertical displacement of the center of the ith tire; is the vertical velocity of the center of the ith tire; is the vertical acceleration of the center of the ith tire; k wi is the stiffness of the ith tire; q i is the input displacement at the ith tire; k si is the stiffness of the i-th tire suspension; z fi is the displacement of the frame suspension spring corresponding to the i-th tire; is the frame suspension spring speed corresponding to the i-th tire; c si is the frame suspension damping coefficient corresponding to the i-th tire; where i = 1, 2, …, 6.

[0075] The three-degree-of-freedom vibration differential equation of the frame:

[0076]

[0077] In the formula, m f is the frame mass; z f is the vertical displacement of the frame; is the vertical acceleration of the frame; z bj is the unsprung displacement corresponding to the jth cab suspension; z cj is the sprung displacement corresponding to the jth cab suspension; f j is the controllable damping force of the jth cab suspension; I fxx ,I fyy is the moment of inertia of the frame about the center of mass; (P fxi ,P fyi ) is the coordinate of the projection point of the center of the ith tire in the coordinate system with the center of mass of the frame as the origin, (P bxj ,P byj ) is the coordinate of the jth cab suspension in the coordinate system with the center of mass of the frame as the origin; k cj is the stiffness of the jth cab suspension; α f is the angular displacement of the frame roll, γ f is the angular displacement of the frame pitch; is the angular acceleration of the frame roll, is the angular acceleration of the frame pitch; where j = 1, 2, …, 4.

[0078] The three-degree-of-freedom vibration differential equation of the cab part:

[0079]

[0080] In the formula, m c is the cab mass; c is the vertical displacement of the cab; is the vertical acceleration of the cab; I cxx ,I cyy is the moment of inertia of the cab about its center of mass; (P cxj ,P cyj ) is the coordinate of the cab suspension in the coordinate system with the cab center of mass as the origin; α c is the angular displacement of the cab roll, γ c is the angular displacement of the cab pitch; is the angular acceleration of the cab roll, is the angular acceleration of the cab pitch.

[0081] The kinematic relationship between the three degrees of freedom (roll, pitch and vertical) motion of the frame and cab and each suspension is as follows:

[0082]

[0083] The random road model is selected as white noise road:

[0084] According to ISO / TC108 / SC2N67 and GB7031-86 "Vehicle Vibration Input - Road Surface Flatness Representation Method Standard", the random road surface model is established using the filtered white noise method, which can be expressed by the following formula:

[0085]

[0086] Where n 00 is the lower cutoff spatial frequency, u is the vehicle speed, z r (t) is the random road elevation displacement, is the vertical vibration velocity of the random road surface, n0 is the reference spatial frequency, G q (n0) is the road roughness coefficient, and W(t) is Gaussian white noise with a mean of 0.

[0087] The resonant frequency of the vehicle dynamics system, that is, the vertical resonant frequency ω0 of the cab part, is:

[0088]

[0089] Step S2, respectively design a low-pass IIR filter and a high-pass IIR filter, combine the vehicle dynamics system resonance frequency ω0, preliminarily determine the two IIR filter orders and cutoff frequency points, design a filter frequency division controller, and preliminarily determine the sampling time length and the high and low frequency judgment threshold values, as follows:

[0090] Step S2.1, design a low-pass IIR filter and a high-pass IIR filter:

[0091] The transfer function expression of the first-order low-pass IIR filter is:

[0092]

[0093] Among them, ω L is the cutoff frequency of the low-pass IIR filter, s represents the complex variable in the Laplace transform, and Y L represents the complex domain low-pass IIR filter output, X L Represents the low-pass IIR filter input in the complex domain.

[0094] Discretize the above formula, that is, let s = (1-z -1 ) / T s , we get the difference equation:

[0095]

[0096] Among them, y L,j (n) is the filtering value of the sprung low-pass IIR filter corresponding to the nth and jth cab suspension, y L,j (n-1) is the filtering value of the sprung low-pass IIR filter corresponding to the j-th cab suspension at the n-1th order, is the sprung acceleration sampling value corresponding to the jth cab suspension at the nth time, T s is the sampling time, and z represents the complex variable in the Z transform.

[0097] The transfer function expression of the first-order high-pass IIR filter is:

[0098]

[0099] Among them, ω H is the cutoff frequency of the high-pass IIR filter, Y H represents the complex domain high-pass IIR filter output, X H Represents the high-pass IIR filter input in the complex domain.

[0100] Discretize the above equation and get the difference equation:

[0101]

[0102] Among them, y H,j (n) is the filtering value of the sprung high-pass IIR filter corresponding to the nth and jth cab suspension, y H,j (n-1) is the filtering value of the sprung high-pass IIR filter corresponding to the j-th cab suspension at the n-1th time. 1) is the sprung acceleration sampling value corresponding to the j-th cab suspension at the n-1th time.

[0103] Set the low-pass IIR filter cutoff frequency ω L and high-pass IIR filter cutoff frequency ω H , which is initially set to the vehicle dynamics system resonant frequency ω0 calculated in step S1.

[0104] Step S2.2, design the filter frequency division controller:

[0105]

[0106] In the formula, f j is the controllable damping force of the jth cab suspension, is the unsprung speed corresponding to the jth cab suspension; is the sprung velocity corresponding to the jth cab suspension; c min is the minimum controllable damping coefficient of the semi-active shock absorber; c maxis the maximum controllable damping coefficient of the semi-active shock absorber; Th is the damping output judgment threshold; k L k is the number of sampling points for low-frequency signals; H is the number of sampling points for high frequency signals; L,j (m) is the filtering value of the sprung low-pass IIR filter corresponding to the mth jth cab suspension; y H,j (l) is the filtering value of the sprung high-pass IIR filter corresponding to the lth jth cab suspension.

[0107] Step S3: Taking the comfort index of the sprung mass as the optimization target, the fmincon optimization algorithm is used to determine the optimal solutions of various parameters in the low-pass IIR filter, the high-pass IIR filter and the filter frequency division controller, so as to optimize the comprehensive performance of the semi-active suspension:

[0108] Select the low-pass IIR filter cutoff frequency ω L , high-pass IIR filter cutoff frequency ω H , and the number of sampling points k of the low-frequency signal of the filter frequency division controller L , high frequency signal sampling points k H And the threshold Th is optimized, and the objective function J is as follows:

[0109]

[0110] Where T is the simulation running time length, It is the vertical acceleration vibration signal of the cab.

[0111] According to the objective function J, the cutoff frequency ω of the low-pass IIR filter is L , high-pass IIR filter cutoff frequency ω H , and the number of sampling points k of the low-frequency signal of the filter frequency division controller L , high frequency signal sampling points k H The optimization is performed based on the threshold Th to make the objective function reach the global minimum value to ensure the smoothness of vehicle driving.

[0112] In this example, the vehicle dynamics parameters are selected as shown in Table 1:

[0113] Table 1 Vehicle dynamics model parameters

[0114]

[0115] In this example, the vehicle geometry parameters are selected as shown in Table 2:

[0116] Table 2 Vehicle geometric parameters

[0117]

[0118] In this example, the initial parameters and optimization results of the low / high frequency filter and controller parameters are shown in Table 3:

[0119] Table 3 Values ​​of parameters before and after optimization

[0120]

[0121] The optimized parameters are introduced into the actual vehicle test, and the results are shown in Table 4:

[0122] Table 4 Comparison of actual vehicle test data

[0123]

[0124] It can be found from Table 4 that the filter frequency division control algorithm can effectively improve the comfort index of the vehicle in the x-direction and z-direction. From the comprehensive data, when the threshold is 0.9 and 1.2, the improvement effect is similar, which is improved by 11.0% and 10.2% respectively, among which the highest optimization is 21.3% in the x-direction and 11.7% in the z-direction. Figure 3 It can be seen that under the excitation of random road surfaces, by continuously increasing the controller threshold, the amplitude above the crossover point increases, the amplitude below the crossover point decreases, and the optimal effect is achieved when the threshold is 0.9, which is significantly improved compared to the passive suspension. In addition, this control strategy requires a small number of sensors, low complexity, and low hardware computing power requirements, which can effectively reduce the cost of semi-active suspension, has certain engineering practical application value, and is highly scalable.

Claims

1. A semi-active suspension control strategy based on IIR filter frequency division, characterized in that: Here are the steps: Step S1, establishing a vehicle dynamics model and a random road surface model, preliminarily calculating the resonance frequency of the vehicle dynamics system, and then proceeding to step S2; Step S2, respectively design a low-pass IIR filter and a high-pass IIR filter, combine the resonance frequency ω0 of the vehicle dynamics system, preliminarily determine the orders and cutoff frequency points of the two IIR filters, design a filter frequency division controller, preliminarily determine the sampling time length and the size of the high and low frequency judgment thresholds, and then proceed to step S3; Step S3, taking the comfort index of the sprung mass as the optimization target, using the fmincon optimization algorithm to determine the optimal solutions for various parameters in the low-pass IIR filter, the high-pass IIR filter and the filter frequency division controller, so as to optimize the comprehensive performance of the semi-active suspension.

2. The semi-active suspension control strategy based on IIR filter frequency division according to claim 1, characterized in that: In step S1, the vehicle dynamics model is a 12-degree-of-freedom commercial vehicle model; the random road surface model is selected as a white noise road surface.

3. The semi-active suspension control strategy based on IIR filter frequency division according to claim 2, characterized in that: In step S1, a vehicle dynamics model is established, which is specifically as follows: the vehicle dynamics model includes three vibration differential equations and a kinematic relationship; The three vibration differential equations include the vertical vibration differential equations of the six tire parts, the three-degree-of-freedom vibration differential equations of the frame part, and the three-degree-of-freedom vibration differential equations of the cab part.

4. The semi-active suspension control strategy based on IIR filter frequency division according to claim 3 is characterized in that: In step S1, a vehicle dynamics model is established, as follows: Differential equations for vertical vibration of the six tire parts: In the formula, m wi is the mass of the i-th tire; z wi is the vertical displacement of the center of the ith tire; is the vertical velocity of the center of the ith tire; is the vertical acceleration of the center of the ith tire; k wi is the stiffness of the ith tire; q i is the input displacement at the ith tire; k si is the stiffness of the i-th tire suspension; z fi is the displacement of the frame suspension spring corresponding to the i-th tire; is the frame suspension spring speed corresponding to the i-th tire; is the acceleration of the frame suspension spring corresponding to the i-th tire; c si is the frame suspension damping coefficient corresponding to the i-th tire; where i=1,2,…,6; The three-degree-of-freedom vibration differential equation of the frame is: In the formula, m f is the frame mass; z f is the vertical displacement of the frame; is the vertical acceleration of the frame; z bj is the unsprung displacement corresponding to the jth cab suspension; z cj is the sprung displacement corresponding to the jth cab suspension; f j is the controllable damping force of the jth cab suspension; I fxx ,I fyy is the moment of inertia of the frame about the center of mass; (P fxi ,P fyi ) is the coordinate of the projection point of the center of the ith tire in the coordinate system with the center of mass of the frame as the origin, (P bxj ,P byj ) is the coordinate of the jth cab suspension in the coordinate system with the center of mass of the frame as the origin; k cj is the stiffness of the jth cab suspension; α f is the angular displacement of the frame roll, γ f is the angular displacement of the frame pitch; is the angular acceleration of the frame roll, is the angular acceleration of the frame pitch; where j = 1, 2, …, 4; The three-degree-of-freedom vibration differential equation of the cab part: In the formula, m c is the cab mass; z c is the vertical displacement of the cab; is the vertical acceleration of the cab; I cxx ,I cyy is the moment of inertia of the cab about its center of mass; (P cxj ,P cyj ) is the coordinate of the cab suspension in the coordinate system with the cab center of mass as the origin; α c is the angular displacement of the cab roll, γ c is the angular displacement of the cab pitch; is the angular acceleration of the cab roll, is the angular acceleration of the cab pitch; The kinematic relationship between the three-degree-of-freedom motion of the frame and cab and each suspension is as follows: In the formula, z fi is the displacement of the frame suspension spring corresponding to the i-th tire; z bj is the unsprung displacement corresponding to the jth cab suspension; z cj is the sprung displacement corresponding to the jth cab suspension.

5. The semi-active suspension control strategy based on IIR filter frequency division according to claim 4, characterized in that: In step S1, a random road surface model is established, as follows: The random road surface model is established using the filtered white noise method, as follows: Where n 00 is the lower cutoff spatial frequency, u is the vehicle speed, z r (t) is the random road elevation displacement, is the vertical vibration velocity of the random road surface, n0 is the reference spatial frequency, G q (n0) is the road roughness coefficient, W(t) is Gaussian white noise with a mean of 0; The resonant frequency of the vehicle dynamics system, that is, the vertical resonant frequency ω0 of the cab part, is: k cj is the stiffness of the jth cab suspension.

6. The semi-active suspension control strategy based on IIR filter frequency division according to claim 5, characterized in that: In step S2, a low-pass IIR filter and a high-pass IIR filter are designed respectively, and the orders and cutoff frequency points of the two IIR filters are preliminarily determined in combination with the resonant frequency ω0 of the vehicle dynamics system, as follows: The transfer function expression of the first-order low-pass IIR filter is: Among them, ω L is the cutoff frequency of the low-pass IIR filter, s represents the complex variable in the Laplace transform, and Y L represents the complex domain low-pass IIR filter output, X L represents the low-pass IIR filter input in the complex domain; Discretize the above formula, that is, let s = (1-z -1 ) / T s , we get the difference equation: Among them, y L,j (n) is the filtering value of the sprung low-pass IIR filter corresponding to the nth jth cab suspension, y L,j (n)-1) is the filtering value of the sprung low-pass IIR filter corresponding to the j-th cab suspension at the n-1th order, is the sprung acceleration sampling value corresponding to the jth cab suspension at the nth time, T s is the sampling time, z represents the complex variable in the Z transform; The transfer function expression of the first-order high-pass IIR filter is: Among them, ω H is the cutoff frequency of the high-pass IIR filter, Y H represents the complex domain high-pass IIR filter output, X H represents the high-pass IIR filter input in the complex domain; Discretize the above equation and get the difference equation: Among them, y H,j (n) is the filtering value of the sprung high-pass IIR filter corresponding to the nth and jth cab suspension, y H,j (n-1) is the filtering value of the sprung high-pass IIR filter corresponding to the j-th cab suspension at the n-1th time. is the sprung acceleration sampling value corresponding to the j-th cab suspension at the n-1th time; Set the low-pass IIR filter cutoff frequency ω L and high-pass IIR filter cutoff frequency ω H , which is initially set to the resonant frequency ω0 of the vehicle dynamics system.

7. The semi-active suspension control strategy based on IIR filter frequency division according to claim 6, characterized in that: In step S2, a filter frequency division controller is designed, as follows: In the formula, f j is the controllable damping force of the jth cab suspension, is the unsprung speed corresponding to the jth cab suspension; is the sprung velocity corresponding to the jth cab suspension; c min is the minimum controllable damping coefficient of the semi-active shock absorber; c max is the maximum controllable damping coefficient of the semi-active shock absorber; Th is the damping output judgment threshold; k L k is the number of sampling points for low-frequency signals; H is the number of sampling points for high frequency signals; L,j (m) is the filtering value of the sprung low-pass IIR filter corresponding to the mth jth cab suspension; y H,j (l) is the filtering value of the sprung high-pass IIR filter corresponding to the lth jth cab suspension.

8. The semi-active suspension control strategy based on IIR filter frequency division according to claim 7, characterized in that: In step S3, the comfort index of the sprung mass is taken as the optimization target, and the fmincon optimization algorithm is used to determine the optimal solutions of various parameters in the low-pass IIR filter, the high-pass IIR filter and the filter frequency division controller, so as to optimize the comprehensive performance of the semi-active suspension: Select the low-pass IIR filter cutoff frequency ω L , high-pass IIR filter cutoff frequency ω H , and the number of sampling points k of the low-frequency signal of the filter frequency division controller L , high frequency signal sampling points k H And the threshold Th is optimized, and the objective function J is as follows: Where T is the simulation running time length, is the vertical acceleration vibration signal of the cab part, t is the time; According to the objective function J, the cutoff frequency ω of the low-pass IIR filter is L , high-pass IIR filter cutoff frequency ω H , and the number of sampling points k of the low-frequency signal of the filter frequency division controller L , high frequency signal sampling points k H The optimization is performed based on the threshold Th to make the objective function reach the global minimum value to ensure the smoothness of vehicle driving.

Images