A vehicle suspension dynamics state observation method and control method

Abstract

The present invention proposes a vehicle suspension dynamic state observation method and control method, which can estimate the state quantity of the suspension system and the road elevation by only using the easily measurable dynamic stroke (or dynamic velocity) signal. The estimation method does not require the sprung or unsprung mass acceleration signal required by the traditional method, which can significantly reduce the cost of subsequent processing of the control system sensor and controller signals, avoids the integral operation of the measurement signal and the problems caused by the use and processing of high-frequency signals, and is more conducive to engineering implementation and improving system reliability. The state estimation method proposed in the present invention can be used for other control systems of vehicle chassis dynamics, that is, the engineering implementation of the control system requires a control method that relies on the vehicle suspension dynamic state for state feedback.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle engineering, and in particular relates to a vehicle suspension dynamics state observation method and a control method. Background Art

[0002] The chassis dynamics system of unmanned vehicles has many controllers, such as suspension controllers, steering controllers, brake controllers, etc. The engineering implementation of these controllers requires that all state variables of the system can be measured. However, in actual engineering, the increase in the number of sensors will greatly increase the cost of the control system and reduce the reliability of the system; the more complicated problem is that some state quantities such as road elevation, tire deformation, absolute displacement of sprung and unsprung mass, absolute speed and other signals are difficult to measure through sensors. There are many reasons for the difficulty in measurement, such as the road elevation, the absolute displacement of the sprung and unsprung mass of the suspension is difficult to measure due to the lack of a reference system, and there are also practical problems encountered in engineering implementation such as price, installation space and reliability. Therefore, the problem of obtaining state variables is a practical problem that must be faced in the engineering implementation of the control algorithm of the vehicle chassis dynamics controller. Summary of the invention

[0003] In view of this, an object of the present invention is to provide a method for feedback control of vehicle chassis dynamics states, which can significantly reduce the cost of subsequent processing of control system sensor and controller signals.

[0004] A vehicle suspension dynamic state observation method, comprising:

[0005] According to Newton's second law, the dynamic equation of the suspension system is as follows:

[0006]

[0007] Among them, m b is the sprung mass, that is, the vehicle mass, m w is the unsprung mass, that is, the wheel mass, k s is the spring stiffness, that is, the stiffness of the suspension, c is the damping coefficient of the suspension, k t is the wheel stiffness, u is the suspension control force, t represents the time variable, and w(t) is the road excitation;

[0008] The road surface adopts the form of integrated white noise:

[0009]

[0010] Where f0 is the cutoff spatial frequency, n0 is the reference spatial frequency, G q (n0) is the road roughness coefficient, v0 is the vehicle speed, and w(t) is the unit white noise;

[0011] Select the state variables as: Written in state space form:

[0012]

[0013] Where w is the model noise, v is the measurement noise, and y is the output of the system, i.e., the suspension travel x b -x w ;

[0014] in:

[0015]

[0016] C = [1 0 -1 0 0];

[0017] The state observer equation is:

[0018]

[0019] In the formula, is the estimated value of state x; is the estimated value of the output y; L is the Kalman filter gain matrix.

[0020] The Kalman filter gain matrix L is:

[0021] L=PC T V -1 (5)

[0022] Where P is the solution of the following Riccati equation:

[0023]

[0024] Where V is the measurement noise variance matrix and W is the model noise variance matrix.

[0025] A control method for a vehicle suspension dynamics state observation method uses the following control rate to control the vehicle suspension:

[0026]

[0027] Where K is the state feedback gain matrix.

[0028] Preferably, the control method for the vehicle suspension adopts a linear quadratic LQ control method.

[0029] Preferably, the vehicle suspension control method adopts a model predictive control (MPC) method.

[0030] Preferably, the control method for the vehicle suspension adopts a sliding mode control SMC method.

[0031] The present invention has the following beneficial effects:

[0032] The present invention proposes a vehicle suspension dynamic state observation method and control method, which can estimate the state quantity of the suspension system and the road elevation by only using the easily measurable dynamic stroke (or dynamic velocity) signal. The estimation method does not require the sprung or unsprung mass acceleration signal required by the traditional method, which can significantly reduce the cost of subsequent processing of the control system sensor and controller signals, avoids the integral operation of the measurement signal and the problems caused by the use and processing of high-frequency signals, and is more conducive to engineering implementation and improving system reliability. The state estimation method proposed in the present invention can be used for other control systems of vehicle chassis dynamics, that is, the engineering implementation of the control system requires a control method that relies on the vehicle suspension dynamic state for state feedback. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is the active suspension mechanics diagram;

[0034] Figure 2 Implement a block diagram for the state observer;

[0035] Figure 3 is the sprung mass displacement estimate;

[0036] Figure 4 is the sprung mass velocity estimate;

[0037] Figure 5 is the unsprung mass velocity estimate;

[0038] Figure 6 is the unsprung mass displacement estimate;

[0039] Figure 7 For road surface estimation;

[0040] Figure 8 is the sprung mass displacement estimate;

[0041] Fig. 9 Estimated for road surface. DETAILED DESCRIPTION

[0042] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0043] The engineering implementation of a vehicle chassis dynamics state feedback control method proposed in the present invention is to estimate the state variables through a designed state observer. The specific idea is:

[0044] A state-space dynamic model that is consistent with the state feedback quantity of the control algorithm is established in the controller. Certain state quantities that can be or are convenient to measure in the actual system are used as inputs of the dynamic model in the controller to stimulate the model and generate a response. A feedback matrix of the deviation between the model response and the measured state quantity is determined through a specific algorithm to implement closed-loop feedback regulation of the model response. This allows the estimation of all state variables of the system to be achieved while satisfying the performance of the controller. The estimated quantities in the controller are then used instead of actual physical measurements to meet the needs of engineering implementation of a type of chassis dynamics state feedback control method.

[0045] The present invention proposes a new method for estimating the entire state of the suspension by only measuring the dynamic stroke or dynamic speed of the suspension. The first feature of the estimation method of the present invention is that the dynamic stroke or dynamic speed of the suspension is easier to measure than signals such as the acceleration of the sprung and unsprung masses, and the sensor is inexpensive and easy to install. The dynamic stroke of the suspension can be measured by an angle sensor or a linear displacement sensor. In this case, the dynamic speed can be obtained by numerically differentiating the dynamic stroke. For a type of active suspension with a permanent magnet motor as an actuator, the dynamic speed of the suspension can be directly obtained by equivalently transforming the rotation speed of the motor shaft head; the dynamic stroke can be obtained by the shaft head absolute position encoder. The second feature of the estimation method of the present invention is that compared with the previous measurement of the sprung mass and unsprung mass acceleration, the dynamic stroke is a low-frequency, slow-changing signal, which is conducive to the sampling and processing of the controller, greatly reducing the cost of signal processing of the control system and improving the system reliability. The third feature of the estimation method of the present invention is that the road elevation is augmented into the system matrix, and the road elevation signal can be estimated together with the state estimation.

[0046] 1. Observer Design Method

[0047] like Figure 1 As shown in the figure, it is the active suspension mechanics diagram; b is the sprung mass, that is, the vehicle mass, m w is the unsprung mass, that is, the wheel mass, k s is the spring stiffness, that is, the stiffness of the suspension, c is the damping coefficient of the suspension, k t is the wheel stiffness, u is the suspension control force, t represents the time variable, and w(t) is the road excitation;

[0048] According to Newton's second law, the dynamic equation of the suspension system is as follows:

[0049]

[0050] The road surface adopts the form of integrated white noise:

[0051]

[0052] Where f0 is the cutoff spatial frequency, n0 is the reference spatial frequency, G q (n0) is the road roughness coefficient, v0 is the vehicle speed, and w(t) is the unit white noise.

[0053] Select the state variables as: Written in state space form:

[0054]

[0055] Where w is the model noise, v is the measurement noise, and y is the output of the system, i.e., the suspension travel x b -x w .

[0056] in:

[0057]

[0058] The necessary and sufficient condition for observability is the matrix The rank of is n.

[0059] It has been verified that when only the dynamic stroke or dynamic speed is used as the output, C = [1 0 -1 0 0] or C = [0 10 -1 0]. The matrix N is full rank.

[0060] When the system output is the suspension travel, that is: C = [1 0 -1 0 0];

[0061] The state observer equation is:

[0062]

[0063] In the formula, is the estimated value of state x; is the estimated value of the output y; L is the Kalman filter gain matrix.

[0064] The Kalman filter gain matrix L is:

[0065] L=PC T V -1 (5)

[0066] Where P is the solution of the following Riccati equation:

[0067]

[0068] Where V is the measurement noise variance matrix and W is the model noise variance matrix.

[0069] 2. Control Methods

[0070] The present invention provides a vehicle chassis dynamics control method, which involves a method for estimating vehicle suspension dynamics state parameters, and aims to solve the engineering implementation problem of a vehicle chassis dynamics control method whose control law is a full state feedback form. Applicable control methods include: linear quadratic LQ control, model predictive control MPC method, sliding mode control SMC method, etc.

[0071] The characteristic of the above listed control methods is that the control law is in the form of full-state feedback. When implementing such control algorithms in engineering, the observer design method of the present invention can be used to estimate the required state, and the computer estimated value of the state can be used instead of the actual physical measurement value, thereby solving the problem of difficulty in measuring certain state variables in engineering implementation.

[0072] The control method implementation process is as follows:

[0073] The control law is: u=-Kx (7)

[0074] Where K is the state feedback gain matrix, which is determined by the specific control algorithm.

[0075] Using estimated values ​​instead of measured values, we can get:

[0076]

[0077] When the control rate is in the form of state feedback, the observer equation represented by equation (4) can be specifically written as:

[0078]

[0079] The formation and implementation process of the observation method and control method is as follows: Figure 2 As shown. The physical system is represented by equation (3), and the solution process of P is obtained by equation (6).

[0080] In summary, any control law in the form of full state feedback can be implemented using the above-mentioned state observer structure.

[0081] 3. Simulation Verification of Observation Effect

[0082] The system is stimulated with a random signal and the true value is compared with the estimate of the state observer. Figure 3-7 The figure shows the comparison between the true value and the estimated value of each state under random road excitation. It can be seen that the state of the suspension system can be well estimated by measuring the dynamic stroke alone. The estimation accuracy is shown in Table 1, which shows that the estimation accuracy meets the general engineering requirements.

[0083] The estimated accuracy can be calculated as follows:

[0084]

[0085] Table 1 Estimation accuracy

[0086]

[0087] When a sine wave is used to excite the system, the effect of the state observation method proposed in the present invention is as follows: Figure 8-9 It can be seen that the proposed state observation method also has a good estimation effect under deterministic road surface excitation, that is, the state observation method proposed in the present invention is still applicable to deterministic road surface input.

[0088] IV. Conclusion

[0089] The present invention proposes a vehicle chassis suspension state observer design method, which can effectively estimate all states of the suspension system and the road surface at the same time under the premise of measuring only the dynamic stroke. It has good estimation effect in both random road surface and deterministic road surface conditions. On the one hand, only measuring the dynamic stroke reduces the number of required sensors, reduces costs, and increases the reliability of the system; on the other hand, the dynamic stroke (or dynamic speed) is easy to obtain, which is conducive to engineering implementation. The state estimation method proposed in the present invention can be applied to the dynamic control of vehicle chassis related systems, and is suitable for a type of vehicle chassis dynamics control method whose control law is a full-state feedback form. When implementing such control algorithms in engineering, the observation method of the present invention is used to estimate the required state, and the estimated value is used instead of the actual physical measurement value, which can solve the problem that all state variables cannot be obtained in engineering implementation. The present invention provides the specific process of forming and implementing the estimation method and the control method.

[0090] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A vehicle suspension dynamic state observation method, characterized in that: include: According to Newton's second law, the dynamic equation of the suspension system is as follows: Among them, m b is the sprung mass, that is, the vehicle mass, m w is the unsprung mass, that is, the wheel mass, k s is the spring stiffness, that is, the stiffness of the suspension, c is the damping coefficient of the suspension, k t is the wheel stiffness, u is the suspension control force, t represents the time variable, and w(t) is the road excitation; The road surface adopts the form of integrated white noise: Where f0 is the cutoff spatial frequency, n0 is the reference spatial frequency, G q (n0) is the road roughness coefficient, v0 is the vehicle speed, and w(t) is the unit white noise; Select the state variables as: Written in state space form: Where w is the model noise, v is the measurement noise, and y is the output of the system, i.e., the suspension travel x b -x w ; in: C＝[1 0-1 0 0]； The state observer equation is: In the formula, is the estimated value of state x; is the estimated value of the output y; L is the Kalman filter gain matrix; The Kalman filter gain matrix L is: L=PC T V -1 (5) Where P is the solution of the following Riccati equation: Where V is the measurement noise variance matrix, W is the model noise variance matrix; the following control rate is used to control the vehicle suspension: Where K is the state feedback gain matrix; The control method for the vehicle suspension adopts a linear quadratic LQ control method, a model predictive control MPC method or a sliding mode control SMC method.

Images