Automobile height control method based on electronic control air suspension

Abstract

The invention discloses an automobile height control method based on an electronic control air suspension. The automobile height control method comprises the following steps that firstly, the displacement variation, the vibration speed and the working state air pressure in an air chamber of an automobile body are collected; step 2, constructing an automobile height state equation of the electric control air suspension; 3, according to the automobile height state equation, a sliding mode surface controller is constructed, and the expected opening degree of an electromagnetic valve is obtained; fourthly, an error adjusting network is constructed, the expected opening degree of the electromagnetic valve is output to the error adjusting network, and the opening degree error of the electromagnetic valve is adjusted; wherein the error adjusting network comprises an RBF (Radial Basis Function) network identifier and a PID controller; the RBF network identifier comprises two input layer nodes, six hidden layer nodes and one output layer node; the electric control air suspension has the advantages that comfort is guaranteed, and meanwhile the adjusting accuracy of the electric control air suspension can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle assembly control, and more specifically, the present invention relates to a vehicle height control method based on an electronically controlled air suspension. Background Art

[0002] In order to balance the comfort and off-road performance of a vehicle, an air suspension has emerged. As the most important part of a vehicle system, it can adjust the damping of the suspension in real time according to different road conditions and different riding requirements, and then adjust the height of the vehicle body in real time.

[0003] Regarding the wide application of electronically controlled air suspensions, their control strategies play a crucial role in performance improvement. In existing research, such as PID control, optimal control, adaptive control, fuzzy control, neural network control, or robust control, etc., there have been varying degrees of progress. However, most of them cannot adapt to the real-time adjustment of the vehicle body height under complex working conditions, and there is a time delay. After the electronic control signal is sent, the suspension has an overshoot phenomenon and cannot accurately reach the expected position of the suspension, and it cannot effectively achieve the control performance of the electronically controlled air suspension under complex working conditions. Summary of the Invention

[0004] The purpose of the present invention is to design and develop a vehicle height control method based on an electronically controlled air suspension, which improves the adjustment accuracy of the vehicle body height through solenoid valve opening control and its error adjustment.

[0005] The technical solution provided by the present invention is as follows:

[0006] A vehicle height control method based on an electronically controlled air suspension, comprising the following steps:

[0007] Step 1: Collect the displacement change amount, vibration speed, and working state air pressure in the air chamber of the vehicle body;

[0008] Step 2: Construct a vehicle height state equation of the electronically controlled air suspension;

[0009] Step 3: According to the vehicle height state equation, construct a sliding mode surface controller to obtain the desired solenoid valve opening;

[0010] Step 4: Construct an error adjustment network, output the desired solenoid valve opening to the error adjustment network, and adjust the solenoid valve opening error;

[0011] Among them, the error adjustment network includes an RBF network identifier and a PID controller;

[0012] The RBF network identifier includes 2 input layer nodes, 6 hidden layer nodes, and 1 output layer node;

[0013] The output layer node satisfies:

[0014]

[0015] Wherein, h m (k) is the output value of the RBF network identifier, w j (k) is the weight vector, g j is the hidden layer Gaussian function, c j is the center position of the network neuron, b j is the baseband parameter of the network neuron.

[0016] Preferably, the vehicle height state equation of the electronically controlled air suspension is:

[0017]

[0018] Wherein, X is the state variable, z is the displacement change of the vehicle body, is the vibration velocity, p as is the working state air pressure in the air chamber;

[0019] The state variable satisfies:

[0020]

[0021] Wherein, u is the vehicle body height adjustment amount, i.e., the solenoid valve switch signal, m s is the vehicle body mass, V as is the air chamber volume, A as is the effective cross-sectional area of the air spring, R is the gas constant, T is the gas temperature, q m is the gas mass flow rate, k is the gas variable index, P as is the air pressure in the air chamber.

[0022] Preferably, the sliding mode surface controller includes:

[0023] The sliding mode surface function is:

[0024]

[0025] Wherein, s is the sliding mode surface, [c 1 , c 2 , c 3 is the coefficient matrix, e is the error of the vehicle body displacement state parameter, is the derivative of e;

[0026] The sliding mode surface output is:

[0027] u = [v eq -v sw -F(x)]G -1 (x) + f(t);

[0028] where \(u\) is the desired solenoid valve opening, and \(v\) eq is the equivalent control, and \(v\) sw is the switching control, \(F(x)\) is the first intermediate parameter, \(G(x)\) is the second intermediate parameter, and \(f(t)\) is the external disturbance quantity.

[0029] Preferably, the error of the vehicle body displacement state parameter satisfies:

[0030] \(e = a\) d - \(a\) 1 ;

[0031] where \(a\) 1 is the vehicle body displacement, and \(a\) d is the target vehicle body displacement.

[0032] Preferably, the equivalent control satisfies:

[0033]

[0034] Preferably, the switching control satisfies:

[0035] \(v\) sw = -ksgns;

[0036] where \(k\) is the switching gain coefficient and \(sgns\) is the sign function.

[0037] Preferably, the update of the sign function satisfies:

[0038]

[0039] where \(\varphi\) is the boundary layer thickness.

[0040] Preferably, the input layer nodes of the RBF network identifier include the desired solenoid valve opening and the derivative of the desired solenoid valve opening;

[0041] The hidden layer nodes of the RBF network identifier are radial basis vectors composed of Gaussian fuzziness.

[0042] Preferably, the update of the RBF network satisfies:

[0043]

[0044] where \(\alpha\) is the momentum factor and \(\eta\) is the first learning rate.

[0045] Preferably, the control error of the PID controller satisfies:

[0046] \(e(k)=u\) e (k) - u(k);

[0047] where \(u\)e (k) Set the control target quantity under dynamic operation;

[0048] The three input parameters of the PID controller are:

[0049]

[0050] The control formula of the PID controller is:

[0051] I(k) = I(k - 1)+ΔI(k);

[0052] ΔI(k) = K p xc(1)+K i xc(2)+K d xc(3);

[0053] In the formula, I(k) is the duty ratio of the solenoid valve switch control quantity, K p is the proportionality coefficient, K i is the integral coefficient, K d is the differential coefficient.

[0054] The beneficial effects of the present invention:

[0055] A vehicle height control method based on an electronically controlled air suspension designed and developed by the present invention, according to the non-linear characteristics of the air suspension, combined with the dynamics theory, through a sliding mode surface controller combined with an error adjustment network, realizes the precise control of the air suspension height adjustment. The sliding mode surface controller predicts the solenoid valve opening control quantity, and then the error adjustment network adjusts the error of the solenoid valve opening, thereby realizing the height feedback correction of the electronically controlled air suspension, ensuring comfort while improving the adjustment accuracy of the electronically controlled air suspension. Description of the Drawings

[0056] Figure 1 is a schematic flow chart of the vehicle height control method based on the electronically controlled air suspension of the present invention.

[0057] Figure 2 is a schematic diagram of the body height change curve in the embodiment of the present invention. Detailed Embodiments

[0058] The following further describes the present invention in detail with reference to the accompanying drawings of the specification, so that those skilled in the art can implement it according to the text of the specification.

[0059] As Figure 1 shown, a vehicle height control method based on an electronically controlled air suspension provided by the present invention includes the following steps:

[0060] Step 1. Collect the displacement change amount, vibration speed and working state air pressure in the air chamber of the vehicle body;

[0061] Step 2: Construct the vehicle height state equation of the electronically controlled air suspension:

[0062]

[0063] In the formula, X is the state variable, z is the displacement change of the vehicle body, is the vibration velocity, p as is the working state air pressure in the air chamber;

[0064] That is

[0065] Among them, u is the vehicle body height adjustment amount, that is, the solenoid valve switch signal, m s is the vehicle body mass, V as is the air chamber volume, A as is the effective cross-sectional area of the air spring, R is the gas constant, T is the gas temperature, q m is the gas mass flow rate, k is the gas variable index, P as is the air pressure in the air chamber;

[0066] Then the conversion formula between the converted control amount and the vehicle body height adjustment amount is:

[0067]

[0068] In the formula,

[0069]

[0070] Step 3: Construct a sliding mode controller;

[0071] Among them, the sliding mode surface is:

[0072]

[0073] In the formula, s is the sliding mode surface, [c 1 , c 2 , c 3 is the coefficient matrix, e is the error of the vehicle body displacement state parameter, is the derivative of e;

[0074] In the linear state space, the error of the vehicle body displacement state parameter:

[0075] e = a d - a 1 ;

[0076] In the formula, a 1 is the vehicle body displacement, a d is the target vehicle body displacement;

[0077] And the vehicle body displacement and the target vehicle body displacement satisfy:

[0078] a 1 = h - h(k);

[0079] a d = h e - h(k);

[0080] In the formula, h e is the target height, which is a constant under static specified commands, and h(k) is the initial vehicle height;

[0081] v is the output of the linear space sliding mode control quantity, and satisfies:

[0082] v = v eq + v sw ;

[0083] In the formula, v eq is the equivalent control, and v sw is the switching control;

[0084] Among them, during equivalent control, it drives the state parameters to change following the sliding mode switching surface, and finally makes the regulated variable enter the switching surface

[0085]

[0086] That is, when The equivalent control can be expressed as:

[0087]

[0088] During switching control, it drives the state of the system to move in the direction of approaching the sliding mode surface s = 0, and the switching control term can be defined as:

[0089] v sw = -ksgns;

[0090] In the formula, k is the switching gain coefficient, and sgns is the sign function;

[0091] The output of the sliding mode control can be expressed as:

[0092]

[0093] According to the sliding mode control conditions, an appropriate k value needs to be selected to meet the Lyapunov stability requirements, and the Lyapunov function is selected:

[0094]

[0095] Only when k > η, the sliding mode control system meets the Lyapunov stability requirements.

[0096] Update the sign function:

[0097]

[0098] In the formula, φ is the boundary layer thickness;

[0099] Update the output parameter to:

[0100]

[0101] Furthermore, the expected solenoid valve opening of the sliding mode surface output can be obtained as:

[0102]

[0103] In the formula, f(t) is the external disturbance quantity, and f(t) ≤ ε;

[0104] Step 4: Construct an error adjustment network, and the error adjustment network includes an RBF network identifier and a PID controller;

[0105] Among them, the network structure of the RBF network identifier includes 2 input layer nodes, 6 hidden layer nodes and 1 output layer node;

[0106] The input layer nodes include u and

[0107] The hidden layer nodes are radial basis vectors composed of Gaussian functions,

[0108] The output layer node is:

[0109]

[0110] In the formula, w j (k) is the weight vector, g j is the Gaussian function of the hidden layer, c j is the center position of the network neuron, b j is the baseband parameter of the network neuron;

[0111] The identification performance index is:

[0112]

[0113] In the formula, h(k) is the actual output of the controlled object;

[0114] The network update satisfies:

[0115]

[0116] In the formula, α is the momentum factor and η is the first learning rate.

[0117] The PID control satisfies:

[0118] e(k) = u e (k) - u(k);

[0119] Wherein, u e (k) is the set control target quantity under dynamic operation, and the three input parameters of the PID are:

[0120]

[0121] For the electronic control of the switching solenoid valve, the gas mass flow control signal u(k) is converted into the duty cycle I(k) of the solenoid valve switching control quantity:

[0122] I(k) = f(u(k));

[0123] Therefore, the expression of the PID controller is:

[0124] I(k) = I(k - 1) + ΔI(k);

[0125] ΔI(k) = K p xc(1) + K i xc(2) + K d xc(3);

[0126] Wherein, K p is the proportionality coefficient, K i is the integral coefficient, K d is the differential coefficient;

[0127] The tuning indexes of the PID are:

[0128]

[0129] The gradient descent method is used to adjust the K p , K i , K d parameters in real time:

[0130]

[0131] Wherein, β is the second learning rate, is the Jacobian matrix equation of the target control object system, which can be obtained by the system identifier, that is, the relationship between the opening of the switching solenoid valve and the trigger signal.

[0132] In this embodiment, the momentum factor is 0.05, the first learning rate is 0.5, the second learning rate is 0.15, the inputs for network identification are I(k), h(k), and h(k - 1), the initial weight value is 0.1, the initial Gaussian function parameter is 10, the initial center vector is 0, the initial proportional coefficient in the PID control parameters is 0.015, the initial integral coefficient is 0, and the initial derivative coefficient is 0.015. Under dynamic conditions, with the B-level standard road surface as the excitation input signal, the road surface unevenness coefficient G = 6.4×10 -5 m 2 / m -1 , the vehicle speed is 20 km / h, with the goal of increasing the vehicle body height by 0.025 m, the change in the vehicle body height is as Figure 2 shown. It can be seen that the vehicle body height can effectively track the setting of the control command under dynamic conditions, reach the target value within 3 s, and the vibration amplitude of the vehicle body displacement is effectively controlled under the double closed-loop control, avoiding the frequent charging and discharging operations caused by overcharging and over-discharging of the air spring, thus saving energy and reducing the mechanical fatigue of the air spring components.

[0133] A vehicle height control method based on an electronically controlled air suspension designed and developed by the present invention, according to the non-linear characteristics of the air suspension, combines dynamic theory, and through a sliding mode surface controller combined with an error adjustment network, realizes the precise control of the air suspension height adjustment. The sliding mode surface controller predicts the control quantity of the solenoid valve opening, and then the error adjustment network adjusts the error of the solenoid valve opening, thereby realizing the height feedback correction of the electronically controlled air suspension, ensuring comfort while improving the adjustment accuracy of the electronically controlled air suspension.

[0134] Although the embodiments of the present invention have been disclosed as above, it is not limited to the applications listed in the specification and embodiments. It can be fully applied to various fields suitable for the present invention. For those familiar with the field, additional modifications can be easily made. Therefore, without departing from the general concept defined by the claims and the equivalent scope, the present invention is not limited to the specific details and the embodiments shown and described herein.

Claims

1. A vehicle height control method based on an electronically controlled air suspension, characterized in that: The steps include: Step 1: Collect the displacement change, vibration speed and working state air pressure of the vehicle body; Step 2: Construct the vehicle height state equation of the electronically controlled air suspension; Step 3: construct a sliding surface controller according to the vehicle height state equation to obtain the desired solenoid valve opening; Step 4: construct an error adjustment network, output the desired solenoid valve opening to the error adjustment network, and adjust the solenoid valve opening error; Wherein, the error adjustment network includes an RBF network identifier and a PID controller; The RBF network identifier includes 2 input layer nodes, 6 hidden layer nodes and 1 output layer node; The output layer nodes satisfy: In the formula, h m (k) is the output value of the RBF network identifier, w j (k) is the weight vector, g j is the hidden layer Gaussian function, c j is the center position of the network neuron, b j is the baseband parameter of the network neurons.

2. The vehicle height control method based on the electronically controlled air suspension according to claim 1, characterized in that: The vehicle height state equation of the electronically controlled air suspension is: In the formula, X is the state variable, z is the displacement change of the vehicle body, is the vibration speed, p as The working pressure in the air chamber; The state variables satisfy: Among them, u is the vehicle height adjustment amount, that is, the solenoid valve switch signal, m s is the vehicle mass, V as is the volume of the air chamber, A as is the effective cross-sectional area of ​​the air spring, R is the gas constant, T is the gas temperature, q m is the gas mass flow rate, k is the gas variable index, P as is the air pressure in the air chamber.

3. The vehicle height control method based on the electronically controlled air suspension according to claim 2, characterized in that: The sliding surface controller comprises: The sliding surface function is: Where s is the sliding surface, [c1, c2, c3] is the coefficient matrix, e is the body displacement state parameter error, is the derivative of e; The sliding surface output is: u=[v eq -v sw -F(x)]G -1 (x)+f(t); Where u is the desired solenoid valve opening, v eq For equivalent control, v sw is switching control, F(x) is the first intermediate parameter, G(x) is the second intermediate parameter, and f(t) is the external disturbance.

4. The vehicle height control method based on the electronically controlled air suspension according to claim 3, characterized in that: In the linear state space, the body displacement state parameter error satisfies: it = a d -a1; Where a1 is the displacement of the vehicle body, a d is the target body displacement.

5. The vehicle height control method based on the electronically controlled air suspension according to claim 4, characterized in that: The equivalent control satisfies:

6. The vehicle height control method based on the electronically controlled air suspension according to claim 5, characterized in that: The switching control satisfies: v sw =-ksgns; Where k is the switching gain coefficient and sgns is the sign function.

7. The vehicle height control method based on the electronically controlled air suspension according to claim 6, characterized in that: The update of the symbolic function satisfies: Where φ is the boundary layer thickness.

8. The vehicle height control method based on the electronically controlled air suspension according to claim 7, characterized in that: The input layer nodes of the RBF network identifier include the desired solenoid valve opening and the derivative of the desired solenoid valve opening; The hidden layer nodes of the RBF network identifier are radial basis vectors formed by Gaussian fuzziness.

9. The vehicle height control method based on the electronically controlled air suspension as claimed in claim 8, characterized in that: The update of the RBF network satisfies: Where α is the momentum factor and η is the first learning rate.

10. The vehicle height control method based on the electronically controlled air suspension according to claim 9, characterized in that: The control error of the PID controller satisfies: e(k)=u e (k)-u(k); In the formula, u e (k) Setting control target quantity under dynamic operation; The three input parameters of the PID controller are: The control formula of the PID controller is: I(k)=I(k-1)+ΔI(k); ΔI(k)=K p xc(1)+K i xc(2)+K d xc(3); Where I(k) is the duty cycle of the solenoid valve switch control quantity, K p is the proportionality coefficient, K i is the integration coefficient, K d is the differential coefficient.

Images