CN106218715A - A kind of control method of four-wheel independent steering vehicle - Google Patents

Abstract

The present invention provides a kind of four-wheel independent steering control method for vehicle, comprises the following steps: link, trailing wheel corner sliding mode controller and yaw moment sliding mode controller are estimated in A, default preferable vehicle steering model, front wheel angle observer, interference border；B, the side slip angle measuring vehicle in real time and yaw velocity, be calculated the front wheel angle estimated value in corresponding moment；C, in real time calculate front wheel angle estimated value, real-time side slip angle, yaw velocity, it is calculated the control error of real-time side slip angle and yaw velocity；D, by interference border estimate link be calculated real-time interference boundary parameter；E, combine real-time front wheel angle estimated value, real-time interference boundary parameter and control Error Calculation and obtain real-time trailing wheel corner and yaw moment, and vehicle is controlled.The method can not only overcome sensor fault to reduce the defect of intact stability, and have and control the feature effective, stability is high.

Description

A kind of control method of four-wheel independent steering vehicle

Technical field

The present invention relates to Vehicular turn control field, be specifically related to the control method of a kind of four-wheel independent steering vehicle.

Background technology

The control stability of vehicle is an important performance of relation vehicle safe driving.Four-wheel steering (4WS) technology is The important composition of active chassis control system, is that modern vehicle improves control stability and the development trend of active safety.Control System strategy is the important research aspect of 4WS technology, by regulation trailing wheel controlling angle vehicle centroid side drift angle and yaw velocity, The control stability of high vehicle speeds and the maneuverability of lower-speed state can be effectively improved.So far, after people are for active Round to 4WS control problem, it is proposed that the proportional feedforward of front and back wheel corner, yaw rate feedback control, neural The methods such as network control.Direct yaw moment control (DYC) is also that in Current vehicle dynamic system stability contorting one is more Effective Chassis Control Technology, it by distributing the yaw fortune producing yaw moment with regulation vehicle to longitudinal force of tire Dynamic, so that it is guaranteed that vehicle run stability.At present, the most more about the report using yaw moment control intact stability, its Include optimum control, robust control, fuzzy control etc..

Owing to actual vehicle tire and ground action by contact have nonlinear characteristic, simultaneously vehicle parameter (as complete vehicle quality, Vehicle rotary inertia etc.) change the control performance of automobile body state can be produced interference effect impact.Therefore, either The rear-axle steering of 4WS vehicle controls or DYC means, and the control stability improving vehicle traveling is had by single control strategy Limit, particularly under the limiting condition such as vehicle high-speed, zig zag, it is impossible to obtains satisfied vehicle and travels control stability.

Vehicle front wheel angle sensor is the important measurement device of automobile body electric stabilizing system (ESP), in system Controller provide front wheel angle measure the signal of telecommunication, its measure accuracy and reliability direct relation controller control effect Really, so impact driving control stability and safety.But, the Vehicular vibration that the unevenness on road surface causes, or vehicle body The part device that internal problem causes is inharmonious and damages, thus can cause sensor experiences failure phenomenon so that controller without Method obtains correct front wheel angle input signal, and this will result in vehicle stability controlled system and cannot normally work, and even can lead Cause to reduce vehicle body stability, the serious consequence of traffic safety hidden danger occurs.

Summary of the invention

It is desirable to provide a kind of four-wheel independent steering control method for vehicle, this control method overcomes prior art to work as car During front wheel angle sensor fault, intact stability reduces defect affect traffic safety, has and controls effective, stability High feature.

Technical scheme is as follows: the control method of a kind of four-wheel independent steering vehicle, comprises the following steps:

A, default preferable vehicle steering model, front wheel angle estimate that initial value, front wheel angle observer, interference border are estimated Link, trailing wheel corner sliding mode controller and yaw moment sliding mode controller；

B, using vehicle craspedodrome state as initial time, measure side slip angle and the yaw velocity of vehicle in real time, with Time front wheel angle estimated the preferable vehicle steering model of initial value input, initially expected side slip angle and initially expected horizontal stroke Pivot angle speed；By initially expect side slip angle and initial expectation yaw velocity respectively with the real-time barycenter lateral deviation in corresponding moment Angle compares with yaw velocity, and the side slip angle respectively obtaining the corresponding moment controls error and yaw velocity control Error；Side slip angle controls error, the input of initial horizontal pivot angle speed controlling error obtains correspondence to front wheel angle observer The front wheel angle estimated value in moment；

C, carry out the calculating of front wheel angle estimated value in real time, the real-time preferable vehicle of front wheel angle estimated value input is turned To model, obtain real-time expectation side slip angle and expectation yaw velocity, by real-time expectation side slip angle and expectation Yaw velocity compares with real-time side slip angle, yaw velocity, thus obtains real-time side slip angle and control Error, yaw velocity control error；

D, real-time side slip angle is controlled error, yaw velocity controls error input nonlinearities border and estimates link, Obtain real-time interference boundary parameter；

E, by real-time front wheel angle estimated value, side slip angle controls error, yaw velocity controls error and dry Disturbing boundary parameter and be jointly input to trailing wheel corner sliding mode controller and yaw moment sliding mode controller, output obtains real-time respectively Trailing wheel corner and yaw moment, and use real-time trailing wheel corner and yaw moment that vehicle is controlled in real time.

Preferably, in described step A, the construction process of preferable vehicle steering model is as follows:

Set up following Vehicular turn kinematics model:

$\left\{\begin{array}{c}m({\stackrel{\·}{v}}_x-v_y\mathrm{\γ})=(F_{x1}+F_{x2}){\mathrm{cos\δ}}_f-(F_{y1}+F_{y2}){\mathrm{sin\δ}}_f+(F_{x3}+F_{x4}){\mathrm{cos\δ}}_r-(F_{y3}+F_{y4}){\mathrm{sin\δ}}_r\\ m({\stackrel{\·}{v}}_y+v_x\mathrm{\γ})=(F_{x1}+F_{x2}){\mathrm{sin\δ}}_f+(F_{y1}+F_{y2}){\mathrm{cos\δ}}_f+(F_{x3}+F_{x4}){\mathrm{sin\δ}}_r+(F_{y3}+F_{y4}){\mathrm{cos\δ}}_r\\ I_z\stackrel{\·}{\mathrm{\γ}}=a(F_{y1}+F_{y2}){\mathrm{cos\δ}}_f-b(F_{y3}+F_{y4}){\mathrm{cos\δ}}_r+0.5W\[(F_{y1}+F_{y2}){\mathrm{sin\δ}}_f+(F_{y3}-F_{y4}){\mathrm{sin\δ}}_r\]+M\\ J_{wi}{\stackrel{\·}{\mathrm{\ω}}}_i=M_{di}-F_{xi}R-M_{bi}(i=1,2,3,4)\end{array}\right.---\left(1\right);$

In formula: m is complete vehicle quality；v_x、v_yRepresent automobile systemic velocity V velocity component on x and y-axis respectively； Represent automobile systemic velocity V component of acceleration on x and y-axis respectively；γ is automobile yaw velocity,Then represent yaw angle Acceleration；A and b is automobile barycenter to front axle and the distance of rear axle, vehicle wheel base L=a+b respectively；F_xi、F_yiRepresent automobile respectively The longitudinal force of tire and cross force, wherein subscript i=1,2,3,4 the most corresponding the near front wheels, off-front wheel, left rear wheel and off hind wheel； δ_f、δ_rIt is front and rear wheel steering angle respectively；I_zFor automobile around the rotary inertia of z-axis；J_wiAnd ω_iIt is respectively the rotary inertia of each tire And rotational angular velocity,Represent the angle of rotation acceleration of each tire；M_diIt it is the output moment of torsion on differential side；R represents tire Radius；M_biBraking moment suffered by tire；W is wheelspan, i.e. front tread B_fWith rear tread B_rIt is equal to W；M represents suffered by wheel Longitudinal force is produced the yaw moment of additional control:

M=a (F_x1+F_x2)sinδ_f-b(F_x3+F_x4)sinδ_r+0.5W[(F_x2-F_x1)cosδ_f+(F_x4-F_x3)cosδ_r] (2)；

Vehicle centroid side drift angle: β=arctan (v_x/v_y)；

The side drift angle α of front and back wheel_i:

$\left\{\begin{array}{c}{\mathrm{\α}}_1={\mathrm{\α}}_2\≈\mathrm{\β}+a\mathrm{\γ}/v_x-{\mathrm{\δ}}_f\\ {\mathrm{\α}}_3={\mathrm{\α}}_4\≈\mathrm{\β}-b\mathrm{\γ}/v_x-{\mathrm{\δ}}_r\end{array}\right.---\left(3\right);$

The wherein the most corresponding the near front wheel of subscript i=1,2,3,4, off-front wheel, left rear wheel and off hind wheel；

Assuming that under the driving cycle that automobile is in the non-emergent state of normal speed per hour scope and low-angle turns to, have v_x≈ V, And only consider defective steering stabilizer and weaving, i.e. select side slip angle and yaw velocity as the measurement master of control stability Want index, convolution (1) and (3) can obtain the kinetics equation of vehicle 2DOF linear single track model:

$\left\{\begin{array}{c}mu(\stackrel{\·}{\mathrm{\β}}+\mathrm{\γ})=F_{y1}+F_{y2}+F_{y3}+F_{y4}\\ I_z\stackrel{\·}{\mathrm{\γ}}=a(F_{y1}+F_{y2})-b(F_{y3}+F_{y4})+M\end{array}\right.---\left(4\right);$

In formula: F_y1+F_y2、F_y3+F_y4Represent the lateral deviation power of axle tire respectively

$\left\{\begin{array}{c}{F}_{y1}+F_{y2}=-k_f(\mathrm{\β}+a\mathrm{\γ}/V-{\mathrm{\δ}}_f)\\ F_{y3}+F_{y4}=-k_r(\mathrm{\β}-b\mathrm{\γ}/V-{\mathrm{\δ}}_r)\end{array}\right.---\left(5\right);$

Wherein k_fAnd k_rIt is respectively the comprehensive cornering stiffness of front axle both sides tire, the comprehensive cornering stiffness of rear axle both sides tire, Its numerical value is 2 times of front and rear wheel cornering stiffness；

Definition system state vector x=[β, γ]^TWith control input vector u=[δ_r,M]^T, set up according to formula (4) and (5) Following model state space equation is:

$\stackrel{\·}{x}=Ax+Bu+{\mathrm{C\δ}}_f---\left(6\right);$

In formula:ForSytem matrix

Control input matrixFront wheel angle input matrix

Use front wheel angle observer output estimation valueApproach actual front wheel corner information δ_f, consider that vehicle turns simultaneously To the changing factor of the systematic parameter function influence to system, then formula (6) then becomes

$\stackrel{\·}{x}=(A+\mathrm{\Δ}A)x+(B+\mathrm{\Δ}B)u+(C+\mathrm{\Δ}C){\widehat{\mathrm{\δ}}}_f---\left(7\right);$

In formula:WithSystem when representing system parameter variations respectively System matrix, the changing value that control input matrix is the most corresponding with front wheel angle input matrix；

Formula (7) can arrange further:

$\stackrel{\·}{x}=Ax+Bu+C{\widehat{\mathrm{\δ}}}_f+d\left(t\right)---\left(8\right)$

In formulad₁(t)、d₂T () represents when vehicle parameter changes respectively, The changing value that side slip angle is corresponding with yaw velocity；

. use following preferable auto model:

${\stackrel{\·}{x}}_d=A_d{x}_d+C_d{\widehat{\mathrm{\δ}}}_f---\left(9\right);$

In formula: the state vector of ideal modelWherein β_d、γ_dIt is respectively expectation side slip angle and phase Hope yaw velocity；The sytem matrix of ideal modelInput matrixIts Middle coefficient k_γAnd τ_γBeing proportional gain and the lag time constant of first-order lag link respectively, expression formula is as follows:

$k_{\mathrm{\γ}}=\frac{k_0(b_{11}{a}_{21}-a_{11}{b}_{21})+(c_1{a}_{21}-c_2{a}_{11})}{(a_{11}{a}_{22}-a_{12}{a}_{21})};k_0=\frac{c_1{a}_{22}-c_2{a}_{12}}{a_{12}{b}_{21}-a_{22}{b}_{11}};{\mathrm{\τ}}_{\mathrm{\γ}}=\frac{k_{\mathrm{\γ}}}{k_0{b}_{21}+c_2};$

Formula (9) is the expression formula of preferable vehicle steering model.

Preferably, in described step A, the specific configuration process of front wheel angle observer is as follows:

Set front wheel angle and estimate initial valueFront wheel angle observer expression formula is as follows:

${\widehat{\mathrm{\δ}}}_f\left(t\right)={\∫}_0^t{e}^T(C-C_d)d\mathrm{\τ}---\left(10\right);$

In formula: e is the error vector between the actual side slip angle of automobile and yaw velocity and ideal model state, express Formula is as follows:

$e=x-x_d=\left[\begin{array}{c}{e}_1\\ e_2\end{array}\right]=\left[\begin{array}{c}{e}_{\mathrm{\β}}\\ e_{\mathrm{\γ}}\end{array}\right]=\left[\begin{array}{c}\mathrm{\β}-{\mathrm{\β}}_d\\ \mathrm{\γ}-{\mathrm{\γ}}_d\end{array}\right]=\left[\begin{array}{c}\mathrm{\β}\\ \mathrm{\γ}-{\mathrm{\γ}}_d\end{array}\right]---\left(11\right);$

e_β、e_γIt is respectively side slip angle and controls error and yaw velocity control error；

The ART network rule understanding front wheel angle according to formula (10) is

${\stackrel{\·}{\widehat{\mathrm{\δ}}}}_f=e^T(C-C_d)---\left(12\right);$

Control error equation is derived according to formula (8) and formula (9):

$\stackrel{\·}{e}=Ae+(A-A_d)x_d+Bu+(C-C_d){\widehat{\mathrm{\δ}}}_f+d\left(t\right)---\left(13\right).$

Preferably, described step A is disturbed border estimate that the specific configuration process of link is as follows:

The ART network rule on definition interference border is as follows:

$\left\{\begin{array}{c}{\stackrel{\·}{\widehat{\mathrm{\ψ}}}}_1={\mathrm{\ϵ}}_1{e}_{\mathrm{\β}}\mathrm{sgn}\left(e_{\mathrm{\β}}\right)={\mathrm{\ϵ}}_1\left|e_{\mathrm{\β}}\right|\\ {\stackrel{\·}{\widehat{\mathrm{\ψ}}}}_2={\mathrm{\ϵ}}_2{e}_{\mathrm{\γ}}\mathrm{sgn}\left(e_{\mathrm{\γ}}\right)={\mathrm{\ϵ}}_2\left|e_{\mathrm{\γ}}\right|\end{array}\right.---\left(14\right);$

In formula: sgn (.) represents symbol switch function；Represent interference boundary parameter ψ respectively₁And ψ₂Estimated value； ε₁、ε₂It is called the estimation coefficient on interference border, and is all higher than 1；

Assuming that wheel steering initial timeWithThe mathematic(al) representation of link is estimated on interference border As follows:

$\left\{\begin{array}{c}{\widehat{\mathrm{\ψ}}}_1\left(t\right)={\∫}_0^t{\mathrm{\ϵ}}_1\left|e_{\mathrm{\β}}\right|d\mathrm{\τ}\\ {\widehat{\mathrm{\ψ}}}_2\left(t\right)={\∫}_0^t{\mathrm{\ϵ}}_2\left|e_{\mathrm{\γ}}\right|d\mathrm{\τ}\end{array}\right.---\left(15\right);$

Estimate to draw according to formula (15).

Preferably, trailing wheel corner sliding mode controller and the specific configuration of yaw moment sliding mode controller in described step A Process is as follows:

Definition sliding-mode surface function s=e, sliding mode controllerWherein, after sliding mode controller u comprises simultaneously Wheel corner sliding mode controller and yaw moment sliding mode controller, and wheel corner δ later_rWith yaw moment M as controlled quentity controlled variable, u_eqFor Sliding formwork equivalent controller, u_sFor switch controller；Ignore shock wave d (t) caused by systematic parameter, according toAnd Utilize formula (15) that sliding formwork equivalent controller u can be derived_eqExpression formula as follows:

$u_{eq}=-B^{-1}\[-Ke-Ae-(A-A_d)x_d-(C-C_d){\widehat{\mathrm{\δ}}}_f\]---\left(16\right);$

In formula: K is control gain matrix undetermined,k₁And k₂It is all higher than zero, wherein Diag (.) represents diagonal matrix；

Switch controller u_sExpression formula as follows:

$u_s=-B^{-1}diag({\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1,{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2)sgn\left(e\right)---\left(17\right);$

In formula:For switching control u_sIn control gain；

Expression formula according to the available sliding mode controller of formula (16) and (17) is as follows:

$u=\left[\begin{array}{c}{\mathrm{\δ}}_r\\ M\end{array}\right]=u_{eq}+u_s=-B^{-1}\[Ke+Ae+(A-A_d)x_d+(C-C_d){\widehat{\mathrm{\δ}}}_f+diag({\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1,{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2)sgn\left(e\right)\]---\left(18\right).$

Preferably, described step E particularly as follows:

By matrix A, A_d、B、C、C_dSubstitute into formula (18) with the element of K, obtain trailing wheel corner sliding mode controller by arrangement Concrete form is as follows:

$\begin{array}{c}{\mathrm{\δ}}_r=-\{\[(k_1+a_{11})b_{22}-a_{21}{b}_{12}\]e_{\mathrm{\β}}+\[a_{12}{b}_{22}-b_{12}(k_2+a_{22})\]e_{\mathrm{\γ}}+\[a_{12}{b}_{22}-b_{12}(a_{22}-a_{22d})\]{\mathrm{\γ}}_d\\ +\[b_{22}{c}_1-b_{12}(c_2-c_{2d})\]{\widehat{\mathrm{\δ}}}_f+b_{22}{\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1\mathrm{sgn}\left(e_1\right)-b_{12}{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2\mathrm{sgn}\left(e_2\right)\}/(b_{11}{b}_{22}-b_{12}{b}_{21})\end{array}---\left(19\right);$

The concrete form of yaw moment sliding mode controller is:

$\begin{array}{c}M=\{\[(k_1+a_{11})b_{21}-a_{21}{b}_{11}\]e_{\mathrm{\β}}+\[a_{12}{b}_{21}-b_{11}(k_2+a_{22})\]e_{\mathrm{\γ}}+\[a_{12}{b}_{21}-b_{11}(a_{22}-a_{22d})\]{\mathrm{\γ}}_d\\ +\[b_{21}{c}_1-b_{11}(c_2-c_{2d})\]{\widehat{\mathrm{\δ}}}_f+b_{21}{\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1\mathrm{sgn}\left(e_1\right)-b_{11}{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2\mathrm{sgn}\left(e_2\right)\}/(b_{11}{b}_{22}-b_{12}{b}_{21})\end{array}---\left(20\right);$

Use trailing wheel corner obtained above and yaw moment that vehicle is controlled in real time.

Four-wheel independent steering control method for vehicle of the present invention passes through the front wheel angle observer of independent innovational design, interference edge Boundary estimates that the combination of link, trailing wheel corner sliding mode controller and yaw moment control device controls, and is on the one hand seen by front wheel angle Survey device to realize actual front wheel corner is estimated the most accurately, on the other hand make automobile side slip angle and yaw velocity with Error between ideal model correspondence output is the least, allows vehicle obtain good tracing control characteristic, with satisfied traveling The stability requirement of state, it is to avoid control power is lost due to vehicle front wheel angle sensor fault, reduces vehicle Stability；Likewise it is preferred that scheme also includes the knot of yaw moment sliding mode controller, trailing wheel controlling angle and yaw moment control Closing, this makes the present invention program control to be better than in effect the control method of single mode, and on the one hand complex controll can guarantee that relatively Good corner accuracy of observation, on the other hand can obtain under the limiting condition such as vehicle high-speed, sharp turn and preferably control effect； Further, the design of the switch controller in the present invention program can suppress or reduce disturbance that mini system Parameters variation brings to control The impact of performance, improves the control robustness of motor turning control stability.

Accompanying drawing explanation

The flow chart of the control method of the four-wheel independent steering vehicle that Fig. 1 provides for the present invention

The control structure schematic diagram of the four-wheel independent steering vehicle that Fig. 2 (a) provides for the present invention

Fig. 2 (b) is the sliding formwork control structure schematic diagram having front wheel angle sensor described in the embodiment of the present invention 1

Fig. 3 is the angle step waveform figure of vehicle front-wheel actual steering

Fig. 4 is the angle sinusoidal wave form figure of vehicle front-wheel actual steering

Fig. 5 (a) is to be applied with front wheel angle sensor sliding formwork to control (being abbreviated as: have sensor SMC), automobile speed Side slip angle when 30km/h, front-wheel are turned to by angle step waveform controls oscillogram.

Fig. 5 (b) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 100km/h, front-wheel by angle step waveform Side slip angle when turning to controls oscillogram.

Fig. 5 (c) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 30km/h, front-wheel by angle sinusoidal wave form Side slip angle when turning to controls oscillogram.

Fig. 5 (d) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 100km/h, front-wheel by angle sinusoidal wave form Side slip angle when turning to controls oscillogram.

Fig. 6 (a) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 30km/h, front-wheel by angle step waveform Yaw velocity when turning to controls oscillogram.

Fig. 6 (b) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 100km/h, front-wheel by angle step waveform Yaw velocity when turning to controls oscillogram.

Fig. 6 (c) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 30km/h, front-wheel by angle sinusoidal wave form Yaw velocity when turning to controls oscillogram.

Fig. 6 (d) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 100km/h, front-wheel by angle sinusoidal wave form Yaw velocity when turning to controls oscillogram.

Fig. 7 (a) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 30km/h, front-wheel by angle step waveform Speed change curve when turning to.

Fig. 7 (b) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 100km/h, front-wheel by angle step waveform Speed change curve when turning to.

Fig. 7 (c) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 30km/h, front-wheel by angle sinusoidal wave form Speed change curve when turning to.

Fig. 7 (d) is to be applied with the control of front wheel angle sensor sliding formwork, automobile speed 100km/h, front-wheel by angle sinusoidal wave form Speed change curve when turning to.

Fig. 8 (a) is that the unmatched rotary angle transmitter sliding formwork of taking turns applying the present embodiment controls (being abbreviated as: without sensor SMC), vapour Side slip angle when car speed 30km/h, front-wheel are turned to by angle step waveform controls oscillogram.

Fig. 8 (b) be apply the present embodiment unmatched take turns the control of rotary angle transmitter sliding formwork, automobile speed 100km/h, front-wheel by Side slip angle when angle step waveform turns to controls oscillogram.

Fig. 8 (c) be apply the present embodiment unmatched take turns the control of rotary angle transmitter sliding formwork, automobile speed 30km/h, front-wheel by Side slip angle when angle sinusoidal wave form turns to controls oscillogram.

Fig. 8 (d) be apply the present embodiment unmatched take turns the control of rotary angle transmitter sliding formwork, automobile speed 100km/h, front-wheel by Side slip angle when angle sinusoidal wave form turns to controls oscillogram.

Fig. 9 (a) be apply the present embodiment unmatched take turns the control of rotary angle transmitter sliding formwork, automobile speed 30km/h, front-wheel by Yaw velocity when angle step waveform turns to controls oscillogram.

Fig. 9 (b) be apply the present embodiment unmatched take turns the control of rotary angle transmitter sliding formwork, automobile speed 90km/h, front-wheel by Yaw velocity when angle step waveform turns to controls oscillogram.

Fig. 9 (c) be apply the present embodiment unmatched take turns the control of rotary angle transmitter sliding formwork, automobile speed 30km/h, front-wheel by Yaw velocity when angle sinusoidal wave form turns to controls oscillogram.

Fig. 9 (d) be apply the present embodiment unmatched take turns the control of rotary angle transmitter sliding formwork, automobile speed 90km/h, front-wheel by Yaw velocity when angle sinusoidal wave form turns to controls oscillogram.

Figure 10 (a) is to apply the unmatched sliding formwork control of rotary angle transmitter, automobile speed 30km/h, the front-wheel of taking turns by angle step wave Speed change curve when shape turns to.

Figure 10 (b) is to apply the unmatched of the present embodiment to take turns the control of rotary angle transmitter sliding formwork, automobile speed 100km/h, front-wheel Front wheel angle curve chart when turning to by angle step waveform.

Figure 10 (c) be apply the present embodiment unmatched take turns the control of rotary angle transmitter sliding formwork, automobile speed 30km/h, front-wheel by Speed change curve when angle sinusoidal wave form turns to.

Figure 10 (d) is to apply the unmatched of the present embodiment to take turns the control of rotary angle transmitter sliding formwork, automobile speed 100km/h, front-wheel Speed change curve when turning to by angle sinusoidal wave form.

Figure 11 (a) is to apply the unmatched control of rotary angle transmitter sliding formwork, automobile speed 30km/h, the front-wheel of taking turns by angle step waveform Front wheel angle observer output front wheel angle estimation curve figure when turning to.

Figure 11 (b) is to apply the unmatched control of rotary angle transmitter sliding formwork, automobile speed 100km/h, the front-wheel of taking turns by angle step wave Front wheel angle observer output front wheel angle estimation curve figure when shape turns to.

Figure 11 (c) is to apply the unmatched control of rotary angle transmitter sliding formwork, automobile speed 30km/h, the front-wheel of taking turns by angle sinusoidal wave form Front wheel angle observer output front wheel angle estimation curve figure when turning to.

Figure 11 (d) is to apply the unmatched control of rotary angle transmitter sliding formwork, automobile speed 100km/h, the front-wheel of taking turns by angle sine wave Front wheel angle observer output front wheel angle estimation curve figure when shape turns to.

Detailed description of the invention

The present invention is illustrated below in conjunction with the accompanying drawings with embodiment.

Embodiment 1

As it is shown in figure 1, the control method of the four-wheel independent steering vehicle of the present embodiment offer comprises the following steps:

A, default preferable vehicle steering model, front wheel angle estimate that initial value, front wheel angle observer, interference border are estimated Link, trailing wheel corner sliding mode controller and yaw moment sliding mode controller；

The construction process of described preferable vehicle steering model is as follows:

Set up following Vehicular turn kinematics model:

$\left\{\begin{array}{c}m({\stackrel{\·}{v}}_x-v_y\mathrm{\γ})=(F_{x1}+F_{x2}){\mathrm{cos\δ}}_f-(F_{y1}+F_{y2}){\mathrm{sin\δ}}_f+(F_{x3}+F_{x4}){\mathrm{cos\δ}}_r-(F_{y3}+F_{y4}){\mathrm{sin\δ}}_r\\ m({\stackrel{\·}{v}}_y+v_x\mathrm{\γ})=(F_{x1}+F_{x2}){\mathrm{sin\δ}}_f+(F_{y1}+F_{y2}){\mathrm{cos\δ}}_f+(F_{x3}+F_{x4}){\mathrm{sin\δ}}_r+(F_{y3}+F_{y4}){\mathrm{cos\δ}}_r\\ I_z\stackrel{\·}{\mathrm{\γ}}=a(F_{y1}+F_{y2}){\mathrm{cos\δ}}_f-b(F_{y3}+F_{y4}){\mathrm{cos\δ}}_r+0.5W\[(F_{y1}+F_{y2}){\mathrm{sin\δ}}_f+(F_{y3}-F_{y4}){\mathrm{sin\δ}}_r\]+M\\ J_{wi}{\stackrel{\·}{\mathrm{\ω}}}_i=M_{di}-F_{xi}R-M_{bi}(i=1,2,3,4)\end{array}\right.---\left(1\right);$

In formula: m is complete vehicle quality；v_x、v_yRepresent automobile systemic velocity V velocity component on x and y-axis respectively； Represent automobile systemic velocity V component of acceleration on x and y-axis respectively；γ is automobile yaw velocity,Then represent yaw angle Acceleration；A and b is automobile barycenter to front axle and the distance of rear axle, vehicle wheel base L=a+b respectively；F_xi、F_yiRepresent automobile respectively The longitudinal force of tire and cross force, wherein subscript i=1,2,3,4 the most corresponding the near front wheels, off-front wheel, left rear wheel and off hind wheel； δ_f、δ_rIt is front and rear wheel steering angle respectively；I_zFor automobile around the rotary inertia of z-axis；J_wiAnd ω_iIt is respectively the rotary inertia of each tire And rotational angular velocity,Represent the angle of rotation acceleration of each tire；M_diIt it is the output moment of torsion on differential side；R represents tire Radius；M_biBraking moment suffered by tire；W is wheelspan, i.e. front tread B_fWith rear tread B_rIt is equal to W；M represents suffered by wheel Longitudinal force is produced the yaw moment of additional control:

M=a (F_x1+F_x2)sinδ_f-b(F_x3+F_x4)sinδ_r+0.5W[(F_x2-F_x1)cosδ_f+(F_x4-F_x3)cosδ_r] (2)；

Vehicle centroid side drift angle: β=arctan (v_x/v_y)；

The side drift angle α of front and back wheel_i:

$\left\{\begin{array}{c}{\mathrm{\α}}_1={\mathrm{\α}}_2\≈\mathrm{\β}+a\mathrm{\γ}/v_x-{\mathrm{\δ}}_f\\ {\mathrm{\α}}_3={\mathrm{\α}}_4\≈\mathrm{\β}-b\mathrm{\γ}/v_x-{\mathrm{\δ}}_r\end{array}\right.---\left(3\right);$

The wherein the most corresponding the near front wheel of subscript i=1,2,3,4, off-front wheel, left rear wheel and off hind wheel；

Assuming that under the driving cycle that automobile is in the non-emergent state of normal speed per hour scope and low-angle turns to, have v_x≈ V, And only consider defective steering stabilizer and weaving, i.e. select side slip angle and yaw velocity as the measurement master of control stability Want index, convolution (1) and (3) can obtain the kinetics equation of vehicle 2DOF linear single track model:

$\left\{\begin{array}{c}mu(\stackrel{\·}{\mathrm{\β}}+\mathrm{\γ})=F_{y1}+F_{y2}+F_{y3}+F_{y4}\\ I_z\stackrel{\·}{\mathrm{\γ}}=a(F_{y1}+F_{y2})-b(F_{y3}+F_{y4})+M\end{array}\right.---\left(4\right);$

In formula: F_y1+F_y2、F_y3+F_y4Represent the lateral deviation power of axle tire respectively

$\left\{\begin{array}{c}{F}_{y1}+F_{y2}=-k_f(\mathrm{\β}+a\mathrm{\γ}/V-{\mathrm{\δ}}_f)\\ F_{y3}+F_{y4}=-k_r(\mathrm{\β}-b\mathrm{\γ}/V-{\mathrm{\δ}}_r)\end{array}\right.---\left(5\right);$

Wherein k_fAnd k_rIt is respectively the comprehensive cornering stiffness of front axle both sides tire, the comprehensive cornering stiffness of rear axle both sides tire, Its numerical value is 2 times of front and rear wheel cornering stiffness；

Definition system state vector x=[β, γ]^TWith control input vector u=[δ_r,M]^T, set up according to formula (4) and (5) Following model state space equation is:

$\stackrel{\·}{x}=Ax+Bu+{\mathrm{C\δ}}_f---\left(6\right);$

In formula:ForSytem matrix

Control input matrixFront wheel angle input matrix

Use front wheel angle observer output estimation valueApproach actual front wheel corner information δ_f, consider that vehicle turns simultaneously To the changing factor of the systematic parameter function influence to system, then formula (6) then becomes

$\stackrel{\·}{x}=(A+\mathrm{\Δ}A)x+(B+\mathrm{\Δ}B)u+(C+\mathrm{\Δ}C){\widehat{\mathrm{\δ}}}_f---\left(7\right);$

In formula:WithSystem when representing system parameter variations respectively System matrix, the changing value that control input matrix is the most corresponding with front wheel angle input matrix；

Formula (7) can arrange further:

$\stackrel{\·}{x}=Ax+Bu+C{\widehat{\mathrm{\δ}}}_f+d\left(t\right)---\left(8\right)$

In formulaSide slip angle and horizontal stroke when representing vehicle parameter change The changing value that pivot angle speed is corresponding；

Use following preferable auto model:

${\stackrel{\·}{x}}_d=A_d{x}_d+C_d{\widehat{\mathrm{\δ}}}_f---\left(9\right);$

In formula: the state vector of ideal modelWherein β_d、γ_dIt is respectively expectation side slip angle and phase Hope yaw velocity；The sytem matrix of ideal modelInput matrixIts Middle coefficient k_γAnd τ_γBeing proportional gain and the lag time constant of first-order lag link respectively, expression formula is as follows:

$k_{\mathrm{\γ}}=\frac{k_0(b_{11}{a}_{21}-a_{11}{b}_{21})+(c_1{a}_{21}-c_2{a}_{11})}{(a_{11}{a}_{22}-a_{12}{a}_{21})};k_0=\frac{c_1{a}_{22}-c_2{a}_{12}}{a_{12}{b}_{21}-a_{22}{b}_{11}};{\mathrm{\τ}}_{\mathrm{\γ}}=\frac{k_{\mathrm{\γ}}}{k_0{b}_{21}+c_2};$

Formula (9) is the expression formula of preferable vehicle steering model；

In described step A, the specific configuration process of front wheel angle observer is as follows:

Set front wheel angle and estimate initial valueFront wheel angle observer expression formula is as follows:

${\widehat{\mathrm{\δ}}}_f\left(t\right)={\∫}_0^t{e}^T(C-C_d)d\mathrm{\τ}---\left(10\right);$

In formula: e is the error vector between the actual side slip angle of automobile and yaw velocity and ideal model state, express Formula is as follows:

$e=x-x_d=\left[\begin{array}{c}{e}_1\\ e_2\end{array}\right]=\left[\begin{array}{c}{e}_{\mathrm{\β}}\\ e_{\mathrm{\γ}}\end{array}\right]=\left[\begin{array}{c}\mathrm{\β}-{\mathrm{\β}}_d\\ \mathrm{\γ}-{\mathrm{\γ}}_d\end{array}\right]=\left[\begin{array}{c}\mathrm{\β}\\ \mathrm{\γ}-{\mathrm{\γ}}_d\end{array}\right]---\left(11\right);$

e_β、e_γIt is respectively side slip angle and controls error and yaw velocity control error；

The ART network rule understanding front wheel angle according to formula (10) is

${\stackrel{\·}{\widehat{\mathrm{\δ}}}}_f=e^T(C-C_d)---\left(12\right);$

Control error equation is derived according to formula (8) and formula (9):

$\stackrel{\·}{e}=Ae+(A-A_d)x_d+Bu+(C-C_d){\widehat{\mathrm{\δ}}}_f+d\left(t\right)---\left(13\right);$

Described interference border estimates that the specific configuration process of link is as follows:

The ART network rule on definition interference border is as follows:

$\left\{\begin{array}{c}{\stackrel{\·}{\widehat{\mathrm{\ψ}}}}_1={\mathrm{\ϵ}}_1{e}_{\mathrm{\β}}\mathrm{sgn}\left(e_{\mathrm{\β}}\right)={\mathrm{\ϵ}}_1\left|e_{\mathrm{\β}}\right|\\ {\stackrel{\·}{\widehat{\mathrm{\ψ}}}}_2={\mathrm{\ϵ}}_2{e}_{\mathrm{\γ}}\mathrm{sgn}\left(e_{\mathrm{\γ}}\right)={\mathrm{\ϵ}}_2\left|e_{\mathrm{\γ}}\right|\end{array}\right.---\left(14\right);$

In formula: sgn (.) represents symbol switch function；Represent interference boundary parameter ψ respectively₁And ψ₂Estimated value； ε₁、ε₂It is called the estimation coefficient on interference border, and is all higher than 1；

Assuming that wheel steering initial timeWithThe mathematic(al) representation of link is estimated on interference border As follows:

$\left\{\begin{array}{c}{\widehat{\mathrm{\ψ}}}_1\left(t\right)={\∫}_0^t{\mathrm{\ϵ}}_1\left|e_{\mathrm{\β}}\right|d\mathrm{\τ}\\ {\widehat{\mathrm{\ψ}}}_2\left(t\right)={\∫}_0^t{\mathrm{\ϵ}}_2\left|e_{\mathrm{\γ}}\right|d\mathrm{\τ}\end{array}\right.---\left(15\right);$

Estimate to draw according to formula (15)；

Described trailing wheel corner sliding mode controller and the specific configuration process of yaw moment sliding mode controller are as follows:

Definition sliding-mode surface function s=e, sliding mode controllerWherein, after sliding mode controller u comprises simultaneously Wheel corner sliding mode controller and yaw moment sliding mode controller, and wheel corner δ later_rWith yaw moment M as controlled quentity controlled variable, u_eqFor Sliding formwork equivalent controller, u_sFor switch controller；Ignore shock wave d (t) caused by systematic parameter, according toAnd Utilize formula (15) that sliding formwork equivalent controller u can be derived_eqExpression formula as follows:

$u_{eq}=-B^{-1}\[-Ke-Ae-(A-A_d)x_d-(C-C_d){\widehat{\mathrm{\δ}}}_f\]---\left(16\right);$

In formula: K is control gain matrix undetermined,k₁And k₂It is all higher than zero, wherein Diag (.) represents diagonal matrix；

Switch controller u_sExpression formula as follows:

$u_s=-B^{-1}diag({\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1,{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2)sgn\left(e\right)---\left(17\right);$

In formula:For switching control u_sIn control gain；

Expression formula according to the available sliding mode controller of formula (16) and (17) is as follows:

$u=\left[\begin{array}{c}{\mathrm{\δ}}_r\\ M\end{array}\right]=u_{eq}+u_s=-B^{-1}\[Ke+Ae+(A-A_d)x_d+(C-C_d){\widehat{\mathrm{\δ}}}_f+diag({\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1,{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2)sgn\left(e\right)\]---\left(18\right);$

B, using vehicle craspedodrome state as initial time, measure side slip angle and the yaw velocity of vehicle in real time, with Time front wheel angle estimated the preferable vehicle steering model of initial value input, initially expected side slip angle and initially expected horizontal stroke Pivot angle speed；By initially expect side slip angle and initial expectation yaw velocity respectively with the real-time barycenter lateral deviation in corresponding moment Angle compares with yaw velocity, and the side slip angle respectively obtaining the corresponding moment controls error and yaw velocity control Error；Side slip angle controls error, the input of initial horizontal pivot angle speed controlling error obtains correspondence to front wheel angle observer The front wheel angle estimated value in moment；

C, carry out the calculating of front wheel angle estimated value in real time, the real-time preferable vehicle of front wheel angle estimated value input is turned To model, obtain real-time expectation side slip angle and expectation yaw velocity, by real-time expectation side slip angle and expectation Yaw velocity compares with real-time side slip angle, yaw velocity, thus obtains real-time side slip angle and control Error, yaw velocity control error；

D, real-time side slip angle is controlled error, yaw velocity controls error input nonlinearities border and estimates link, Obtain real-time interference boundary parameter；

E, by real-time front wheel angle estimated value, side slip angle controls error, yaw velocity controls error and dry Disturbing boundary parameter and be separately input to trailing wheel corner sliding mode controller and yaw moment sliding mode controller, output obtains real-time respectively Trailing wheel corner and yaw moment, and use real-time trailing wheel corner and yaw moment that vehicle is controlled in real time；

Particularly as follows:

By matrix A, A_d、B、C、C_dSubstitute into formula (18) with the element of K, obtain trailing wheel corner sliding mode controller by arrangement Concrete form is as follows:

$\begin{array}{c}{\mathrm{\δ}}_r=-\{\[(k_1+a_{11})b_{22}-a_{21}{b}_{12}\]e_{\mathrm{\β}}+\[a_{12}{b}_{22}-b_{12}(k_2+a_{22})\]e_{\mathrm{\γ}}+\[a_{12}{b}_{22}-b_{12}(a_{22}-a_{22d})\]{\mathrm{\γ}}_d\\ +\[b_{22}{c}_1-b_{12}(c_2-c_{2d})\]{\widehat{\mathrm{\δ}}}_f+b_{22}{\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1\mathrm{sgn}\left(e_1\right)-b_{12}{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2\mathrm{sgn}\left(e_2\right)\}/(b_{11}{b}_{22}-b_{12}{b}_{21})\end{array}---\left(19\right);$

The concrete form of yaw moment sliding mode controller is:

$\begin{array}{c}M=\{\[(k_1+a_{11})b_{21}-a_{21}{b}_{11}\]e_{\mathrm{\β}}+\[a_{12}{b}_{21}-b_{11}(k_2+a_{22})\]e_{\mathrm{\γ}}+\[a_{12}{b}_{21}-b_{11}(a_{22}-a_{22d})\]{\mathrm{\γ}}_d\\ +\[b_{21}{c}_1-b_{11}(c_2-c_{2d})\]{\widehat{\mathrm{\δ}}}_f+b_{21}{\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1\mathrm{sgn}\left(e_1\right)-b_{11}{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2\mathrm{sgn}\left(e_2\right)\}/(b_{11}{b}_{22}-b_{12}{b}_{21})\end{array}---\left(20\right);$

Use trailing wheel corner obtained above and yaw moment that vehicle is controlled in real time.

The control structure schematic diagram of the four-wheel independent steering vehicle that Fig. 2 (a) provides for the present invention；The present embodiment uses table 1 In parameter be simulated, the present embodiment unmatched is taken turns the sliding-mode control of rotary angle transmitter (referred to as without sensor cunning Mould control method) with sliding-mode control (referred to as having sensor sliding-mode control) and the nothing having front wheel angle sensor 3 kinds of situations of vehicle (referred to as FWS vehicle) that sliding formwork controls carry out contrast experiment；

Table 1 vehicle and control parameter

Title	Numerical value	Title	Numerical value
				Complete vehicle quality m/kg	1479	Barycenter is to front axle distance a/m	1.058
Rotary inertia Iz/(kg.m2)	2731	Barycenter is to rear axle distance b/m	1.756
				Front-wheel comprehensive cornering stiffness kf/(N.rad-1)	115600	Axletree is away from L/m	2.814
Trailing wheel comprehensive cornering stiffness kr/(N.rad-1)	119600	Tire rolling radius R/m	0.3075
				Wheelspan W/m	1.55	Vehicle wheel rotation inertia Jw/(kg.m2)	1.25
Coefficient of road adhesion μ	0.8	Control gain matrix K	Diag (900,500)
				Interference border estimation coefficient ε1、ε2	10

Considering the automobile running working condition that vehicle turns in different speeds and different wave, wherein, speed operating mode is: 30km/h (8.333m/s)、100km/h(27.778m/s)；The waveform operating mode that waveform turns to is: non-ideal angle step waveform, angle sine wave Shape (S-shaped)；By speed operating mode and waveform operating mode combination of two, form 4 kinds of composite conditions；Wherein set non-ideal angle step waveform 0s starts to jump, and jump rise time and amplitude are respectively 0.5s and 0.07rad；Angle sinusoidal wave form initial time 2s is set, Cycle, angular amplitude are respectively 4s and 0.07rad, and Fig. 3, Fig. 4 respectively illustrate angle step waveform and angle sinusoidal wave form；

In view of the quality in vehicle parameter and rotary inertia easily change, therefore suppose in table 1 in contrast test is whole Car quality and rotary inertia all increase+15%；

(a, b, c, d) (a, b, c d) have respectively illustrated under sensor sliding formwork control condition, various combination work-Fig. 7 Fig. 5 The time-domain response curve of side slip angle, yaw velocity and speed under condition, the output that sensor sliding formwork will be had to control respectively Expect with reality and contrast without control situation；(a, b, c, d) (a, b, c d) then sets forth the present embodiment to-Figure 10 to Fig. 8 Under the control condition of sliding formwork without sensor, the time domain of side slip angle, yaw velocity and speed under the conditions of each driving cycle is rung Answer curve；(a, b, c, in the case of d) reflecting without the control of sensor sliding formwork, vehicle difference driving cycle, observer export Figure 11 Front wheel angle estimate waveform.

By comparison diagram 5 (a, b, c, d) and Fig. 8 (a, b, c, d) visible, for uncontrolled FWS vehicle, side slip angle Steady-state response non-zero, and numerical value during high speed is relatively big and contrary with front wheel angle input direction, which increase the whipping of vehicle with Sideslip trend；4WS vehicle is under having sensor or trailing wheel corner and yaw moment sliding formwork control condition without sensor, though car Different speeds and different wave turn to, and all can realize vehicle centroid side drift angle is zero, reaches preferably to expect steady statue, makes Obtain 4WS vehicle and can maintain body gesture well, there is good path trace ability, significantly improve the manipulation of vehicle Property.

(a, b, c, d) (a, b, c, d) find out comparison diagram 6, and during low speed, 4WS vehicle is having sensor or without sensor with Fig. 9 Sliding formwork control condition under, yaw velocity all can obtain stability contorting, its numerical value be more than uncontrolled FWS vehicle, this shows By controlling so that 4WS vehicle is fewer than FWS vehicle beats steering wheel, is effectively reduced radius of turn, improves turn inside diameter Maneuverability.During high-speed cruising, there is the biggest overshoot in FWS yaw rate, and produces the vibration fluctuation of higher magnitude Phenomenon, this reflects the unstability that vehicle travels.Under two kinds of control method operative condition, the yaw velocity of 4WS vehicle Both less than FWS vehicle, and oscillatory occurences is substantially inhibited, particularly when step waveform turns to, and yaw velocity non-overshoot And oscillation phenomenon, the stability of the 4WS vehicle that this not only shows is improved, it is to avoid and reduce driving under high-speed travel state Member hits the danger that steering wheel causes.

(a, b, c, d) (a, b, c, d) visible, speed is that the low speed of 30km/h (8.333m/s) turns to comparison diagram 7 with Figure 10 Time, the speed in the case of having sensor or the sliding formwork without sensor to control all than has declined during without controlling, but it is equal to decline degree Less；When speed is the high speed steering of 100km/h (27.778m/s), controlled 4WS vehicle speed keeps effect to be better than without control The FWS vehicle of system.This shows, sliding mode control strategy is ensureing that vehicle obtains preferable turning path and follows the tracks of ability and vehicle body stability While, it is the most little that speed reduces degree, and this can make 4WS vehicle keep bigger speed to carry out turning according to target trajectory safely Curved traveling.

Comprehensive Correlation has front wheel angle sensor or the unmatched sliding formwork taken turns under the conditions of rotary angle transmitter to control effect, in difference Under speed driving cycle, even if vehicle parameter changes, the present embodiment controls without sensor sliding formwork also can be preferably to ideal The estimation expectation that Controlling model produces carries out tenacious tracking control, can obtain the control almost identical with there being sensor sliding formwork control The impact of performance.Meanwhile, according to Figure 11 (a, b, c, d) find out, in the case of different speeds and difference turn to, front wheel angle observer Output can preferably approach actual front wheel corner configurations and amplitude, and this also indicates that unmatched under the conditions of taking turns rotary angle transmitter, this enforcement The sliding-mode control of the front wheel angle observer of example is effective and feasible.

Claims (6)

1. the control method of a four-wheel independent steering vehicle, it is characterised in that comprise the following steps:

A, default preferable vehicle steering model, front wheel angle estimate that ring is estimated on initial value, front wheel angle observer, interference border Joint, trailing wheel corner sliding mode controller and yaw moment sliding mode controller；

B, using vehicle craspedodrome state as initial time, measure side slip angle and the yaw velocity of vehicle in real time, will simultaneously Front wheel angle estimates the preferable vehicle steering model of initial value input, is initially expected side slip angle and initially expects yaw angle Speed；By initially expect side slip angle and initial expectation yaw velocity respectively with the real-time side slip angle in corresponding moment and Yaw velocity compares, and respectively obtains side slip angle control error and the yaw velocity in corresponding moment and controls by mistake Difference；Side slip angle is controlled error, initial horizontal pivot angle speed controlling error input to front wheel angle observer obtain to correspondence time The front wheel angle estimated value carved；

C, carry out the calculating of front wheel angle estimated value in real time, by the real-time preferable Vehicular turn mould of front wheel angle estimated value input Type, obtains real-time expectation side slip angle and expectation yaw velocity, by real-time expectation side slip angle and expectation yaw Angular velocity compares with real-time side slip angle, yaw velocity, thus obtain real-time side slip angle control error, Yaw velocity controls error；

D, real-time side slip angle is controlled error, yaw velocity controls error input nonlinearities border and estimates link, obtain Real-time interference boundary parameter；

E, by real-time front wheel angle estimated value, side slip angle controls error, yaw velocity controls error and interference edge Boundary's parameter is input to trailing wheel corner sliding mode controller and yaw moment sliding mode controller jointly, and output obtains real-time trailing wheel respectively Corner and yaw moment, and use real-time trailing wheel corner and yaw moment that vehicle is controlled in real time.

2. the control method of four-wheel independent steering vehicle as claimed in claim 1, it is characterised in that:

In described step A, the construction process of preferable vehicle steering model is as follows:

Set up following Vehicular turn kinematics model:

$\left\{\begin{array}{c}m({\stackrel{\·}{v}}_x-v_y\mathrm{\γ})=(F_{x1}+F_{x2}){\mathrm{cos\δ}}_f-(F_{y1}+F_{y2}){\mathrm{sin\δ}}_f+(F_{x3}+F_{x4}){\mathrm{cos\δ}}_r-(F_{y3}+F_{y4}){\mathrm{sin\δ}}_r\\ m({\stackrel{\·}{v}}_y+v_x\mathrm{\γ})=(F_{x1}+F_{x2}){\mathrm{sin\δ}}_f+(F_{y1}+F_{y2}){\mathrm{cos\δ}}_f+(F_{x3}+F_{x4}){\mathrm{sin\δ}}_r+(F_{y3}+F_{y4}){\mathrm{cos\δ}}_r\\ I_z\stackrel{\·}{\mathrm{\γ}}=a(F_{y1}+F_{y2}){\mathrm{cos\δ}}_f-b(F_{y3}+F_{y4}){\mathrm{cos\δ}}_r+0.5W\[(F_{y1}+F_{y2}){\mathrm{sin\δ}}_f+(F_{y3}-F_{y4}){\mathrm{sin\δ}}_r\]+M\\ J_{wi}{\stackrel{\·}{\mathrm{\ω}}}_i=M_{di}-F_{xi}R-M_{bi},(i=1,2,3,4)\end{array}\right.---\left(1\right);$

In formula: m is complete vehicle quality；v_x、v_yRepresent automobile systemic velocity V velocity component on x and y-axis respectively； Table respectively Show automobile systemic velocity V component of acceleration on x and y-axis；γ is automobile yaw velocity,Then represent yaw angle acceleration； A and b is automobile barycenter to front axle and the distance of rear axle, vehicle wheel base L=a+b respectively；F_xi、F_yiRepresent automobile tire respectively Longitudinal force and cross force, wherein subscript i=1,2,3,4 the most corresponding the near front wheels, off-front wheel, left rear wheel and off hind wheel；δ_f、δ_rPoint It it not front and rear wheel steering angle；I_zFor automobile around the rotary inertia of z-axis；J_wiAnd ω_iIt is respectively rotary inertia and the rotation of each tire Angular velocity,Represent the angle of rotation acceleration of each tire；M_diIt it is the output moment of torsion on differential side；R represents tire radius； M_biBraking moment suffered by tire；W is wheelspan, i.e. front tread B_fWith rear tread B_rIt is equal to W；M represents longitudinal suffered by wheel Power is produced the yaw moment of additional control:

M=a (F_x1+F_x2)sinδ_f-b(F_x3+F_x4)sinδ_r+0.5W[(F_x2-F_x1)cosδ_f+(F_x4-F_x3)cosδ_r] (2)；

Vehicle centroid side drift angle: β=arctan (v_x/v_y)；

The side drift angle α of front and back wheel_i:

$\left\{\begin{array}{c}{\mathrm{\α}}_1={\mathrm{\α}}_2\≈\mathrm{\β}+a\mathrm{\γ}/v_x-{\mathrm{\δ}}_f\\ {\mathrm{\α}}_3={\mathrm{\α}}_4\≈\mathrm{\β}-b\mathrm{\γ}/v_x-{\mathrm{\δ}}_r\end{array}\right.---\left(3\right);$

The wherein the most corresponding the near front wheel of subscript i=1,2,3,4, off-front wheel, left rear wheel and off hind wheel；

Assuming that under the driving cycle that automobile is in the non-emergent state of normal speed per hour scope and low-angle turns to, have v_x≈ V, and only Consider defective steering stabilizer and weaving, i.e. select side slip angle and yaw velocity to refer mainly to as the measurement of control stability Mark, convolution (1) and (3) can obtain the kinetics equation of vehicle 2DOF linear single track model:

$\left\{\begin{array}{c}mu(\stackrel{\·}{\mathrm{\β}}+\mathrm{\γ})=F_{y1}+F_{y2}+F_{y3}+F_{y4}\\ I_z\stackrel{\·}{\mathrm{\γ}}=a(F_{y1}+F_{y2})-b(F_{y3}+F_{y4})+M\end{array}\right.---\left(4\right);$

In formula: F_y1+F_y2、F_y3+F_y4Represent the lateral deviation power of axle tire respectively

$\left\{\begin{array}{c}{F}_{y1}+F_{y2}=-k_f(\mathrm{\β}+a\mathrm{\γ}/V-{\mathrm{\δ}}_f)\\ F_{y3}+F_{y4}=-k_r(\mathrm{\β}-b\mathrm{\γ}/V-{\mathrm{\δ}}_r)\end{array}\right.---\left(5\right);$

Wherein k_fAnd k_rIt is respectively the comprehensive cornering stiffness of front axle both sides tire, the comprehensive cornering stiffness of rear axle both sides tire, its number Value is 2 times of front and rear wheel cornering stiffness；

Definition system state vector x=[β, γ]^TWith control input vector u=[δ_r,M]^T, set up as follows according to formula (4) and (5) Model state space equation be:

$\stackrel{\·}{x}=Ax+Bu+{\mathrm{C\δ}}_f---\left(6\right);$

In formula:ForSytem matrix

Control input matrixFront wheel angle input matrix

Use front wheel angle observer output estimation valueApproach actual front wheel corner information δ_f, consider vehicle steering system simultaneously The changing factor of the system parameter function influence to system, then formula (6) then becomes

$\stackrel{\·}{x}=(A+\mathrm{\Δ}A)x+(B+\mathrm{\Δ}B)u+(C+\mathrm{\Δ}C){\widehat{\mathrm{\δ}}}_f---\left(7\right);$

In formula:WithSystem square when representing system parameter variations respectively Battle array, the changing value that control input matrix is the most corresponding with front wheel angle input matrix；

Formula (7) can arrange further:

$\stackrel{\·}{x}=Ax+Bu+C{\widehat{\mathrm{\δ}}}_f+d\left(t\right)---\left(8\right)$

In formulad₁(t)、d₂T () represents when vehicle parameter changes respectively, barycenter The changing value that side drift angle is corresponding with yaw velocity；

Use following preferable auto model:

${\stackrel{\·}{x}}_d=A_d{x}_d+C_d{\widehat{\mathrm{\δ}}}_f---\left(9\right);$

In formula: the state vector of ideal modelWherein β_d、γ_dIt is respectively expectation side slip angle and expectation horizontal stroke Pivot angle speed；The sytem matrix of ideal modelInput matrixIt is wherein Number k_γAnd τ_γBeing proportional gain and the lag time constant of first-order lag link respectively, expression formula is as follows:

$k_{\mathrm{\γ}}=\frac{k_0(b_{11}{a}_{21}-a_{11}{b}_{21})+(c_1{a}_{21}-c_2{a}_{11})}{(a_{11}{a}_{22}-a_{12}{a}_{21})};k_0=\frac{c_1{a}_{22}-c_2{a}_{12}}{a_{12}{b}_{21}-a_{22}{b}_{11}};{\mathrm{\τ}}_{\mathrm{\γ}}=\frac{k_{\mathrm{\γ}}}{k_0{b}_{21}+c_2};$

Formula (9) is the expression formula of preferable vehicle steering model.

3. the control method of four-wheel independent steering vehicle as claimed in claim 2, it is characterised in that:

In described step A, the specific configuration process of front wheel angle observer is as follows:

Set front wheel angle and estimate initial valueFront wheel angle observer expression formula is as follows:

${\widehat{\mathrm{\δ}}}_f\left(t\right)={\∫}_0^t{e}^T(C-C_d)d\mathrm{\τ}---\left(10\right);$

In formula: e is the error vector between the actual side slip angle of automobile and yaw velocity and ideal model state, and expression formula is such as Under:

$e=x-x_d=\left[\begin{array}{c}{e}_1\\ e_2\end{array}\right]=\left[\begin{array}{c}{e}_{\mathrm{\β}}\\ e_{\mathrm{\γ}}\end{array}\right]=\left[\begin{array}{c}\mathrm{\β}-{\mathrm{\β}}_d\\ \mathrm{\γ}-{\mathrm{\γ}}_d\end{array}\right]=\left[\begin{array}{c}\mathrm{\β}\\ \mathrm{\γ}-{\mathrm{\γ}}_d\end{array}\right]---\left(11\right);$

e_β、e_γIt is respectively side slip angle and controls error and yaw velocity control error；

The ART network rule understanding front wheel angle according to formula (10) is

${\stackrel{\·}{\widehat{\mathrm{\δ}}}}_f=e^T(C-C_d)---\left(12\right);$

Control error equation is derived according to formula (8) and formula (9):

$\stackrel{\·}{e}=Ae+(A-A_d)x_d+Bu+(C-C_d){\widehat{\mathrm{\δ}}}_f+d\left(t\right)---\left(13\right).$

4. the control method of four-wheel independent steering vehicle as claimed in claim 3, it is characterised in that:

Described step A is disturbed border estimate that the specific configuration process of link is as follows:

The ART network rule on definition interference border is as follows:

$\left\{\begin{array}{c}{\stackrel{\·}{\widehat{\mathrm{\ψ}}}}_1={\mathrm{\ϵ}}_1{e}_{\mathrm{\β}}\mathrm{sgn}\left(e_{\mathrm{\β}}\right)={\mathrm{\ϵ}}_1\left|e_{\mathrm{\β}}\right|\\ {\stackrel{\·}{\widehat{\mathrm{\ψ}}}}_2={\mathrm{\ϵ}}_2{e}_{\mathrm{\γ}}\mathrm{sgn}\left(e_{\mathrm{\γ}}\right)={\mathrm{\ϵ}}_2\left|e_{\mathrm{\γ}}\right|\end{array}\right.---\left(14\right);$

In formula: sgn (.) represents symbol switch function；Represent interference boundary parameter ψ respectively₁And ψ₂Estimated value；ε₁、ε₂ It is called the estimation coefficient on interference border, and is all higher than 1；

Assuming that wheel steering initial timeWithThe mathematic(al) representation of link is estimated such as in interference border Under:

$\left\{\begin{array}{c}{\widehat{\mathrm{\ψ}}}_1\left(t\right)={\∫}_0^t{\mathrm{\ϵ}}_1\left|e_{\mathrm{\β}}\right|d\mathrm{\τ}\\ {\widehat{\mathrm{\ψ}}}_2\left(t\right)={\∫}_0^t{\mathrm{\ϵ}}_2\left|e_{\mathrm{\γ}}\right|d\mathrm{\τ}\end{array}\right.---\left(15\right);$

Estimate to draw according to formula (15).

5. the control method of four-wheel independent steering vehicle as claimed in claim 4, it is characterised in that:

In described step A, the specific configuration process of trailing wheel corner sliding mode controller and yaw moment sliding mode controller is as follows:

Definition sliding-mode surface function s=e, sliding mode controllerWherein, sliding mode controller u comprises rear round simultaneously Angle sliding mode controller and yaw moment sliding mode controller, and wheel corner δ later_rWith yaw moment M as controlled quentity controlled variable, u_eqFor sliding formwork Equivalent controller, u_sFor switch controller；Ignore shock wave d (t) caused by systematic parameter, according toAnd utilize Formula (15) can derive sliding formwork equivalent controller u_eqExpression formula as follows:

$u_{eq}=-B^{-1}\[-Ke-Ae-(A-A_d)x_d-(C-C_d){\widehat{\mathrm{\δ}}}_f\]---\left(16\right);$

In formula: K is control gain matrix undetermined,k₁And k₂It is all higher than zero, wherein diag (.) Represent diagonal matrix；

Switch controller u_sExpression formula as follows:

$u_s=-B^{-1}diag({\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1,{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2)sgn\left(e\right)---\left(17\right);$

In formula:For switching control u_sIn control gain；

Expression formula according to the available sliding mode controller of formula (16) and (17) is as follows:

$u=\left[\begin{array}{c}{\mathrm{\δ}}_r\\ M\end{array}\right]=u_{eq}+u_s=-B^{-1}\[Ke+Ae+(A-A_d)x_d+(C-C_d){\widehat{\mathrm{\δ}}}_f+diag({\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1,{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2)sgn\left(e\right)\]---\left(18\right).$

6. the control method of four-wheel independent steering vehicle as claimed in claim 5, it is characterised in that:

Described step E particularly as follows:

By matrix A, A_d、B、C、C_dSubstitute into formula (18) with the element of K, obtain the concrete of trailing wheel corner sliding mode controller by arrangement Form is as follows:

$\begin{array}{c}{\mathrm{\δ}}_r=-\{\[(k_1+a_{11})b_{22}-a_{21}{b}_{12}\]e_{\mathrm{\β}}+\[a_{12}{b}_{22}-b_{12}(k_2+a_{22})\]e_{\mathrm{\γ}}+\[a_{12}{b}_{22}-b_{12}(a_{22}-a_{22d})\]{\mathrm{\γ}}_d\\ +\[b_{22}{c}_1-b_{12}(c_2-c_{2d})\]{\widehat{\mathrm{\δ}}}_f+b_{22}{\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1\mathrm{sgn}\left(e_1\right)-b_{12}{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2\mathrm{sgn}\left(e_2\right)\}/(b_{11}{b}_{22}-b_{12}{b}_{21})\end{array}---\left(19\right);$

The concrete form of yaw moment sliding mode controller is:

$\begin{array}{c}M=\{\[(k_1+a_{11})b_{21}-a_{21}{b}_{11}\]e_{\mathrm{\β}}+\[a_{12}{b}_{21}-b_{11}(k_2+a_{22})\]e_{\mathrm{\γ}}+\[a_{12}{b}_{21}-b_{11}(a_{22}-a_{22d})\]{\mathrm{\γ}}_d\\ +\[b_{21}{c}_1-b_{11}(c_2-c_{2d})\]{\widehat{\mathrm{\δ}}}_f+b_{21}{\mathrm{\ϵ}}_1{\widehat{\mathrm{\ψ}}}_1\mathrm{sgn}\left(e_1\right)-b_{11}{\mathrm{\ϵ}}_2{\widehat{\mathrm{\ψ}}}_2\mathrm{sgn}\left(e_2\right)\}/(b_{11}{b}_{22}-b_{12}{b}_{21})\end{array}---\left(20\right);$

Use trailing wheel corner obtained above and yaw moment that vehicle is controlled in real time.