CN104925054A - Vehicle stable steering integrated control method based on differential flatness - Google Patents

Abstract

The invention relates to a vehicle stable steering control method, in particular to a unitary and integral type stable steering control method based on differential flatness. According to the method, the expected longitudinal velocity, the expected side velocity and the expected yaw velocity of a vehicle expected stable state are deduced from path information and serve as expectation of differential flatness control, through the combination of the differential flatness control and PID feedback regulation, a state variable of a vehicle system is kept to be around an expectancy value, and the steering stability of the vehicle which drives at a high speed is guaranteed; by utilizing a weighted control mode of a differential flatness corner and a driver preview corner, the weighting coefficient under different vehicle speeds is obtained, the vehicle steering angle is controlled jointly by a controller and a driver, and a desired trajectory can be tracked. By means of the method, the steering stability of the vehicle which drives at a high speed is guaranteed, and meanwhile a great trajectory tracking performance is achieved.

Description

A kind of vehicle stabilization based on differential flat turns to integrated control method

Technical field

The present invention relates to a kind of vehicle stabilization rotating direction control method, particularly a kind of integral type integrated form stable turning control method based on differential flat.

Background technology

Vehicle is promptly keeping away at a high speed barrier, tempo turn traveling and on low attachment road surface under the limiting condition such as turning driving, and tire very easily enters nonlinear operation region.If driving (braking) moment acting on wheel is excessive, tire easily produces sideslip, causes vehicle unstability; If steering wheel angle value is not suitable for, then easily causes vehicle unstability, cause a traffic accident.Therefore be necessary to drive Vehicular turn angle and tire or lock torque jointly controls, make tire produce rational longitudinal force and side force combination, finally order about vehicle smooth excessively curved.

From the viewpoint of vehicular drive and the control algorithm turning to integrating control, mainly contain Linear Control and nonlinear control method two kinds.At present, after generally considering to utilize differential flat method or feedback linearization method to process non-linear Vehicular system, the control problem of Vehicular system is solved with linear control method.

Existing differential flat controller Problems existing is, will have the expectation of driving data as controller of experience chaufeur, for the expectation not having the road of drive the cross just cannot provide controller, is difficult to engineer applied.In addition, the design of controller is not all for a certain fixed vehicle speed or fixing initial velocity and deceleration/decel, also or fixing desired speed sequence or the control effects under a certain path, study Vehicle Speed different or travel the intervention strength problem of Time Controller on the road that difference is linear.

Summary of the invention

Object of the present invention is exactly the defect travelling the interventional technique existence of Time Controller for solving current vehicle on the road that difference is linear, propose a kind of vehicle stabilization based on differential flat and turn to integrated control method, steadily smooth excessively curved under the operating mode of various paths to realize vehicle.

A kind of vehicle stabilization based on differential flat of the present invention turns to integrated control method, comprises the following steps:

Step S1. is by the road information-road curvature radius R of Real-time Collection ^refwith car status information-vehicular longitudinal velocity v _x, calculate the expectation state variable of Vehicular system, i.e. expectation-expectation the longitudinal velocity of controller expect side velocity with expectation yaw velocity ω ^ref;

1) according to the road curvature radius R collected ^ref, adopt and be calculated as follows expectation longitudinal velocity into the curved principle curved acceleration of slowing down

In formula: v ₀for entering curved moment speed;

2) according to road curvature radius R ^refwith the above-mentioned expectation longitudinal velocity calculated adopt the expectation yaw velocity of the kinematics formulae discovery vehicle of steady-state quantities:

3) according to road curvature radius R ^refwith the above-mentioned expectation longitudinal velocity calculated be calculated as follows the expectation side velocity of vehicle:

The expectation longitudinal velocity that step S2. calculates with step S1 expect side velocity with expectation yaw velocity ω ^refas the expectation of controller, computing controller output-output torque with output deflection angle

1) according to the expectation longitudinal velocity that step S2 calculates expect side velocity with expectation yaw velocity ω ^refand derivative calculation expectation differential flat output and derivative

Meanwhile, according to the actual longitudinal velocity v of the vehicle of onboard sensor Real-time Collection _x, side velocity v _ywith yaw velocity ω, calculate actual differential flat output y ₁, y ₂:

With in above formula: l _ffor front axle is to centroid distance; M is car mass; I _zfor automobile is around the rotor inertia of z-axis;

2) with the above-mentioned expectation differential flat output calculated and derivative with actual differential flat output y ₁, y ₂as input, calculate PID output

In formula:

3) with the above-mentioned PID output calculated as input, the output-output torque of computing controller with output deflection angle

Here f ₁, f ₂, f ₃be respectively:

Wherein, m is car mass, I _zfor automobile is around the rotor inertia of z-axis, l _ffor front-wheel is to the distance of barycenter, l _rfor trailing wheel is to the distance of barycenter, l is wheelbase, C _f, C _rfor front and back wheel cornering stiffness, ω _f, ω _rfor automobile front and back wheel cireular frequency, for automobile front and back wheel angular acceleration, R _efor vehicle wheel roll radius, I _ωfor vehicle wheel rotation inertia, ρ is density of air, C _xfor longitudinal aerodynamic drag factor, A _xfor longitudinal wind area.

The controller output torque that step S3. calculates step S2 input Vehicular system after distributing, calculate the moment T be applied on four wheels _fl, T _fr, T _fr, T _rr:

time, input Vehicular system the moment being applied to four wheels are:

input Vehicular system the moment being applied to four wheels are:

In formula: T _flfor the near front wheel, T _frfor off front wheel, T _rlfor left rear wheel, T _rrfor off hind wheel, M ₁, M ₂for Torque distribution coefficient; Preferred value is M ₁=0.325, M ₂=0.175;

Controller controls the moment T be applied on four wheels with this allocation scheme _fl, T _fr, T _fr, T _rr, the stable turning applying to realize vehicle controls;

The controller that step S4. calculates according to step S3 exports deflection angle corner δ is taken aim in advance with chaufeur _dthe demand deflection angle of Vehicular system when weighted calculation obtains the different speed of a motor vehicle

1) first judge whether controller is got involved:

in formula, v _facefor the critical speed of setting;

The critical speed preferably needing controller to get involved is v _face=15m/s;

2), when needing controller to get involved, the intervention intensity of different speed of a motor vehicle Time Controller is calculated as follows, i.e. coefficient of weight K ₁:

K ₁＝-0.002162v _x ²+0.1288v _x-1.516

3) according to coefficient of weight K ₁, calculate by controller and chaufeur co-controlling and input the Vehicular system demand deflection angle of vehicle

After being controlled by controller and chaufeur weighting, by Vehicular system demand deflection angle ^cinput Vehicular system, under applying to realize the different speed of a motor vehicle, Vehicular system is to the traceability of expected path;

Step S5. judges whether control process stops, road curvature radius R ^refwhen=0, vehicle rolls bend away from, and control process terminates, otherwise re-executes above step and realize cycle control.

Beneficial effect of the present invention: the present invention proposes a kind of integral type integrated control method based on differential flat, first, propose to expect stabilized conditions from routing information derivation vehicle based on nonlinear kinetics, it can be used as the expectation of differential flat controller, have more objectivity and operability; Secondly, controlled and PID feedback regulation by differential flat, realize the stability of vehicle high-speed Turning travel; Again, the present invention proposes vehicle differential flat corner and chaufeur when moving velocity is different and take aim at the transmission diversity weighting control method of corner in advance, consider by controller and chaufeur co-controlling Vehicular turn angle, thus ensure the track following performance of vehicle.

Accompanying drawing explanation

Fig. 1 is control method schematic flow sheet figure of the present invention;

Fig. 2 is control method integrating control general diagram of the present invention;

Fig. 3 be in the embodiment of the present invention under stable equilibrium state K value with the function changing relation diagram of curves of longitudinal velocity;

Fig. 4 be in the embodiment of the present invention coefficient of weight with speed of a motor vehicle variation relation matched curve figure;

Fig. 5 be in the embodiment of the present invention vehicle-state variable it is expected follow situation map;

Fig. 6 is track following design sketch in the embodiment of the present invention.

Detailed description of the invention

By the further specific descriptions of following examples, to do to understand further to content of the present invention, but be not to concrete restriction of the present invention.

Embodiment 1

With reference to Fig. 1,2, a kind of integral type integrated form stable turning control method based on differential flat, comprises the following steps:

Step S1. Real-time Collection road information: road curvature radius R ^refand car status information: vehicular longitudinal velocity v _x, side velocity v _y, yaw velocity ω.

Step S2. is that the stable turning realizing hot-short controls, and first needs the expectation state variable calculating Vehicular system, i.e. the expectation of controller: expect longitudinal velocity expect side velocity with expectation yaw velocity ω ^ref.

S2.1, according to gather road curvature radius R ^ref, calculation expectation longitudinal velocity

Adopt and the principle of curved acceleration into curved deceleration, calculate vehicle and expect longitudinal velocity

In formula: v ₀for entering curved moment speed, preferred value is 25m/s.

The expectation yaw velocity ω of S2.2, calculating vehicle ^ref:

According to road curvature radius R ^refwith the expectation longitudinal velocity that step S2.1 calculates adopt the kinematics formula of steady-state quantities, calculate the expectation yaw velocity of vehicle

S2.3, according to road curvature radius R ^ref, the expectation longitudinal velocity that calculates of step S2.1 with the expectation yaw velocity ω that step S2.2 calculates ^ref, calculate the expectation side velocity of vehicle

In stabilized zone, vehicle is expected side velocity and is expected that yaw velocity exists proportionate relationship: w ^ref=λ v _y ^ref.Point of stable equilibrium corresponding when utilizing genetic algorithm to try to achieve the longitudinal initial velocity of different vehicle, thus adopt least square fitting to draw corresponding λ value.λ value sees Fig. 3 with the function changing relation fitting result of longitudinal velocity and formula is as follows:

λ＝-55630v _x ^-4.039-0.07462

Thus calculation expectation side velocity

The vehicle expectation state variable that step S3. calculates with step S2: expect longitudinal velocity expect side velocity with expectation yaw velocity ω ^refas the expectation of controller, computing controller output: output torque with output deflection angle

S3.1, first need the smooth output of computing differential, comprise and expect differential flat output and derivative with actual differential flat output y ₁, y ₂:

First according to the expectation longitudinal velocity that step S2 calculates expect side velocity with expectation yaw velocity ω ^refand derivative calculation expectation differential flat output and derivative

In formula: l _ffor front axle is to centroid distance; M is car mass; I _zfor automobile is around the rotor inertia of z-axis.

Secondly according to the actual longitudinal velocity v of the vehicle of onboard sensor Real-time Collection _x, side velocity v _ywith yaw velocity ω, calculate actual differential flat output y ₁, y ₂:

L _ffor front axle is to centroid distance; M is car mass; I _zfor automobile is around the rotor inertia of z-axis.

S3.2, the differential flat output calculated with step S3.1: expect differential flat output and derivative with actual differential flat output y ₁, y ₂as input, calculate PID output

In formula:

S3.3, the PID output calculated with step S3.2 as input, the output of computing controller: output torque with output deflection angle

Here f ₁, f ₂, f ₃be respectively:

Wherein m is car mass, I _zfor automobile is around the rotor inertia of z-axis, l _ffor front-wheel is to the distance of barycenter, l _rfor trailing wheel is to the distance of barycenter, l is wheelbase, C _f, C _rfor front and back wheel cornering stiffness, ω _f, ω _rfor automobile front and back wheel cireular frequency, for automobile front and back wheel angular acceleration, R _efor vehicle wheel roll radius, I _ωfor vehicle wheel rotation inertia, ρ is density of air, C _xfor longitudinal aerodynamic drag factor, A _xfor longitudinal wind area.

The controller output torque that step S4. calculates step S3 input Vehicular system after distributing, realize the stable turning of vehicle:

Drive torque distributes at two first-class squares of front-wheel; The squares such as lock torque left and right wheels distribute, and according to actual demands of engineering, determine that antero posterior axis braking torque distribution ratio is M1/M2=0.65/0.35=1.85, namely front axle brake moment is rear axle lock torque is

Judge controller output torque sign:

time, for drive torque, input Vehicular system the moment being applied to four wheels are:

time, for lock torque, input Vehicular system the moment being applied to four wheels are:

In formula: T _flfor the near front wheel, T _frfor off front wheel, T _rlfor left rear wheel, T _rrfor off hind wheel, M ₁, M ₂for Torque distribution coefficient: preferred value is M ₁=0.325, M ₂=0.175.

Controller controls the moment T be applied on four wheels with this allocation scheme _fl, T _fr, T _fr, T _rr, apply to the stable turning realizing vehicle.

The controller that step S5. calculates according to step S3 exports deflection angle the Vehicular system demand deflection angle inputting vehicle is needed when calculating the different speed of a motor vehicle

The Vehicular system demand deflection angle of input Vehicular system because stability and track following need be taken into account, need to export deflection angle by controller corner δ is taken aim in advance with chaufeur _dweighted calculation obtains.By controller and chaufeur co-controlling Vehicular system demand deflection angle realize good path trace ability.

S5.1, first judge whether controller is got involved:

The critical speed preferably needing controller to get involved is v _face=15m/s.

S5.2, calculate the intervention intensity of different speed of a motor vehicle Time Controller, i.e. coefficient of weight K ₁:

S5.2.1, when needing controller to get involved, calculate the intervention intensity of different speed of a motor vehicle Time Controller, i.e. coefficient of weight K ₁, by controller and chaufeur co-controlling Vehicular turn angle.Adopt least square fitting, obtain weight coefficient and vehicular longitudinal velocity relational expression: K ₁=-0.002162v _x ²+ 0.1288v _x-1.516, corresponding matched curve as shown in Figure 4.

S5.2.2, when not needing controller to get involved, the intervention intensity of controller, i.e. coefficient of weight K ₁=0, by chaufeur individual operation Vehicular turn angle.

S5.3, the coefficient of weight K calculated according to step S5.2 ₁, calculate by controller and chaufeur co-controlling and input the Vehicular system demand deflection angle of Vehicular system

After being controlled by controller and chaufeur weighting, by Vehicular system demand deflection angle input Vehicular system, under applying to realize the different speed of a motor vehicle, Vehicular system is to the traceability of expected path.

Step S6. judges whether control process stops:

Judge whether vehicle rolls bend away from, road curvature radius R ^refwhen=0, vehicle rolls bend away from, and control process terminates, otherwise re-executes above step and realize cycle control.

Fig. 5-Figure 6 shows that vehicle travels initial velocity and gets 25m/s, the control effects figure obtained.As seen from Figure 5, the existing condition after wagon control is followed substantially near expectation state, illustrates that proposed control method achieves and controls the stability of Vehicular system.From Fig. 6 viewed from vehicle running path, only have slight deviations in laterally offset maximum distance apart and track when going out curved, but deviate is less, proves that this control method has good track following effect thus.

Claims (3)

1. the vehicle stabilization based on differential flat turns to an integrated control method, it is characterized in that comprising the following steps:

Step S1. is by the road information-road curvature radius R of Real-time Collection ^refwith car status information-vehicular longitudinal velocity v _x, calculate the expectation state variable of Vehicular system, i.e. expectation-expectation the longitudinal velocity of controller expect side velocity with expectation yaw velocity ω ^ref;

1) according to the road curvature radius R collected ^ref, adopt and be calculated as follows expectation longitudinal velocity into the curved principle curved acceleration of slowing down

$v_x^{ref}=10.75+0.875v_0-1002.53/R^{ref}$ In formula: v ₀for entering curved moment speed;

2) according to road curvature radius R ^refwith the above-mentioned expectation longitudinal velocity calculated adopt the expectation yaw velocity of the kinematics formulae discovery vehicle of steady-state quantities:

3) according to road curvature radius R ^refwith the above-mentioned expectation longitudinal velocity calculated be calculated as follows the expectation side velocity of vehicle: ${v_y}^{ref}=\frac{{v_x}^{ref}/R^{ref}}{-55630{\left({v_x}^{ref}\right)}^{-4.039}-0.07462}$

The expectation longitudinal velocity that step S2. calculates with step S1 expect side velocity with expectation yaw velocity ω ^refas the expectation of controller, computing controller output-output torque with output deflection angle

1) according to the expectation longitudinal velocity that step S2 calculates expect side velocity with expectation yaw velocity ω ^refand derivative calculation expectation differential flat output and derivative

$\begin{array}{cc}\left\{\begin{array}{c}{y_1}^{ref}={v_x}^{ref}\\ {y_2}^{ref}=l_f{{\mathrm{mv}}_y}^{ref}-I_z{w}^{ref}\end{array}\right.& \left\{\begin{array}{c}{{\stackrel{\·}{y}}_1}^{ref}={{\stackrel{\·}{v}}_x}^{ref}\\ {{\stackrel{\·\·}{y}}_2}^{ref}=l_{f}m{{\stackrel{\·\·}{v}}_y}^{ref}-I_z{\stackrel{\·\·}{\mathrm{\ω}}}^{ref}\end{array}\right.\end{array}$

Meanwhile, according to the actual longitudinal velocity v of the vehicle of onboard sensor Real-time Collection _x, side velocity v _ywith yaw velocity ω, calculate actual differential flat output y ₁, y ₂:

$\left\{\begin{array}{c}{y}_1=v_x\\ y_2=l_f{\mathrm{mv}}_y-I_z\mathrm{\ω}\end{array}\right.$

With in above formula: l _ffor front axle is to centroid distance; M is car mass; I _zfor automobile is around the rotor inertia of z-axis;

2) with the above-mentioned expectation differential flat output calculated and derivative with actual differential flat output y ₁, y ₂as input, calculate PID output

$\left\{\begin{array}{c}{\stackrel{\·}{y}}_1^c={\stackrel{\·}{y}}_1^{ref}+e_{y1}\\ {\stackrel{\·\·}{y}}_2^c={\stackrel{\·\·}{y}}_2^{ref}+60e_{y2}+12{\stackrel{\·}{e}}_{y2}\end{array}\right.$

In formula: $e_{y1}=y_1^{ref}-y_1;e_{y2}=y_2^{ref}-y_2;{\stackrel{\·}{e}}_{y2}={\stackrel{\·}{y}}_2^{ref}-{\stackrel{\·}{y}}_2.$

3) with the above-mentioned PID output calculated as input, the output-output torque of computing controller with output deflection angle

$\left\{\begin{array}{c}{T_w}^{ref}=\[{\mathrm{\Δ}}_{22}({\stackrel{\·}{y}}_1^c-F_1){\mathrm{\Δ}}_{12}({\stackrel{\·\·}{y}}_2^c-F_2)\]/({\mathrm{\Δ}}_{11}\·{\mathrm{\Δ}}_{22}-{\mathrm{\Δ}}_{12}\·{\mathrm{\Δ}}_{21})\\ {{\mathrm{\δ}}_f}^{ref}=\[{\mathrm{\Δ}}_{11}({\stackrel{\·\·}{y}}_2^c-F_2){\mathrm{\Δ}}_{21}({\stackrel{\·}{y}}_1^c-F_1)\]/({\mathrm{\Δ}}_{11}\·{\mathrm{\Δ}}_{22}-{\mathrm{\Δ}}_{12}\·{\mathrm{\Δ}}_{21})\end{array}\right.$

${\mathrm{\Δ}}_{11}=\frac{1}{mR}$

${\mathrm{\Δ}}_{12}=\frac{C_f}{m}\left(\frac{v_y+{\mathrm{wl}}_f}{v_x}\right)$

${\mathrm{\Δ}}_{21}=\frac{C_{r}l(v_y-{\mathrm{wl}}_r)-l_f{{\mathrm{mwv}}_x}^2}{{\mathrm{mR}}_e{v_x}^2}$

$\begin{array}{c}{\mathrm{\Δ}}_{\mathrm{22}}=\frac{(l_r{C}_{r}l-l_f{{\mathrm{mv}}_x}^2)l_f(C_f{R}_e-I_w{\stackrel{\·}{w}}_f)}{v_x{I}_z{R}_e}+\\ \frac{\left(C_{r}l\right(v_y-{\mathrm{wl}}_r)-l_f{{\mathrm{mwv}}_x}^2)C_f(v_y+{\mathrm{wl}}_f)}{{{\mathrm{mv}}_x}^3}-\frac{C_{r}l(C_f{R}_e-I_w{\stackrel{\·}{w}}_f)}{{\mathrm{mR}}_e{v}_x}\end{array}$

$F_1={\mathrm{wv}}_y-\frac{I_w}{mR}({\stackrel{\·}{w}}_r+{\stackrel{\·}{w}}_f)-\frac{\mathrm{\ρ}}{2m}{C}_x{A}_x{V_x}^2$

$F_2=\frac{\left(C_{r}l\right(v_y-{\mathrm{wl}}_r)-l_f{{\mathrm{mwv}}_x}^2}{{v_x}^2}{f}_1-\frac{C_{r}l}{v_x}{f}_2+\frac{(C_r{l}_{r}l-l_f{{\mathrm{mv}}_x}^2)}{v_x}{f}_3$

Here f ₁, f ₂, f ₃be respectively:

$f={\mathrm{wv}}_y-\frac{I_w({\stackrel{\·}{w}}_f+{\stackrel{\·}{w}}_r)}{{\mathrm{mR}}_e}$

$f_2=-{\mathrm{wv}}_x-\frac{C_f(v_y+{\mathrm{wl}}_f)+C_r(v_y-{\mathrm{wl}}_r)}{{\mathrm{mv}}_x}$

$f_3=\frac{-l_f{C}_f(v_y+{\mathrm{wl}}_f)+l_r{C}_r(v_y-{\mathrm{wl}}_r)}{v_x}$

Wherein, m is car mass, I _zfor automobile is around the rotor inertia of z-axis, l _ffor front-wheel is to the distance of barycenter, l _rfor trailing wheel is to the distance of barycenter, l is wheelbase, C _f, C _rfor front and back wheel cornering stiffness, ω _f, ω _rfor automobile front and back wheel cireular frequency, for automobile front and back wheel angular acceleration, R _efor vehicle wheel roll radius, I _ωfor vehicle wheel rotation inertia, ρ is density of air, C _xfor longitudinal aerodynamic drag factor, A _xfor longitudinal wind area.

The controller output torque that step S3. calculates step S2 input Vehicular system after distributing, calculate the moment T be applied on four wheels _fl, T _fr, T _fr, T _rr:

time, input Vehicular system the moment being applied to four wheels are: $\left\{\begin{array}{c}{T}_{fl}=T_{fr}=0.5T_w^{ref}\\ T_{rl}=T_{rr}=0\end{array}\right.;$

time, input Vehicular system the moment being applied to four wheels are: $\{\begin{array}{c}{T}_{fl}=T_{fr}=M_1*T_w^{ref}\\ T_{rl}=T_{rr}=M_2*T_w^{ref}\end{array};$

In formula: T _flfor the near front wheel, T _frfor off front wheel, T _rlfor left rear wheel, T _rrfor off hind wheel, M ₁, M ₂for Torque distribution coefficient;

Controller controls the moment T be applied on four wheels with this allocation scheme _fl, T _fr, T _fr, T _rr, the stable turning applying to realize vehicle controls;

The controller that step S4. calculates according to step S3 exports deflection angle corner δ is taken aim in advance with chaufeur _dthe demand deflection angle of Vehicular system when weighted calculation obtains the different speed of a motor vehicle

1) first judge whether controller is got involved:

in formula, v _facefor the critical speed of setting;

2), when needing controller to get involved, the intervention intensity of different speed of a motor vehicle Time Controller is calculated as follows, i.e. coefficient of weight K ₁:

K ₁＝-0.002162v _x ²+0.1288v _x-1.516

3) according to coefficient of weight K ₁, calculate by controller and chaufeur co-controlling and input the Vehicular system demand deflection angle of vehicle ${\mathrm{\δ}}_f^c=K_1{\mathrm{\δ}}_f^{ref}+(1-K_1){\mathrm{\δ}}_d;$

After being controlled by controller and chaufeur weighting, by Vehicular system demand deflection angle input Vehicular system, under applying to realize the different speed of a motor vehicle, Vehicular system is to the traceability of expected path;

Step S5. judges whether control process stops, road curvature radius R ^refwhen=0, vehicle rolls bend away from, and control process terminates, otherwise re-executes above step and realize cycle control.

2. a kind of vehicle stabilization based on differential flat according to claim 1 turns to integrated control method, it is characterized in that, the Torque distribution coefficient M described in step S3 ₁=0.325, M ₂=0.175.

3. a kind of vehicle stabilization based on differential flat according to claim 1 turns to integrated control method, it is characterized in that, the critical speed needing controller to get involved described in step S4 is v _face=15m/s.