CN104002808A - Active anti-collision automatic brake control system of automobile and working method - Google Patents

Abstract

The invention discloses an active anti-collision automatic brake control system of an automobile and a working method. The active anti-collision automatic brake control system comprises a pedestrian information acquiring and processing module, a driving safety state judging module and an automatic control module. The method includes steps that acquiring information of the pedestrian before the automobile, judging the driving safety state and controlling automatically. The active anti-collision automatic brake control system of the automobile and the working method use the learning rate mu<I> for integral I, use the learning rate mu<P> for proportion P and use the learning rate mu<D> for differential D so as to facilitate regulating different weight coefficients, the online correction for weighting coefficients is not based entirely on a neural network learning algorithm but formulated according to the actual situation, so that the timeliness and precision requirements are better satisfied. According to the active anti-collision automatic brake control system, the control system with an upper layer and a lower layer is designed based on sliding-mode control and mono-neuron PID control, the system buffeting phenomenon is effectively inhibited, the outside interference is overcame, the automatic stopping control for preventing the vehicle colliding with the pedestrian is realized, and the vehicle driving safety is guaranteed.

Description

A kind of automobile active anti-corrosion autobrake control system and method for work

Technical field

The invention belongs to automobile active safety field of intelligent control, relate to vehicle autobrake control method, specially refer to a kind of automobile active anti-corrosion autobrake control system and method for work.

Background technology

Automobile active anti-corrosion forewarn system has become the focus of domestic and international research.Active anti-collision system for car comprises following several gordian techniquies: running information perception and processing, the judgement of traffic safety state, vehicle dynamics modeling and control and control execution technique.The main goal in research of considering current active anti-collision system for car is using front vehicles as crashproof target, and research purpose is mainly positioned at and is improved on this car driving safety, there is no to embody the active safety protection to traffic disadvantaged group such as goers outside vehicles.

Due to the complexity of traffic environment information, variety and the randomness of pedestrian's motion feature; how to set up the traffic safety evaluation module based on pedestrains safety and be intended to protect pedestrian's auto control system in vehicles, this is also the important composition of future automobile active anti-corrosion forewarn system.

Summary of the invention

The problems referred to above that exist for solving prior art, the present invention will design one can detect the front pedestrian every trade car analysis of going forward side by side in real time, can under precarious position, realize automobile active anti-corrosion autobrake control system and the method for work of vehicle automatic train stop control.

For achieving the above object, technical scheme of the present invention is as follows:

A kind of automobile active anti-corrosion autobrake control system, comprise that pedestrian information obtains and processing module, traffic safety condition judgment module and automatic control module, described pedestrian information is obtained with processing module successively order and is connected in series with traffic safety condition judgment module and automatic control module; Automatic control module comprises acceleration/accel sliding formwork upper strata controller and single neuron PID lower floor controller; First utilize the infrared camera that is arranged on vehicle front to obtain the front vehicles pedestrian information every trade human information processing of going forward side by side, result after treatment is flowed to traffic safety condition judgment module, traffic safety condition judgment module sends danger signal to automatic control module, utilizes the braking of automatic control module control car retardation.

A method of work for automobile active anti-corrosion autobrake control system, comprises the following steps:

A, obtain vehicle front pedestrian information

Pedestrian information is obtained the infrared camera that utilization is installed on automobile with processing module and is obtained front pedestrian information, and real-time vehicle front pedestrian is detected and processed obtains accurately, failure-free pedestrian information.

B, carry out the judgement of traffic safety state

B1, calculating braking critical range

Vehicle and pedestrian are in same track, and the computing formula of vehicle and pedestrian's minimum braking critical range is:

$d_{w1}=v_h{t}_r+\frac{(v_h-v_l)t_i}{2}+\frac{{v_h}^2-{v_l}^2}{2\mathrm{fg}}+d_{\min }$

Wherein: v _hfor the rate of onset of vehicle, t _rfor time of driver's reaction and brake coordination time sum, value is 0.8～1.0s, t _ifor the time that deceleration/decel increases, get 0.1～0.2s, f is the adhesion value between tire and ground, d _minfor stopping the distance of minimum requirements between car pedestrian and vehicle, v _lfor the average velociity of pedestrian's walking, g is acceleration due to gravity, and value is 9.8m/s ².

B2, the judgement of traffic safety state

Obtaining after braking critical range, need to judge traffic safety state.The following formula of basis for estimation:

$\left\{\begin{array}{c}d>d_{w1}\\ d\&le;d_{w1}\end{array}\right.$

If real-time between vehicle and pedestrian is greater than braking critical range d apart from d _w1, drive a vehicle in a safe condition, otherwise it is in the hole to drive a vehicle, need to carry out vehicle and automatically control.

C, automatically control

System automatic control module act on driving real-time automatic control car retardation braking time in the hole, realize pedestrian's safety and the safety of Vehicle Driving Cycle, specifically comprise acceleration/accel sliding formwork upper strata controller and single neuron PID lower floor controller.

The effect of C1, described acceleration/accel sliding formwork upper strata controller is according to keeping suitable distance between vehicle and front pedestrian, determines the expectation acceleration/accel that under present case, vehicle is realized, and specifically comprises the following steps:

C11, ask for acceleration/accel sliding formwork upper strata controller variable parameter

Relative distance between pedestrian and vehicle and relative velocity error are as two indexs of acceleration/accel sliding formwork upper strata controller, and the variable parameter of definition acceleration/accel sliding formwork upper strata controller is

ε＝(y-x)+(L+H)

$\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}=v_l-v_h$

Wherein ε, be respectively relative distance error and relative velocity error; H is pedestrian and vehicle relative distance expectation value; L is bodywork length; Y, x are respectively the lengthwise position of pedestrian and vehicle; v _l, v _hbe respectively pedestrian's speed and the speed of vehicle.

C12, ask for vehicle expect acceleration/accel

Select the sliding formwork diverter surface of acceleration/accel sliding formwork upper strata controller to be:

$S\left(t\right)=\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}+{\mathrm{\&lambda;}}_1\mathrm{\&epsiv;}+{\mathrm{\&lambda;}}_2{\&Integral;}_0^t\mathrm{\&epsiv;dt}$

In formula: λ ₁> 0 and λ ₂> 0 is two parameters of sliding formwork control, to the differentiate of sliding formwork diverter surface,

$\stackrel{\&CenterDot;}{S}\left(t\right)={\stackrel{\&CenterDot;}{v}}_l-{\stackrel{\&CenterDot;}{v}}_h+{\mathrm{\&lambda;}}_1\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}+{\mathrm{\&lambda;}}_2\mathrm{\&epsiv;}$

In the control system of actual vehicle own, the impact of friction, external interference and Parameter Perturbation is difficult to avoid, and at this moment needs to consider to choose suitable inverse amplification factor, and symbolization function sgn (S) is restrained the single order differential of sliding formwork diverter surface S,

$\stackrel{\&CenterDot;}{S}=-\mathrm{\&beta;sgn}\left(S\right)\mathrm{\&beta;}>0$

The expectation acceleration/accel of finally trying to achieve vehicle is

$a_{\mathrm{cdes}}={\stackrel{\&CenterDot;}{v}}_l={\stackrel{\&CenterDot;}{v}}_h-{\mathrm{\&lambda;}}_1\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}-{\mathrm{\&lambda;}}_2\mathrm{\&epsiv;}-\mathrm{\&beta;sgn}\left(S\right)$

The stability of C13, checking acceleration/accel sliding formwork upper strata controller

Be defined as follows Lyapunov function

$V_2=\frac{1}{2}{S}^2$

Obviously V ₂> 0.To V ₂differentiate, have

${\stackrel{\&CenterDot;}{V}}_2=S\stackrel{\&CenterDot;}{S}=-S\&CenterDot;\mathrm{\&beta;sgn}\left(S\right)=-\left|S\right|\&CenterDot;\mathrm{\&beta;}$

Known by Lyapunov stability criterion, because of β > 0, obviously right have system stability, can effectively suppress and weaken the buffeting effect of the control system of actual vehicle own, disturbs and has good robustness to external world.

The effect of C2, described single neuron PID lower floor controller is the output according to acceleration/accel sliding formwork upper strata controller, and Carsim Full Vehicle Dynamics system is controlled, and realizes the acceleration/accel of this expectation, specifically comprises the following steps:

C21, the quantity of state of asking for single neuron PID lower floor controller are x ₁, x ₂, x ₃:

x ₁(t)＝e(t)＝a _cdes(t)-a _x(t)

x ₂(t)＝Δe(t)

x ₃(t)＝e(t)-2e(t-1)+e(t-2)＝e(t-1)-e(t-2)

C22, determine single neuron PID lower floor controller control algorithm and study:

$a_{\mathrm{ccon}}\left(t\right)=a_{\mathrm{ccon}}(t-1)+K\underset{i=1}{\overset{3}{\mathrm{\&Sigma;}}}{w}_i\left(t\right)x_i\left(t\right)$

$w_i\left(t\right)=w_j\left(t\right)/\underset{j=1}{\overset{3}{\mathrm{\&Sigma;}}}\left|w_j\left(t\right)\right|$

w ₁(t)＝w ₁(t-1)+μ _Ie(t)a _ccon(t-1)(e(t)+Δe(t))

w ₂(t)＝w ₂(t-1)+μ _Pe(t)a _ccon(t-1)(e(t)+Δe(t))

w ₃(t)＝w ₃(t-1)+μ _De(t)a _ccon(t-1)(e(t)+Δe(t))

μ in formula _i, μ _p, μ _dbe respectively the learning rate of integration, ratio, differential, K is neuronic proportionality coefficient, K > 0, w _i(t) be corresponding to x _i(t) coefficient of weight, a _cdes(t) for expecting acceleration/accel, a _x(t) be actual acceleration, a _ccon(t) for controlling acceleration/accel.

C3, acceleration/accel sliding formwork upper strata controller are according to speed and the position of vehicle and front pedestrian's current time, adopt sliding-mode control to determine the expectation acceleration/accel that under present case, vehicle is realized, send the expectation acceleration/accel of asking for to single neuron PID lower floor controller; Single neuron PID lower floor controller is according to output valve and the vehicle actual acceleration value of acceleration/accel sliding formwork upper strata controller, adopt Single neuron PID control method to try to achieve vehicle control acceleration/accel, Carsim Full Vehicle Dynamics system is controlled, realize the acceleration/accel of this expectation, and then reach the object of automatic control.

Compared with prior art, the present invention has following beneficial effect:

1, the present invention has adopted respectively different learning rate μ to integration I, ratio P and differential D _i, μ _p, μ _dto different weight coefficients is adjusted respectively, the online correction of coefficient of weight is not exclusively according to Learning Algorithm, but formulate with reference to actual conditions, can better meet the requirement of real-time and accuracy.

2, to have avoided the goal in research of active anti-collision system for car be the limitation of front vehicles in the present invention, design the control system with upper and lower double-layer structure based on sliding formwork control and Single neuron PID control simultaneously, can effectively suppress system chatter phenomenon, overcome the impact of external interference, having realized vehicle is the automatic train stop control that collision avoidance pedestrian carries out, and has ensured the safety of Vehicle Driving Cycle; There are certain practical significance and researching value.

Brief description of the drawings

The present invention has accompanying drawing 4 width, wherein:

Fig. 1 is composition schematic diagram of the present invention.

Fig. 2 is method flow diagram of the present invention.

Fig. 3 is system automatic control module schematic diagram of the present invention.

Fig. 4 is anticollision pedestrian scene schematic diagram of the present invention.

Detailed description of the invention

Below in conjunction with accompanying drawing, the present invention is described in detail.

As shown in Figure 1, a kind of automobile active anti-corrosion autobrake control system, comprises that pedestrian information is obtained and processing module, traffic safety condition judgment module and system automatic control module, and above-mentioned module successively signal connects.

As shown in Figure 2, a kind of method of work of automobile active anti-corrosion autobrake control system, comprises the following steps:

A, obtain vehicle front pedestrian information

Pedestrian information is obtained the infrared camera that utilization is installed on automobile with processing module and is obtained front pedestrian information, and real-time vehicle front pedestrian is detected and processed obtains accurately, failure-free pedestrian information.

B, carry out the judgement of traffic safety state

Vehicle and pedestrian are in same track, and the speed of vehicle is v _h, the real-time distance between vehicle and pedestrian is d, scene is as shown in Figure 4.

B1, calculating braking critical range

Vehicle and pedestrian are in same track, and the computing formula of vehicle and pedestrian's minimum braking critical range is:

$d_{w1}=v_h{t}_r+\frac{(v_h-v_l)t_i}{2}+\frac{{v_h}^2-{v_l}^2}{2\mathrm{fg}}+d_{\min }$

Wherein: v _hfor the rate of onset of vehicle, t _rfor time of driver's reaction and brake coordination time sum, value is 0.8～1.0s, t _ifor the time that deceleration/decel increases, get 0.1～0.2s, f is the adhesion value between tire and ground, d _minfor stopping the distance of minimum requirements between car pedestrian and vehicle, v _lfor the average velociity of pedestrian's walking, g is acceleration due to gravity, and value is 9.8m/s ².

B2, the judgement of traffic safety state

Obtaining after braking critical range, need to judge traffic safety state.The following formula of basis for estimation:

$\left\{\begin{array}{c}d>d_{w1}\\ d\&le;d_{w1}\end{array}\right.$

If real-time between vehicle and pedestrian is greater than braking critical range d apart from d _w1, drive a vehicle in a safe condition, otherwise it is in the hole to drive a vehicle, need to carry out vehicle and automatically control.

C, automatically control

System automatic control module act on driving real-time automatic control car retardation braking time in the hole, realize pedestrian's safety and the safety of Vehicle Driving Cycle, specifically comprise acceleration/accel sliding formwork upper strata controller, single neuron PID lower floor controller and Carsim Full Vehicle Dynamics system.

The effect of C1, described acceleration/accel sliding formwork upper strata controller is according to keeping suitable distance between vehicle and front pedestrian, determines the expectation acceleration/accel that under present case, vehicle is realized, and specifically comprises the following steps:

C11, ask for acceleration/accel sliding formwork upper strata controller variable parameter

Relative distance between pedestrian and vehicle and relative velocity error are as two indexs of acceleration/accel sliding formwork upper strata controller, and the variable parameter of definition acceleration/accel sliding formwork upper strata controller is

ε＝(y-x)+(L+H)

$\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}=v_l-v_h$

Wherein ε, be respectively relative distance error and relative velocity error; H is pedestrian and vehicle relative distance expectation value; L is bodywork length; Y, x are respectively the lengthwise position of pedestrian and vehicle; v _l, v _hbe respectively pedestrian's speed and the speed of vehicle.

C12, ask for vehicle expect acceleration/accel

Select the sliding formwork diverter surface of acceleration/accel sliding formwork upper strata controller to be:

$S\left(t\right)=\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}+{\mathrm{\&lambda;}}_1\mathrm{\&epsiv;}+{\mathrm{\&lambda;}}_2{\&Integral;}_0^t\mathrm{\&epsiv;dt}$

In formula: λ ₁> 0 and λ ₂> 0 is two parameters of sliding formwork control, to the differentiate of sliding formwork diverter surface,

$\stackrel{\&CenterDot;}{S}\left(t\right)={\stackrel{\&CenterDot;}{v}}_l-{\stackrel{\&CenterDot;}{v}}_h+{\mathrm{\&lambda;}}_1\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}+{\mathrm{\&lambda;}}_2\mathrm{\&epsiv;}$

In the control system of actual vehicle own, the impact of friction, external interference and Parameter Perturbation is difficult to avoid, and at this moment needs to consider to choose suitable inverse amplification factor, and symbolization function sgn (S) is restrained the single order differential of sliding formwork diverter surface S,

$\stackrel{\&CenterDot;}{S}=-\mathrm{\&beta;sgn}\left(S\right)\mathrm{\&beta;}>0$

The expectation acceleration/accel of finally trying to achieve vehicle is

$a_{\mathrm{cdes}}={\stackrel{\&CenterDot;}{v}}_l={\stackrel{\&CenterDot;}{v}}_h-{\mathrm{\&lambda;}}_1\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}-{\mathrm{\&lambda;}}_2\mathrm{\&epsiv;}-\mathrm{\&beta;sgn}\left(S\right)$

The stability of C13, checking acceleration/accel sliding formwork upper strata controller

Be defined as follows Lyapunov function

$V_2=\frac{1}{2}{S}^2$

Obviously V ₂> 0.To V ₂differentiate, have

${\stackrel{\&CenterDot;}{V}}_2=S\stackrel{\&CenterDot;}{S}=-S\&CenterDot;\mathrm{\&beta;sgn}\left(S\right)=-\left|S\right|\&CenterDot;\mathrm{\&beta;}$

Known by Lyapunov stability criterion, because of β > 0, obviously right have system stability, can effectively suppress and weaken the buffeting effect of the control system of actual vehicle own, disturbs and has good robustness to external world.

The effect of C2, described single neuron PID lower floor controller is the output according to acceleration/accel sliding formwork upper strata controller, and Carsim Full Vehicle Dynamics system is controlled, and realizes the acceleration/accel of this expectation, specifically comprises the following steps:

C21, the quantity of state of asking for single neuron PID lower floor controller are x ₁, x ₂, x ₃:

x ₁(t)＝e(t)＝a _cdes(t)-a _x(t)

x ₂(t)＝Δe(t)

x ₃(t)＝e(t)-2e(t-1)+e(t-2)＝e(t-1)-e(t-2)

C22, determine single neuron PID lower floor controller control algorithm and study:

$a_{\mathrm{ccon}}\left(t\right)=a_{\mathrm{ccon}}(t-1)+K\underset{i=1}{\overset{3}{\mathrm{\&Sigma;}}}{w}_i\left(t\right)x_i\left(t\right)$

$w_i\left(t\right)=w_j\left(t\right)/\underset{j=1}{\overset{3}{\mathrm{\&Sigma;}}}\left|w_j\left(t\right)\right|$

w ₁(t)＝w ₁(t-1)+μ _Ie(t)a _ccon(t-1)(e(t)+Δe(t))

w ₂(t)＝w ₂(t-1)+μ _Pe(t)a _ccon(t-1)(e(t)+Δe(t))

w ₃(t)＝w ₃(t-1)+μ _De(t)a _ccon(t-1)(e(t)+Δe(t))

μ in formula _i, μ _p, μ _dbe respectively the learning rate of integration, ratio, differential, K is neuronic proportionality coefficient, K > 0, w _i(t) be corresponding to x _i(t) coefficient of weight, a _cdes(t) for expecting acceleration/accel, a _x(t) be actual acceleration, a _ccon(t) for controlling acceleration/accel.

C3, acceleration/accel sliding formwork upper strata controller are according to speed and the position of vehicle and front pedestrian's current time, adopt sliding-mode control to determine the expectation acceleration/accel that under present case, vehicle is realized, send the expectation acceleration/accel of asking for to single neuron PID lower floor controller; Single neuron PID lower floor controller is according to output valve and the vehicle actual acceleration value of acceleration/accel sliding formwork upper strata controller, adopt Single neuron PID control method to try to achieve vehicle control acceleration/accel, Carsim Full Vehicle Dynamics system is controlled, realize the acceleration/accel of this expectation, and then reach the object of automatic control.

As shown in Figure 3, automatic control module of the present invention comprises acceleration/accel sliding formwork upper strata controller, and single neuron PID lower floor controller and Carsim Full Vehicle Dynamics system, comprise following specific implementation process:

The effect of C1, described acceleration/accel sliding formwork upper strata controller is according to keeping suitable distance between vehicle and front pedestrian, determines the expectation acceleration/accel that under present case, vehicle is realized, and specifically comprises the following steps:

C11, ask for acceleration/accel sliding formwork upper strata controller variable parameter

Relative distance between pedestrian and vehicle and relative velocity error are as two indexs of acceleration/accel sliding formwork upper strata controller, and the variable parameter of definition acceleration/accel sliding formwork upper strata controller is

ε＝(y-x)+(L+H)

$\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}=v_l-v_h$

Wherein ε, be respectively relative distance error and relative velocity error; H is pedestrian and vehicle relative distance expectation value; L is bodywork length; Y, x are respectively the lengthwise position of pedestrian and vehicle; v _l, v _hbe respectively pedestrian's speed and the speed of vehicle.

C12, ask for vehicle expect acceleration/accel

Select the sliding formwork diverter surface of acceleration/accel sliding formwork upper strata controller to be:

$S\left(t\right)=\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}+{\mathrm{\&lambda;}}_1\mathrm{\&epsiv;}+{\mathrm{\&lambda;}}_2{\&Integral;}_0^t\mathrm{\&epsiv;dt}$

In formula: λ ₁> 0 and λ ₂> 0 is two parameters of sliding formwork control, to the differentiate of sliding formwork diverter surface,

$\stackrel{\&CenterDot;}{S}\left(t\right)={\stackrel{\&CenterDot;}{v}}_l-{\stackrel{\&CenterDot;}{v}}_h+{\mathrm{\&lambda;}}_1\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}+{\mathrm{\&lambda;}}_2\mathrm{\&epsiv;}$

In the control system of actual vehicle own, the impact of friction, external interference and Parameter Perturbation is difficult to avoid, and at this moment needs to consider to choose suitable inverse amplification factor, and symbolization function sgn (S) is restrained the single order differential of sliding formwork diverter surface S,

$\stackrel{\&CenterDot;}{S}=-\mathrm{\&beta;sgn}\left(S\right)\mathrm{\&beta;}>0$

The expectation acceleration/accel of finally trying to achieve vehicle is

$a_{\mathrm{cdes}}={\stackrel{\&CenterDot;}{v}}_l={\stackrel{\&CenterDot;}{v}}_h-{\mathrm{\&lambda;}}_1\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}-{\mathrm{\&lambda;}}_2\mathrm{\&epsiv;}-\mathrm{\&beta;sgn}\left(S\right)$

The stability of C13, checking acceleration/accel sliding formwork upper strata controller

Be defined as follows Lyapunov function

$V_2=\frac{1}{2}{S}^2$

Obviously V ₂> 0.To V ₂differentiate, have

${\stackrel{\&CenterDot;}{V}}_2=S\stackrel{\&CenterDot;}{S}=-S\&CenterDot;\mathrm{\&beta;sgn}\left(S\right)=-\left|S\right|\&CenterDot;\mathrm{\&beta;}$

Known by Lyapunov stability criterion, because of β > 0, obviously right have system stability, can effectively suppress and weaken the buffeting effect of the control system of actual vehicle own, disturbs and has good robustness to external world.

The effect of C2, described single neuron PID lower floor controller is the output according to acceleration/accel sliding formwork upper strata controller, and Carsim Full Vehicle Dynamics system is controlled, and realizes the acceleration/accel of this expectation, specifically comprises the following steps:

C21, the quantity of state of asking for single neuron PID lower floor controller are x ₁, x ₂, x ₃:

x ₁(t)＝e(t)＝a _cdes(t)-a _x(t)

x ₂(t)＝Δe(t)

x ₃(t)＝e(t)-2e(t-1)+e(t-2)＝e(t-1)-e(t-2)

C22, determine single neuron PID lower floor controller control algorithm and study:

$a_{\mathrm{ccon}}\left(t\right)=a_{\mathrm{ccon}}(t-1)+K\underset{i=1}{\overset{3}{\mathrm{\&Sigma;}}}{w}_i\left(t\right)x_i\left(t\right)$

$w_i\left(t\right)=w_j\left(t\right)/\underset{j=1}{\overset{3}{\mathrm{\&Sigma;}}}\left|w_j\left(t\right)\right|$

w ₁(t)＝w ₁(t-1)+μ _Ie(t)a _ccon(t-1)(e(t)+Δe(t))

w ₂(t)＝w ₂(t-1)+μ _Pe(t)a _ccon(t-1)(e(t)+Δe(t))

w ₃(t)＝w ₃(t-1)+μ _De(t)a _ccon(t-1)(e(t)+Δe(t))

μ in formula _i, μ _p, μ _dbe respectively the learning rate of integration, ratio, differential, K is neuronic proportionality coefficient, K > 0, w _i(t) be corresponding to x _i(t) coefficient of weight, a _cdes(t) for expecting acceleration/accel, a _x(t) be actual acceleration, a _ccon(t) for controlling acceleration/accel.

C3, acceleration/accel sliding formwork upper strata controller are according to speed and the position of vehicle and front pedestrian's current time, adopt sliding-mode control to determine the expectation acceleration/accel that under present case, vehicle is realized, send the expectation acceleration/accel of asking for to single neuron PID lower floor controller; Single neuron PID lower floor controller is according to output valve and the vehicle actual acceleration value of acceleration/accel sliding formwork upper strata controller, adopt Single neuron PID control method to try to achieve vehicle control acceleration/accel, Carsim Full Vehicle Dynamics system is controlled, realize the acceleration/accel of this expectation, and then reach the object of automatic control.

Claims (2)

1. an automobile active anti-corrosion autobrake control system, it is characterized in that: comprise that pedestrian information obtains and processing module, traffic safety condition judgment module and automatic control module, described pedestrian information is obtained with processing module successively order and is connected in series with traffic safety condition judgment module and automatic control module; Automatic control module comprises acceleration/accel sliding formwork upper strata controller and single neuron PID lower floor controller; First utilize the infrared camera that is arranged on vehicle front to obtain the front vehicles pedestrian information every trade human information processing of going forward side by side, result after treatment is flowed to traffic safety condition judgment module, traffic safety condition judgment module sends danger signal to automatic control module, utilizes the braking of automatic control module control car retardation.

2. a method of work for automobile active anti-corrosion autobrake control system, is characterized in that: comprise the following steps:

A, obtain vehicle front pedestrian information

Pedestrian information is obtained the infrared camera that utilization is installed on automobile with processing module and is obtained front pedestrian information, and real-time vehicle front pedestrian is detected and processed obtains accurately, failure-free pedestrian information;

B, carry out the judgement of traffic safety state

B1, calculating braking critical range

Vehicle and pedestrian are in same track, and the computing formula of vehicle and pedestrian's minimum braking critical range is:

$d_{w1}=v_h{t}_r+\frac{(v_h-v_l)t_i}{2}+\frac{{v_h}^2-{v_l}^2}{2\mathrm{fg}}+d_{\min }$

Wherein: v _hfor the rate of onset of vehicle, t _rfor time of driver's reaction and brake coordination time sum, value is 0.8～1.0s, t _ifor the time that deceleration/decel increases, get 0.1～0.2s, f is the adhesion value between tire and ground, d _minfor stopping the distance of minimum requirements between car pedestrian and vehicle, v _lfor the average velociity of pedestrian's walking, g is acceleration due to gravity, and value is 9.8m/s ²;

B2, the judgement of traffic safety state

Obtaining after braking critical range, need to judge traffic safety state; The following formula of basis for estimation:

$\left\{\begin{array}{c}d>d_{w1}\\ d\&le;d_{w1}\end{array}\right.$

If real-time between vehicle and pedestrian is greater than braking critical range d apart from d _w1, drive a vehicle in a safe condition, otherwise it is in the hole to drive a vehicle, need to carry out vehicle and automatically control;

C, automatically control

System automatic control module act on driving real-time automatic control car retardation braking time in the hole, realize pedestrian's safety and the safety of Vehicle Driving Cycle, specifically comprise acceleration/accel sliding formwork upper strata controller and single neuron PID lower floor controller;

The effect of C1, described acceleration/accel sliding formwork upper strata controller is according to keeping suitable distance between vehicle and front pedestrian, determines the expectation acceleration/accel that under present case, vehicle is realized, and specifically comprises the following steps:

C11, ask for acceleration/accel sliding formwork upper strata controller variable parameter

Relative distance between pedestrian and vehicle and relative velocity error are as two indexs of acceleration/accel sliding formwork upper strata controller, and the variable parameter of definition acceleration/accel sliding formwork upper strata controller is

ε＝(y-x)+(L+H)

$\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}=v_l-v_h$

Wherein ε, be respectively relative distance error and relative velocity error; H is pedestrian and vehicle relative distance expectation value; L is bodywork length; Y, x are respectively the lengthwise position of pedestrian and vehicle; v _l, v _hbe respectively pedestrian's speed and the speed of vehicle;

C12, ask for vehicle expect acceleration/accel

Select the sliding formwork diverter surface of acceleration/accel sliding formwork upper strata controller to be:

$S\left(t\right)=\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}+{\mathrm{\&lambda;}}_1\mathrm{\&epsiv;}+{\mathrm{\&lambda;}}_2{\&Integral;}_0^t\mathrm{\&epsiv;dt}$

In formula: λ ₁> 0 and λ ₂> 0 is two parameters of sliding formwork control, to the differentiate of sliding formwork diverter surface,

$\stackrel{\&CenterDot;}{S}\left(t\right)={\stackrel{\&CenterDot;}{v}}_l-{\stackrel{\&CenterDot;}{v}}_h+{\mathrm{\&lambda;}}_1\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}+{\mathrm{\&lambda;}}_2\mathrm{\&epsiv;}$

In the control system of actual vehicle own, the impact of friction, external interference and Parameter Perturbation is difficult to avoid, and at this moment needs to consider to choose suitable inverse amplification factor, and symbolization function sgn (S) is restrained the single order differential of sliding formwork diverter surface S,

$\stackrel{\&CenterDot;}{S}=-\mathrm{\&beta;sgn}\left(S\right)\mathrm{\&beta;}>0$

The expectation acceleration/accel of finally trying to achieve vehicle is

$a_{\mathrm{cdes}}={\stackrel{\&CenterDot;}{v}}_l={\stackrel{\&CenterDot;}{v}}_h-{\mathrm{\&lambda;}}_1\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}-{\mathrm{\&lambda;}}_2\mathrm{\&epsiv;}-\mathrm{\&beta;sgn}\left(S\right)$

The stability of C13, checking acceleration/accel sliding formwork upper strata controller

Be defined as follows Lyapunov function

$V_2=\frac{1}{2}{S}^2$

Obviously V ₂> 0; To V ₂differentiate, have

${\stackrel{\&CenterDot;}{V}}_2=S\stackrel{\&CenterDot;}{S}=-S\&CenterDot;\mathrm{\&beta;sgn}\left(S\right)=-\left|S\right|\&CenterDot;\mathrm{\&beta;}$

Known by Lyapunov stability criterion, because of β > 0, obviously right have system stability, can effectively suppress and weaken the buffeting effect of the control system of actual vehicle own, disturbs and has good robustness to external world;

The effect of C2, described single neuron PID lower floor controller is the output according to acceleration/accel sliding formwork upper strata controller, and Carsim Full Vehicle Dynamics system is controlled, and realizes the acceleration/accel of this expectation, specifically comprises the following steps:

C21, the quantity of state of asking for single neuron PID lower floor controller are x ₁, x ₂, x ₃:

x ₁(t)＝e(t)＝a _cdes(t)-a _x(t)

x ₂(t)＝Δe(t)

x ₃(t)＝e(t)-2e(t-1)+e(t-2)＝e(t-1)-e(t-2)

C22, determine single neuron PID lower floor controller control algorithm and study:

$a_{\mathrm{ccon}}\left(t\right)=a_{\mathrm{ccon}}(t-1)+K\underset{i=1}{\overset{3}{\mathrm{\&Sigma;}}}{w}_i\left(t\right)x_i\left(t\right)$

$w_i\left(t\right)=w_j\left(t\right)/\underset{j=1}{\overset{3}{\mathrm{\&Sigma;}}}\left|w_j\left(t\right)\right|$

w ₁(t)＝w ₁(t-1)+μ _Ie(t)a _ccon(t-1)(e(t)+Δe(t))

w ₂(t)＝w ₂(t-1)+μ _Pe(t)a _ccon(t-1)(e(t)+Δe(t))

w ₃(t)＝w ₃(t-1)+μ _De(t)a _ccon(t-1)(e(t)+Δe(t))

μ in formula _i, μ _p, μ _dbe respectively the learning rate of integration, ratio, differential, K is neuronic proportionality coefficient, K > 0, w _i(t) be corresponding to x _i(t) coefficient of weight, a _cdes(t) for expecting acceleration/accel, a _x(t) be actual acceleration, a _ccon(t) for controlling acceleration/accel;

C3, acceleration/accel sliding formwork upper strata controller are according to speed and the position of vehicle and front pedestrian's current time, adopt sliding-mode control to determine the expectation acceleration/accel that under present case, vehicle is realized, send the expectation acceleration/accel of asking for to single neuron PID lower floor controller; Single neuron PID lower floor controller is according to output valve and the vehicle actual acceleration value of acceleration/accel sliding formwork upper strata controller, adopt Single neuron PID control method to try to achieve vehicle control acceleration/accel, Carsim Full Vehicle Dynamics system is controlled, realize the acceleration/accel of this expectation, and then reach the object of automatic control.