CN105313957A - Power assisted control method for electric power steering system based on compound control - Google Patents

Abstract

The invention provides a power assisted control method for an electric power steering system based on compound control and aims at providing safety, stability and comfort of an automobile in the running process. Firstly, according to the characteristics of nonlinearity of a system and parameter uncertainty, a sliding mode controller is designed; secondly, in order to enhance the robustness of the system, the special chattering phenomenon of sliding mode control is reduced and even eliminated, meanwhile, the self-adaption characteristic of the control rate is improved, and key parameters of the control rate are optimized through fuzzy control on the basis of sliding mode control; finally, in order to improve the control precision of the system, the control rule of fuzzy control is optimized through a genetic algorithm. By means of compound control of three control methods, the control precision and the robustness of the system are finally improved, meanwhile, the chattering phenomenon of the system is reduced and even eliminated, and therefore the automobile running safety, stability and comfort are improved.

Description

A kind of electric boosting steering system power assist control method based on complex controll

Technical field

The present invention relates to electric boosting steering system field, particularly relate to a kind of electric boosting steering system power assist control method based on complex controll.

Background technology

Vehicular electric servo-steering (EPS) system comprises the parts such as electronic control unit (ECU), torque sensor, car speed sensor, current sensor, assist motor and speed reduction gearing, and assist motor is connected with the steering hardware of automobile by transmission device.There is various control pattern in EPS, wherein Power assisted control is its basis.

Method for Power assisted control is diversified, and the control accuracy of traditional control method is low, poor robustness, and response time is long, and motor can be made like this to provide the delay of power-assisted, causes very large impact to the traveling comfort of automobile and safety.Ideal control method is Sliding mode variable structure control, but traditional Sliding mode variable structure control ubiquity chattering phenomenon.Buffeting can affect the stability of running car, and then affects the safety of automobile.In order to solve the problem, need a kind of new electric boosting steering system power assist control method.

The present invention devises the modified fuzzy sliding mode controlling method based on genetic algorithm optimization, not only efficiently reduces the buffeting of system, yet further enhances control accuracy and the robustness of system.And due to Sliding mode variable structure control, this control method can be applied on the automobile of multiple different model.

Summary of the invention

The invention provides a kind of electric boosting steering system power assist control method based on complex controll, object is to reduce even to eliminate the chattering phenomenon existed in traditional Sliding mode variable structure control, the control accuracy of further raising system and robustness.When vehicle enters steering state, the dtc signal that the ECU of electric boosting steering system records according to torque sensor and the vehicle speed signal that car speed sensor records obtain required target current, the actual current recorded by current sensor and the difference of target current are as the input of current controller, guarantee that actual current can follow target current fast and accurately, come to provide power torque fast and accurately with this.The technical solution adopted in the present invention is:

Based on an electric boosting steering system power assist control method for complex controll, it is characterized in that: the first step, the current error of design Sliding Mode Controller to electric boosting steering system controls; Second step, is incorporated in the Sliding Mode Controller of the first step by the fuzzy control that a dual input list exports, and the key parameter during sliding formwork is controlled has certain adaptive characteristic; 3rd step, is incorporated into genetic algorithm optimization in the fuzzy control of second step, is optimized fuzzy control rule.The principal particulars of the method is: fuzzy control is applied to sliding formwork and controls, reach the effect of the key parameter optimized in sliding formwork control, simultaneously by genetic algorithm optimization fuzzy rule, the particularity of fuzzy rule is improved with this, final realization is to the complex controll of EPS, the robustness of raising system and control accuracy, meanwhile, reduce the chattering phenomenon even eliminating system.

In the first step, by the target current I needed for system _mgather the deviation e of the current actual current I of assist motor as state variable with current sensor, the exponential approach rate in simultaneously controlling with reference to sliding formwork, in conjunction with the electrical specification of assist motor, finally tentatively obtains sliding mode controller.Replace symbolic function sgn (s) in desirable sliding mode with saturation function sat (s), by the serialization of established switching controls item, form the boundary 1ayer with certain neighborhood, strengthen the robustness of system.

In second step, the exponential approach rate formula during sliding formwork controls is ε can the performance of obvious influential system, utilizes fuzzy control to carry out adaptive control to it according to Fig. 3, if the input s of fuzzy controller and fuzzy subset on fuzzy domain export the fuzzy subset ε of ε on fuzzy domain={ NB, NS, ZO, PS, PB}.The Fuzzy Sliding Model Controller that assist motor is controlled can be formed thus;

In 3rd step, utilize genetic algorithm to be optimized to the fuzzy rule in the fuzzy control of second step, concrete control law optimizing process is: the parameter of random initializtion initial population, calculate target function value and accordingly just when; Calculate initial population just when summation, value, aviation value and maxim sample sequence number; Calculate individual just when with on average just when ratio, be greater than the individual reproduction of 1 with ratio, keep individual number in population constant; Match between two at random in previous generation population, with crossover probability be 0.85 random selecting intersect position carry out interlace operation, be just that 0.01. produces population of new generation thus with probability, calculate each sample target function value machine maxim and just when; Calculate population of new generation just when summation, value, aviation value and maxim sample sequence number; Judge whether to meet end condition, satisfied then terminate, the multiple said process of discontented lumping weight; By maximum just when sample decode, the control law be optimized.

Primely, current controller of the present invention will adopt based on the fuzzy sliding mode variable structure control of genetic algorithm optimization, because Sliding Mode Controller has switching characteristic, and with the parameter of system itself and disturbance irrelevant, therefore, this control method can be applied on the automobile of different model.

Primely, parameter during the present invention utilizes fuzzy control optimization sliding formwork to control, make parameter can according to different operating mode generation adaptive changes, utilize the control law in genetic algorithm optimization fuzzy control simultaneously, enhance the control accuracy of system, further increase adaptive characteristic and the robustness of system, meanwhile, reduce even to eliminate chattering phenomenon specific to traditional Sliding mode variable structure control.

Below in conjunction with accompanying drawing, describe advantages and features of the invention in detail.

Accompanying drawing explanation

Fig. 1 is electric boosting steering system sub-model schematic diagram;

Fig. 2 is the logic diagram of electric boosting steering system power assist control method;

Fig. 3 is fuzzy sliding mode variable structure control block diagram;

Fig. 4 is genetic algorithm diagram of circuit.

Detailed description of the invention

As Fig. 2, a kind of electric boosting steering system, comprise Micro-processor MCV 1, and the steering wheel torque sensor 2 be attached thereto, car speed sensor 3, PWM assist motor driver module 4, assist motor 5 and feedback current acquisition module 6, the torque signal that Micro-processor MCV 1 Real-time Collection is produced by steering wheel torque sensor 2 and the vehicle speed signal that car speed sensor 3 produces, and the target current of the assist motor that electric boosting steering system should export is obtained by built-in assist characteristic curve, this parameters input is processed to fuzzy controller, the pulse-width signal (PWM) regulated through Micro-processor MCV is exported to PWM assist motor driver module 4, control assist motor 5 by the mode of control PWM ripple dutycycle and export booster torquemoment, feedback current 6 acquisition module Real-time Collection assist motor actual current simultaneously, fuzzy controller 7 in input microprocessor MCU1, form closed loop control.

The invention discloses a kind of electric boosting steering system power assist control method based on complex controll, wherein key is, to the design of current controller, to be specifically implemented as follows:

1. when vehicle enters steering state, the dtc signal that the ECU of electric boosting steering system records according to torque sensor and the vehicle speed signal that car speed sensor records obtain required target current, the actual current recorded by current sensor and the difference of target current are as the input of current controller, guarantee that actual current can follow target current fast and accurately, come to provide power torque fast and accurately with this.

2. by the target current I needed for system _mgather the deviation e of the current actual current I of assist motor as state variable with current sensor, obtain switching function:

S=ce=c (I _m-I) formula (1)

In formula (1), c is proportionality coefficient, I _mfor target current, I is actual current.

To s differentiate:

S=ce=c (I _m-I) formula (2)

Choose exponential approach rate design sliding mode observer, that is:

$\stackrel{\·}{s}=-\mathrm{\ϵ}\·\mathrm{sgn}\left(s\right)+\mathrm{ks}$ Formula (3)

In formula 3, ε is the speed of system motion point convergence diverter surface.。

Actual sliding formwork controls to move to enable system remain on sliding-mode surface, needs to switch between different control logics, thus form " buffeting "." buffeting " phenomenon is by the particularity of influential system, the power consumption of increase system, serious infringement is caused to system, symbolic function sgn (s) in desirable sliding mode is replaced with saturation function sat (s), by the serialization of established switching controls item, form the boundary 1ayer with certain neighborhood, can reduce so even to eliminate buffeting, strengthen the robustness of system.

$\mathrm{sat}\left(s\right)=\left\{\begin{array}{cc}1& s>\mathrm{\&Delta;}\\ \mathrm{ks}& \left|s\right|\&le;\mathrm{\&Delta;}\\ -1& s<-\mathrm{\&Delta;}\end{array}\right.$ Wherein: $k=\frac{1}{\mathrm{\&Delta;}}$ Formula (4)

The switching controls item expression formula that can obtain system is thus:

U _s=ε sat (s)+ks formula (5)

In formula (5), ε is the speed of system motion point convergence diverter surface.

To sum up and in conjunction with the electrical specification of assist motor, the expression formula that can obtain sliding mode controller is:

$u=L{\stackrel{\&CenterDot;}{I}}_m+K_b{\stackrel{\&CenterDot;}{\mathrm{\&theta;}}}_m+\mathrm{RI}+\frac{\mathrm{L\&epsiv;}}{c}\mathrm{sat}\left(s\right)+\frac{\mathrm{Lk}}{c}s$ Formula (6)

K in formula 6 _bfor back EMF coefficient; θ _mfor steering wheel rotation angle; R, L are respectively armature resistance and inductance.

3., in sliding formwork controls, exponential approach rate formula is ε can the performance of obvious influential system.Reduce ε value, obviously can reduce system chatter, but s value is too little, meeting influential system arrives velocity of approach and the traverse time of diverter surface.Consider the uncertainty of parameter, in boundary 1ayer according to s and fuzzy value obfuscation adjustment is carried out to s, and then realize adjustment to switching controls item.

Utilize fuzzy control to carry out adaptive control to it according to Fig. 3, if the input s of fuzzy controller and fuzzy subset on fuzzy domain export the fuzzy subset ε of ε on fuzzy domain={ NB, NS, ZO, PS, PB}.The Fuzzy Sliding Model Controller that assist motor is controlled can be formed thus;

4. utilize genetic algorithm to be optimized the fuzzy rule in fuzzy control, because fuzzy controller is dual input, and be eachly input as 5 fuzzy sets, therefore have 25 control laws, the fuzzy language value of corresponding 1 ε of every rule.Fuzzy control rule is inputted with export s 5 fuzzy language values PB, PS, ZO, NS, NB} encode respectively, and coded system is translated into binary code, is followed successively by 000,001,010,011,100.Be converted in the engineering of binary string in rule, only corresponding fuzzy rule need be connected.

The objective function that the present invention adopts is:

J=∫ t|e|dt formula (7)

In formula (7) | e| is the absolute value of the error of system input and output, and J value is less, then the performance of system is better.

For the ease of realizing, by objective function discretization, can obtain final product

$\mathrm{\&Delta;J}=J(t+\mathrm{\&Delta;t})-J\left(t\right)={\&Integral;}_0^{t+\mathrm{\&Delta;t}}\mathrm{\&tau;}\left|e\right|\mathrm{d\&tau;}-{\&Integral;}_0^t\mathrm{\&tau;}\left|e\right|\mathrm{d\&tau;}={\&Integral;}_t^{t+\mathrm{\&Delta;t}}\mathrm{\&tau;}\left|e\right|\mathrm{d\&tau;}$ Formula (8)

In formula (8), Δ t is the sampling interval, generally very little, when getting τ=t, then and Δ J=t|E| Δ t.

Fitness function can be done suitable conversion to objective function and obtain.Here, following formula can be utilized determine:

$f=\frac{1}{1+J}$ Formula (9)

Concrete control law optimizing process is as Fig. 4, finally obtains the Fuzzy Sliding Model Controller based on genetic algorithm.

These embodiments are interpreted as only being not used in for illustration of the present invention limiting the scope of the invention above.After the content of reading record of the present invention, technical personnel can make various changes or modifications the present invention, and these equivalence changes and modification fall into the scope of the claims in the present invention equally.

Claims (3)

1. the electric boosting steering system power assist control method based on complex controll, the vehicle speed signal that the method records according to car speed sensor and the dtc signal combined action that torque sensor records are in the ECU of electric boosting steering system, target current is drawn according to the assist characteristic curve in ECU, the difference of actual current target current and current sensor recorded acts on assist motor by current controller, export power torque by assist motor, it comprises the steps:

The first step: utilize Sliding mode variable structure control to obtain the current control of electric boosting steering system.Using the difference of target current and actual current as state variable, obtain switching function input, namely

s＝c·e＝c(I _m-I)(1)

In formula 1, c is proportionality coefficient, I _mfor target current, I is actual current.

To s differentiate:

Choose exponential approach rate design sliding mode observer, that is:

In formula 3, ε is the speed of system motion point convergence diverter surface.

Utilize sat (s) to replace sgn (s), form the boundary 1ayer with certain neighborhood, obtain switching controls item expression formula: U _s=ε sat (s)+ks (4)

To sum up in conjunction with the electrical specification of motor, obtaining current controller is:

K in formula 5 _bfor back EMF coefficient; θ _mfor steering wheel rotation angle; R, L are respectively armature resistance and inductance.

Second step: utilize fuzzy control to make the speed ε of system motion point convergence diverter surface in Sliding Mode Controller have certain adaptive characteristic.

3rd step: utilize the fuzzy control rule in genetic algorithm optimization fuzzy control.

2. a kind of electric boosting steering system power assist control method based on complex controll according to claim 1, its feature d in: in second step, the fuzzy control exported by dual input list is introduced in first step Sliding mode variable structure control, optimize the key parameter ε in sliding formwork control, form the Fuzzy Sliding Model Controller that assist motor is controlled.If the input s of fuzzy controller and fuzzy subset on fuzzy domain the fuzzy subset ε of output ε on fuzzy domain={ NB, NS, ZO, PS, PB} by the obfuscation to input and output, formulate fuzzy rule, the steps such as the anti fuzzy method exported are obtained to the value of ε.

3. a kind of electric boosting steering system power assist control method based on complex controll according to claim 1, is characterized in that: in the 3rd step, genetic algorithm is introduced the fuzzy control of second step, be optimized fuzzy control rule.Optimizing process: the parameter of random initializtion initial population, calculate target function value and accordingly just when; Calculate initial population just when summation, value, aviation value and maxim sample sequence number; Calculate individual just when with on average just when ratio, be greater than the individual reproduction of 1 with ratio, keep individual number in population constant; Match between two at random in previous generation population, with crossover probability be 0.85 random selecting intersect position carry out interlace operation, be just that 0.01. produces population of new generation thus with probability, calculate each sample target function value machine maxim and just when; Calculate population of new generation just when summation, value, aviation value and maxim sample sequence number; Judge whether to meet end condition, satisfied then terminate, the multiple said process of discontented lumping weight; By maximum just when sample decode, the control law be optimized.