CN106945670A - Anti-rollover system for automobiles and control strategy based on driver's input prediction - Google Patents

Abstract

The invention discloses a vehicle anti-rollover system based on driver input prediction and a control strategy. The vehicle anti-rollover system includes a sensor module, a steering wheel angle estimation unit, a front wheel differential braking module, a rollover evaluation unit and a control system. device ECU. The present invention predicts the input value of the steering wheel of the vehicle model online through the gray prediction model, and proposes a predictable vehicle rollover evaluation index based on this, establishes an estimation model of passenger car LTR, and when the vehicle LTR estimated value reaches the rollover threshold, through the front wheel difference The dynamic brake module performs anti-rollover control, which avoids the influence of the time lag of the control system on the rollover judgment.

Description

Vehicle anti-rollover system and control strategy based on driver input prediction

technical field

The invention belongs to the technical field of automobile safety, and in particular relates to an automobile rollover prevention system and control strategy based on driver input prediction.

Background technique

The appearance of automobiles has brought great changes to the way of life of human beings. The rapid development of the automobile industry and the continuous improvement of the economic level have promoted the popularization and use of automobiles. At the same time, it has also brought a series of social problems. With the increasing popularity of automobiles and the increase of road traffic, road traffic accidents have become an increasingly serious public safety problem worldwide and a major public hazard in modern society.

Some vehicles, such as SUVs, buses, and heavy-duty semi-trailers, are prone to rollover accidents due to their high center of mass, relatively large mass and volume, and relatively narrow wheelbase. When a rollover accident occurs, the driver is hardly aware of the rollover. According to the statistics of the U.S. Highway Traffic Safety Administration, from 1992 to 1996, there were as many as 22,700 vehicle rollover accidents in the United States every year, which is second only to frontal collisions. From 1993 to 1998, more than 35,000 people died in traffic accidents, of which non-collision accidents accounted for 10%, and 90% of major non-collision accidents were rollover accidents. With the continuous growth of the number of cars in China and the rapid development of transportation, major traffic accidents such as vehicle rollovers continue to increase. It can be seen that it is necessary to develop a rollover early warning system to avoid frequent rollover accidents.

At present, most traffic accidents are caused by the driver's wrong operation, so there is an urgent need for research on active safety control based on driving intention. However, using the existing anti-rollover control strategy inevitably has the following problems: first, there is a time lag in the control system, so that the obtained rollover evaluation index has a hysteresis; second, the driving intention cannot be considered; Three, it is impossible to estimate the rollover state in the future.

Contents of the invention

Purpose of the invention: In order to overcome the deficiencies in the prior art, the present invention provides a vehicle anti-rollover system and control strategy based on driver input prediction, which can predict the driver's steering wheel input in the future and effectively prevent vehicle rollover.

Technical scheme: in order to achieve the above object, the technical scheme adopted in the present invention is:

A vehicle anti-rollover system based on driver input prediction, comprising a sensor module connected in sequence, a steering wheel angle estimation unit, a vehicle parameter estimation unit and a rollover evaluation unit, wherein the sensor module is also connected with the vehicle parameter estimation unit The unit is connected, and the sensor module includes a vehicle speed sensor and a steering wheel angle sensor, which measure the vehicle speed and the steering wheel angle of the vehicle respectively; the steering wheel angle estimation unit is connected with the steering wheel angle sensor, and the vehicle parameter estimation unit is connected with the steering wheel angle estimation unit and the vehicle speed The sensor is connected, and the rollover evaluation unit is connected with the steering wheel angle estimation unit and the vehicle parameter estimation unit respectively;

The rollover evaluation unit outputs the obtained information to the front wheel differential braking module and the control ECU respectively, wherein the trigger signal is output to the differential braking module, and the rollover evaluation value is output to the ECU; the control ECU outputs the braking pressure The difference signal is sent to the front wheel differential brake module.

The above-mentioned control strategy of the anti-rollover system of the automobile based on driver input prediction comprises the following steps:

Step 1), the steering wheel angle sensor senses the steering wheel angle of the vehicle, and transmits the steering wheel angle value to the steering wheel angle estimation unit; the vehicle speed sensor senses the vehicle speed, and transmits it to the vehicle parameter estimation unit;

Step 2), the steering wheel angle estimation unit estimates the steering wheel input value at the future time based on the steering wheel angle value at the current moment and the gray prediction model theory, and outputs the estimated value to the vehicle parameter estimation unit for vehicle rollover related parameters Prediction;

Step 3), the vehicle parameter estimation unit takes the obtained steering wheel predicted value and vehicle speed as input, and calculates the predicted values of various parameters of the vehicle, including the predicted value of the lateral acceleration of the vehicle body, the predicted value of the vehicle body roll angle, and the predicted value of the vehicle body roll angle velocity value, and input it into the rollover evaluation unit;

Step 4), the rollover evaluation unit calculates the rollover evaluation value according to the received predicted value of vehicle body roll angle, predicted value of vehicle body roll angular velocity, and predicted value of vehicle body lateral acceleration, and transmits it to the controller ECU. Send a trigger signal to the front wheel differential braking module when the rollover evaluation value is greater than the preset rollover threshold;

Step 5), the control ECU compares the received rollover evaluation value with the preset rollover threshold, and if the received rollover evaluation value is greater than the preset rollover threshold, then outputs the brake pressure difference signal to the front Wheel differential brake module;

Step 6), the front wheel differential braking module adjusts the braking pressure of the outer front wheels of the vehicle according to the received braking pressure difference signal, so that the difference between the two front wheel braking pressures and the received front wheel differential braking Brake the vehicle in response to the start signal.

Further, the gray prediction model in step 2) is: adopt the gray prediction GM (1, 1) model, take the steering wheel angle value at the current moment as an input value, and when the sampling time Ts is 0.02s, the number of advance prediction steps k is taken as 5. It means that the advance prediction time T (T=Ts×k) is 0.1s, and the steering wheel angle value after 0.1s in step 5 can be calculated; take the steering wheel angle obtained by the sensor as the input, and use the gray prediction model to predict 0.1 in advance The steering wheel angle value after s, and output it to the car parameter estimation unit.

Further, the step 3) specific implementation steps are as follows:

Step 3.1), establish a linear three-degree-of-freedom vehicle rollover model

Step 3.1.1), vehicle model

The model ignores the dynamic characteristics of the car in the longitudinal and pitch directions, and assumes that the dynamics of the left and right wheels of the car are symmetrical about the X axis, that is, the model consists of a "bicycle model" and a roll plane model, including lateral motion along the Y direction, The yaw motion of the Z axis and the roll motion around the X axis;

Considering the coupling effect between the three degrees of freedom, the vehicle rollover dynamics equation is:

Lateral movement:

Lateral movement:

Rolling movement:

From the yaw and lateral motion coupling relationship, the lateral acceleration at the center of mass of the vehicle can be obtained as:

In the formula: a and b are the distances from the center of mass of the car to the front and rear axles; F _f and F _r are the cornering forces of the front and rear wheels respectively; h is the distance from the roll center to the center of mass; I _x is the roll moment of inertia of the sprung mass; I _z is the yaw moment of inertia; with is the suspension equivalent roll stiffness and equivalent roll damping coefficient; m is the mass of the vehicle; m _s is the sprung mass; is the additional yaw moment output by the control system; r is the yaw rate; v is the lateral velocity; is the roll angle of the sprung mass; u is the longitudinal velocity of the vehicle; a _y is the lateral acceleration of the vehicle; F _f and F _r are the lateral forces of the front and rear wheels;

Step 3.1.2) Build tire model

Considering the effects of roll steering, roll camber, deformation steering and deformation camber on tire lateral characteristics, from the relationship between lateral force, speed and rotation angle, the slip angle of front and rear wheels can be obtained as:

In the formula: k _a is the camber coefficient, δ _sw is the steering wheel angle value, i is the vehicle transmission ratio;

Then the cornering force of the front and rear wheels can be obtained as:

F _f =k _f β _f

F _r =k _r β _r

In the formula: k _f is the cornering stiffness of the front wheel, and k _r is the cornering stiffness of the rear wheel;

Step 3.2), according to the predicted value of the steering wheel angle, the vehicle speed and the above formula, calculate the predicted value of the lateral acceleration of the vehicle body, the predicted value of the body roll angle and the predicted value of the vehicle body roll angle velocity.

Further, the step 4) specific steps are as follows:

Step 4.1) The lateral load transfer rate is used as the rollover factor of the vehicle, and the lateral load transfer rate is defined as the ratio of the difference between the vertical loads of the left and right tires to the total vertical load of the vehicle, and the expression is as follows:

In the formula, F _zl and F _zr are the left and right vertical loads of the vehicle respectively, and it can be seen that:

F _zr +F _zl =mg

In the formula, m is the total mass of the car; the value of LTR is [-1,l]. When LTR=0, the car has no roll, and no rollover will occur. When LTR=±1, there is a vertical load on one side of the wheel If it is 0, one side of the tire is off the ground, and there is a risk of rollover;

Step 4.2), the force balance equation of the car around the roll center is:

The force balance equation of the car around the center point of the wheel base on the ground is:

The lateral load transfer rate of the car can be obtained as:

In the formula: t _w is the wheel base, h _R is the distance from the roll center to the ground.

Further, the specific method of step 5) is: control the ECU to compare the received rollover evaluation value with the preset rollover threshold,

Step 5.1), if the received rollover evaluation value is greater than the preset rollover threshold;

Step 5.1.1), making a difference between the received rollover evaluation value and the preset rollover threshold to obtain a rollover difference, and calculating a brake pressure difference signal according to the rollover difference;

Step 5.1.2), outputting the brake pressure difference signal to the front wheel differential brake module;

Further, the rollover threshold in step 5) is 0.7-0.9.

Beneficial effects: Compared with the prior art, the anti-rollover system and control strategy of the vehicle based on driver input prediction provided by the present invention have the following technical effects:

1. It can predict the driver's steering wheel input in the future, and better simulate the influence of the time-lag of the control system in the traditional control strategy.

2. This method can reflect the driver's steering intention to a certain extent.

Description of drawings

Figure 1 is a work flow chart of the anti-rollover control system;

Fig. 2 is a working flow diagram of the steering wheel angle estimation unit;

Fig. 3 is a working flow chart of the front wheel differential braking module;

Fig. 4 is a schematic diagram of steering wheel angle input in the embodiment.

detailed description

The present invention will be further described below in conjunction with the accompanying drawings.

The present invention is a driver's steering wheel input prediction model based on gray prediction. Using the gray prediction GM (1, 1) model, the steering wheel angle value at the current moment is used as the input value. When the sampling time Ts is 0.02s, the prediction is made in advance. The number of steps k is 5, which means that the advance prediction time T (T=Ts×k) is 0.1s, and the steering wheel angle value at the fifth step (after 0.1s) can be calculated.

The steering wheel angle estimation unit is connected with the steering wheel angle sensor and the rollover evaluation unit. The steering wheel angle obtained by the sensor is taken as input, and the gray prediction model is used to predict the steering wheel angle value after 0.1s in advance, and output to the vehicle parameter estimation unit .

As shown in Figure 1, the present invention discloses a vehicle rollover prevention system, including a sensor module, a front wheel differential braking module, a steering wheel angle estimation unit, a vehicle parameter estimation unit, a rollover evaluation unit and a controller ECU ;

The sensor module includes a vehicle speed sensor and a steering wheel angle sensor.

The sensor module includes a vehicle speed sensor, a steering wheel angle sensor and an acceleration sensor, which are used to measure the vehicle speed and steering wheel angle respectively.

The steering wheel angle estimation unit is connected to the steering wheel angle sensor, and its function is to estimate the steering wheel input value in the future based on the steering wheel angle value at the current moment and the gray prediction model theory, and output the estimated value to the vehicle parameter estimation unit for further calculation. Prediction of parameters related to vehicle rollover;

The vehicle parameter estimation unit is connected with the steering wheel angle estimation unit and the vehicle speed sensor module, and the obtained steering wheel prediction value and vehicle speed are used as input to calculate the prediction values of various parameters of the vehicle (body lateral acceleration prediction value, body roll angle prediction value, predicted value of body roll angular velocity).

The rollover evaluation unit is connected to the steering wheel angle estimation unit and the vehicle parameter estimation unit respectively, and is used to calculate the rollover evaluation value according to the steering wheel angle prediction value and the sensing data of the vehicle speed sensor, and then transmit it to the controller ECU, and then When the evaluation value is greater than the preset rollover threshold, a trigger signal, that is, a front wheel differential braking signal is sent to the front wheel differential braking module;

The braking pressure of the inner front wheels of the vehicle is constant, and the front wheel differential braking module is used to adjust the braking pressure of the outer front wheels of the vehicle according to the received braking pressure difference signal, so that the difference between the braking pressures of the two front wheels and The received front wheel differential braking signal is corresponding;

The control ECU is connected to the rollover evaluation unit and the front wheel differential brake module respectively, and is used to combine the received rollover evaluation value with the preset rollover threshold value when the received rollover evaluation value is greater than the preset rollover threshold value. The differential braking signal of the front wheel is obtained by making a difference at the turning threshold, and the signal of the differential braking pressure is output to the differential braking module of the front wheel.

The implementation of the vehicle anti-rollover control strategy includes the following steps:

Step 1), the steering wheel angle sensor senses the steering wheel angle of the vehicle, and transmits the steering wheel angle value to the steering wheel angle estimation unit; the vehicle speed sensor senses the vehicle speed, and transmits it to the vehicle parameter estimation unit;

Step 2), the steering wheel angle estimation unit estimates the steering wheel input value at the future time based on the steering wheel angle value at the current moment and the gray prediction model theory, and outputs the estimated value to the vehicle parameter estimation unit for vehicle rollover related parameters Prediction;

Step 3), the car takes the obtained steering wheel predicted value and vehicle speed as input, calculates the predicted values of various parameters of the car (predicted value of vehicle body lateral acceleration, predicted value of vehicle body roll angle, predicted value of vehicle body roll angle velocity), and which is input to the rollover evaluation unit;

Step 4), the rollover evaluation unit calculates the rollover evaluation value based on the received predicted value of vehicle body roll angle, predicted value of vehicle body roll angular velocity, and predicted value of vehicle body lateral acceleration, and then transmits it to the controller ECU, and When the evaluation value is greater than the preset rollover threshold, a trigger signal is sent to the front wheel differential braking module;

Step 5), controlling the ECU to compare the received rollover evaluation value with the preset rollover threshold;

Step 5.1), if the received rollover evaluation value is greater than the preset rollover threshold;

Step 5.1.1), making a difference between the received rollover evaluation value and the preset rollover threshold to obtain a rollover difference, and calculating a front wheel differential braking signal according to the rollover difference;

Step 5.1.2), output the brake pressure difference signal to the front wheel differential brake module.

Step 6), the front wheel differential braking adjusts the braking pressure of the outer front wheels of the vehicle according to the received braking pressure difference signal, so that the difference between the two front wheel braking pressures corresponds to the received braking pressure difference signal , to brake the vehicle.

As shown in Figure 2, as a supplementary description, the steering wheel angle estimation unit adopts a driver's steering wheel input estimation model based on gray prediction, using the GM (1, 1) model, and the realization of step 2) has the following steps:

Step 2.1), generate the original input data sequence of the prediction model:

X ⁽⁰⁾ = [x ⁽⁰⁾ (1), x ⁽⁰⁾ (2), ..., x ⁽⁰⁾ (n)]

In the formula: n is the modeling dimension, indicating the amount of historical data used in forecasting; x ⁽⁰⁾ (i), i=1, 2,..., n is the input of the forecasting model.

Step 2.2), the original data can be accumulated once:

X ⁽¹⁾ = [x ⁽¹⁾ (1), x ⁽¹⁾ (2), ..., x ⁽¹⁾ (n)]

In the formula

Step 2.3), form the data matrix B and Y:

Step 2.4), calculate the development coefficient with

Step 2.5), according to the GM(1, 1) model, the predicted value at time j to j+k time Expressed as:

In the formula: is the predicted value at j-n+1 moment; k is the number of steps in advance prediction.

Step 2.6), taking the steering wheel angle as the input, when the sampling time Ts is 0.02s, the number of advance prediction steps k is 5, which means that the advance prediction time T (T=Ts×k) is 0.1s, and can be calculated in step 5 (after 0.1s) the steering wheel angle value.

As a supplementary explanation, the vehicle parameter estimation unit is connected with the steering wheel angle estimation unit and the vehicle speed sensor, and the obtained steering wheel prediction value and vehicle speed are used as input to calculate the prediction values of various parameters of the vehicle (body lateral acceleration prediction value , predicted value of body roll angle, predicted value of body roll angle velocity). Step 3) The specific implementation steps are as follows:

Step 3.1), build a 3-DOF car model. The invention establishes a linear three-degree-of-freedom vehicle rollover model.

Step 3.1.1), vehicle model

The model ignores the dynamic characteristics of the car in the longitudinal and pitch directions, and assumes that the dynamics of the left and right wheels of the car are symmetrical about the X axis, that is, the model consists of a "bicycle model" and a roll plane model, including lateral motion along the Y direction, Yaw motion around the Z axis, and roll motion around the X axis.

Considering the coupling effects among the three degrees of freedom, the vehicle rollover dynamics equation can be obtained according to D'Alembert's principle.

Lateral movement:

Lateral movement:

Rolling movement:

From the yaw and lateral motion coupling relationship, the lateral acceleration at the center of mass of the vehicle can be obtained as:

In the formula: a and b are the distances from the center of mass of the car to the front and rear axles; F _f and F _r are the cornering forces of the front and rear wheels respectively; h is the distance from the roll center to the center of mass; I _x is the roll moment of inertia of the sprung mass; I _z is the yaw moment of inertia; with is the suspension equivalent roll stiffness and equivalent roll damping coefficient; m is the mass of the vehicle; m _s is the sprung mass; is the additional yaw moment output by the control system; r is the yaw rate; v is the lateral velocity; is the roll angle of the sprung mass.

Step 3.1.2) build tire model.

Considering the effects of roll steering, roll camber, deformation steering and deformation camber on tire lateral characteristics, from the relationship between lateral force, speed and rotation angle, the slip angle of front and rear wheels can be obtained as:

In the formula: k _a is the camber coefficient, δ _sw is the steering wheel angle value, and i is the vehicle transmission ratio.

Then the cornering force of the front and rear wheels can be obtained as

F _f =k _f β _f

F _r =k _r β _r

In the formula: _kf is the cornering stiffness of the front wheel, and _kr is the cornering stiffness of the rear wheel.

Step 3.2), according to the predicted value of the steering wheel angle, the vehicle speed and the above formula, calculate the predicted value of the lateral acceleration of the vehicle body, the predicted value of the body roll angle and the predicted value of the vehicle body roll angle velocity.

As a supplementary explanation, the vehicle rollover evaluation unit calculates the rollover evaluation value according to the received predicted value of the vehicle body roll angle, the predicted value of the vehicle body roll angular velocity, and the predicted value of the vehicle body lateral acceleration. The specific steps of step 4) are as follows:

Step 4.1) The vertical load change of the wheel can be described by the lateral load transfer rate (Lateral-load transfer rate, LTR), and the lateral load transfer rate can evaluate the stable state of the vehicle. Therefore, the lateral load transfer rate is used as the rollover factor of the vehicle. The lateral load transfer rate is defined as the ratio of the difference between the vertical load of the left and right tires to the total vertical load of the vehicle, and the expression is as follows:

In the formula, F _zl and F _zr are the left and right vertical loads of the vehicle respectively, and it can be seen that:

F _zr +F _zl =mg

In the formula, m is the total mass of the car. When the car rolls, the vertical load is redistributed on the left and right wheels. Obviously, the value of LTR is between [-1,l]. When LTR=0, the car has no roll. Rollover will not occur. When LTR=±1, the vertical load on one side of the wheel is 0, and the tire on one side is off the ground, so there is a danger of rollover.

Since the occurrence of rollover needs to be predicted in advance, the rollover threshold is taken as 0.85.

Step 4.2), the force balance equation of the car around the roll center is:

The force balance equation of the car around the center point of the wheel base on the ground is:

The lateral load transfer rate of the car can be obtained as:

In the formula: t _w is the wheel base, h _R is the distance from the roll center to the ground.

The control ECU is respectively connected with the rollover evaluation unit and the front wheel differential brake module, and is used to combine the received rollover evaluation value with the preset rollover threshold when the received rollover evaluation value is greater than the preset rollover threshold value. The differential brake pressure signal is obtained by making a difference of the rollover threshold value, and the brake pressure differential signal is output to the front wheel differential brake module. Its specific implementation steps are as follows:

Step 5.1), if the received rollover evaluation value is greater than the preset rollover threshold;

Step 5.1.1), making a difference between the received rollover evaluation value and the preset rollover threshold value to obtain a rollover difference value, and obtaining a brake pressure difference signal through PID control calculation according to the rollover difference value;

Step 5.1.2), outputting the brake pressure difference signal to the front wheel differential brake module.

As shown in Figure 3, the front wheel differential braking module is used to adjust the brake pressure of the outer front wheels of the vehicle according to the received brake pressure difference signal, so that the difference between the two front wheel brake pressures and the received corresponding to the front wheel differential braking signal. The specific implementation steps are as follows:

Step 6.1), the braking pressure of the front wheels inside the vehicle is constant, and the front wheel differential braking module adjusts the braking pressure value of the steering outer front wheel according to the braking pressure difference calculated by the ECU;

Step 6.2), the lateral force of the wheel before braking is F _f , after the braking pressure is applied, the ground braking force of the wheel is F _x1 , and the additional yaw moment generated by the braking force is:

After applying the braking force F _x1 to the outer front wheel, due to the limitation of the friction circle, the lateral force of the wheel will decrease, thus generating an additional yaw moment:

ΔM _y =-(F _y1 -F' _y1 )a

The anti-rollover yaw moment acting on the vehicle is obtained as:

As shown in Figure 3, the differential braking module obtains the anti-rollover yaw moment value and acts on the vehicle to achieve the anti-rollover control effect.

The differential braking module is connected to the ECU and the rollover evaluation unit. The differential braking module includes a differential braking decision maker and the outer front wheel of the vehicle, the differential braking decision maker is connected with the ECU and the rollover evaluation unit, and the outer front wheel is connected with the differential braking decision maker and the vehicle. The rollover evaluation unit sends the trigger signal and LTR estimated value to the differential braking decision maker and ECU respectively. After calculation, the ECU outputs the brake pressure difference to the differential braking decision maker. , output the lateral force F _f of the wheel to the outer front wheel of the vehicle, and the outer front wheel of the vehicle, through the action of the lateral force F _f of the wheel, forms a ground braking force F _x1 with the ground, and the ground braking force generates a yaw moment ΔM, which acts on the vehicle .

Example

The simulation working condition is selected as J-Turn test. In the J-Turn test, the initial speed of the vehicle is 100km/h, the road adhesion coefficient is 0.85, and the maximum steering wheel angle is 180°. The steering wheel angle input is shown in Figure 4:

The flow shown in Figure 1:

(1) The sensor receives the vehicle steering wheel angle signal and outputs the steering wheel angle δ _sw to the steering wheel estimation unit, receives the vehicle speed signal v and outputs it to;

(2) The steering wheel angle estimation unit calculates and estimates the steering wheel input value δ _swf at a future moment, and outputs the estimated value to the vehicle parameter estimation unit to predict vehicle rollover related parameters.

(3) The vehicle parameter estimation module calculates the vehicle body lateral acceleration prediction value a _yf and body roll angle prediction value according to the steering wheel angle prediction value δ _swf and vehicle speed v and body roll rate prediction values

(4) The rollover evaluation unit predicts the value according to the received body roll angle Body roll rate prediction value The predicted value of lateral acceleration a _yf of the vehicle body calculates the rollover evaluation value LTR _f and transmits it to the controller ECU, and sends a trigger signal to the front wheel differential system when the rollover evaluation value LTR is greater than the preset rollover threshold LTRm moving module;

(5) The controller ECU receives the rollover evaluation prediction value LTR _f , and outputs the brake pressure difference signal Δp to the front wheel differential brake module;

(6) The front wheel differential braking module adjusts the braking pressure of the outer front wheels of the vehicle according to the received front wheel differential braking signal Δp, so that the difference between the two front wheel braking pressures and the received front wheel difference The vehicle brakes in response to the dynamic braking signal, reducing the risk of rollover.

Step 6.1), Fig. 2 is the workflow of the steering wheel estimation unit. This working principle is mainly based on the gray prediction model to predict the steering wheel input to compensate for the time lag of the control system.

In the specific implementation, the steering wheel angle is used as the input of the gray prediction model, and the time that needs to be predicted in advance is determined according to the time lag of the control system, and the estimated value of the steering wheel angle in the future is output by the gray prediction model.

Through literature review and technical consultation, it can be known that the time lag of the vehicle anti-rollover control system is generally 0-0.2s. The longer the forecast time in advance, the lower the forecast accuracy. It can be seen from the gray forecasting model that the forecasting time in advance is proportional to the sampling time and the number of forecasting steps in advance, and has nothing to do with the modeling dimension. With the increase of the forecast time in advance, the fluctuation of the forecast curve increases and the forecast accuracy decreases, but within the forecast time of 0.2s, the forecast accuracy is still at a high level. The working process of the steering wheel angle estimation unit is as follows:

(1) The input value of the steering wheel angle between 2s and 3s is the original input object, and the sampling time interval is 0.2s to generate the original input data sequence of the prediction model:

X ⁽⁰⁾ = [0, 45, 87, 137, 180]

(2) Accumulate the original data once and get:

X ⁽¹⁾ = [x ⁽¹⁾ (1), x ⁽¹⁾ (2), ..., x ⁽¹⁾ (5)]

In the formula

(3) Form data matrix B and Y:

(4) Calculate the development coefficient with

(5) When the sampling time Ts is 0.02s, the number of advance prediction steps k is 5, which means that the advance prediction time T (T=Ts×k) is 0.1s, and the steering wheel angle at step 5 (after 0.1s) can be calculated value. According to the GM(1,1) model, the predicted value of the steering wheel angle value in step 5 (after 0.1s) Expressed as:

As shown in Figure 3, the differential braking module is connected to the ECU and the rollover evaluation unit. The differential braking module includes a differential braking decision maker and the outer front wheels of the vehicle, the differential braking decision maker is connected with the ECU and the rollover evaluation unit, and the outer front wheels are connected with the differential braking decision maker and the vehicle connected. The rollover evaluation unit sends the trigger signal and LTR estimated value to the differential braking decision maker and ECU respectively. After calculation, the ECU outputs the brake pressure difference to the differential braking decision maker. , output the lateral force F _f of the wheel to the outer front wheel of the vehicle, and the outer front wheel of the vehicle, through the action of the lateral force F _f of the wheel, forms a ground braking force F _x1 with the ground, and the ground braking force generates a yaw moment ΔM, which acts on the vehicle . The braking pressure of the front wheels inside the vehicle is constant, and the front wheel differential braking module adjusts the braking pressure value of the steering outer front wheel according to the braking pressure difference calculated by the ECU;

Step 6.2), the lateral force of the wheel before braking is F _f , after the braking pressure is applied, the ground braking force of the wheel is F _x1 , and the additional yaw moment generated by the braking force is:

After applying the braking force F _x1 to the outer front wheel, due to the limitation of the friction circle, the lateral force of the wheel will decrease, thus generating an additional yaw moment:

ΔM _y =-(F _y1 -F' _y1 )a

The anti-rollover yaw moment acting on the vehicle is obtained as:

By adding the anti-rollover yaw moment, the rollover risk of the vehicle is effectively reduced.

The time lag of the vehicle anti-rollover control system is generally 0-0.2s. In order to avoid the influence of the time-lag, the control strategy of the present invention uses the gray prediction model to predict the rollover state of the vehicle after 0.1s and based on this The vehicle performs anti-rollover control earlier to reduce the risk of vehicle rollover.

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications are also possible. It should be regarded as the protection scope of the present invention.

Claims (7)

1. a kind of anti-rollover system for automobiles based on driver's input prediction, it is characterised in that：Including the sensor being sequentially connected Module, steering wheel angle estimate unit, automobile parameter and estimate unit and rollover evaluation unit, wherein, the sensor assembly is also Unit is estimated with automobile parameter to be connected, sensor assembly includes vehicle speed sensor, steering wheel angle sensor, and vehicle is measured respectively Speed and steering wheel angle；The steering wheel angle is estimated unit and is connected with steering wheel angle sensor, the automobile parameter Estimate that unit estimates unit with steering wheel angle and vehicle speed sensor is connected, the rollover evaluation unit respectively with steering wheel angle Estimate unit and estimate unit with automobile parameter and be connected；

The rollover evaluation unit exports gained information to front-wheel differential braking module and control ECU respectively, wherein output is touched Differential braking module is signaled to, rollover evaluation of estimate is exported to ECU；The control ECU exports brake pressure difference signal to front-wheel Differential braking module.

2. the control strategy of the anti-rollover system for automobiles according to claim 1 based on driver's input prediction, its feature It is：Comprise the steps of：

Step 1), steering wheel angle sensor senses the steering wheel angle of vehicle, and steering wheel angle value is passed into steering wheel Corner estimates unit；Vehicle speed sensor senses the speed of vehicle, and passes it to automobile parameter and estimate unit；

Step 2), steering wheel angle estimates steering wheel angle value and grey forecasting model theory of the unit based on current time, in advance Estimate the steering wheel input value of future time instance, and this discreet value is output to automobile parameter and estimate unit progress vehicle side turning correlation ginseng Several predictions；

Step 3), automobile parameter estimates unit by obtained steering wheel predicted value and speed as input, and it is each that calculating obtains automobile The predicted value of item parameter, including vehicle body side acceleration predicted value, vehicle roll angle predicted value, vehicle roll angle prediction of speed Value, and it is entered into rollover evaluation unit；

Step 4), rollover evaluation unit is according to the vehicle roll angle predicted value received, vehicle roll angle rate predictions, vehicle body Side acceleration predictor calculation goes out evaluation of estimate of turning on one's side, and passes it to controller ECU, and is more than in the rollover evaluation of estimate Trigger signal is sent during default rollover threshold and gives front-wheel differential braking module；

Step 5), control ECU compares the rollover evaluation of estimate received and default rollover threshold, if the rollover received Evaluation of estimate is more than default rollover threshold, then exports brake pressure difference signal and give front-wheel differential braking module；

Step 6), front-wheel differential braking module adjusts the braking pressure of the outer front-wheel of vehicle according to the brake pressure difference signal received Power so that the difference of two front wheel braking pressures and the front-wheel differential braking signal received are corresponding, are braked to vehicle.

3. the control strategy of the anti-rollover system for automobiles according to claim 2 based on driver's input prediction, its feature It is：Step 2) grey forecasting model is：Using gray prediction GM (1,1) model, by the steering wheel angle at current time Value is as input value, and when sampling time Ts is 0.02s, look-ahead step number k takes 5, represent look-ahead time T (T=Ts × K) it is 0.1s, the steering wheel angle value after the 5th i.e. 0.1s of step can be calculated；It is input to take the steering wheel angle that sensor is obtained Amount, using the steering wheel angle value after grey forecasting model look-ahead 0.1s, and is output to automobile parameter and estimates unit.

4. the control strategy of the anti-rollover system for automobiles according to claim 2 based on driver's input prediction, its feature It is：The step 3) to implement step as follows：

Step 3.1), set up a linear Three Degree Of Freedom vehicle side turning model

Step 3.1.1), auto model

The model ignores automobile longitudinal and the dynamic characteristic of pitch orientation, and assumes that vehicle right and left wheel dynamics is on X-axis Symmetrical, i.e., model is made up of " bicycle model " and roll plane model, including along the transverse movement of Y-direction, horizontal stroke about the z axis Pendular motion, and around the roll motion of X-axis；

Consider the coupling influence between 3 frees degree, vehicle side turning kinetics equation is：

Transverse movement：

Weaving：

Roll motion：

Can obtain Location of Mass Center of Automobiles transverse acceleration by yaw and transverse movement coupled relation is：

In formula：A and b are respectively distance of the automobile barycenter to antero posterior axis；F_fAnd F_rRespectively front and back wheel lateral deviation power；H is roll center To centroid distance；I_xFor the inclination rotary inertia of spring carried mass；I_zFor yaw rotation inertia；WithRolled just for suspension is equivalent Degree and equivalent inclination damped coefficient；M is complete vehicle quality；m_sFor spring carried mass；The additional yaw power exported for control system Square；R is yaw velocity；V is lateral velocity；For spring carried mass angle of heel；U is automobile longitudinal speed；a_yFor automobile side angle plus Speed；F_fAnd F_rFor front and back wheel lateral deviation power；

Step 3.1.2) set up tire model

Consider roll steer, roll flare, the influence to the lateral characteristic of tire of deflection steer and deformation flare, by side force with Speed and angle relation, the side drift angle that can obtain front and back wheel is：

In formula：k_aFor camber coefficient, δ_swFor steering wheel angle value, i is system of vehicle transmission ratio；

The lateral deviation power that then can obtain front and back wheel is：

F_f=k_fβ_f

F_r=k_rβ_r

In formula：k_fFor front-wheel cornering stiffness, k_rFor trailing wheel cornering stiffness；

Step 3.2), according to steering wheel angle predicted value, speed and above-mentioned formula, calculate the vehicle body side acceleration for obtaining automobile Predicted value, vehicle roll angle predicted value and vehicle roll angle rate predictions.

5. the control strategy of the anti-rollover system for automobiles according to claim 2 based on driver's input prediction, its feature It is：The step 4) comprise the following steps that：

Step 4.1) using the transverse load rate of transform as automobile the rollover factor, the transverse load rate of transform be defined as left and right tire hang down The ratio between difference and the total vertical load of vehicle of straight load, expression formula is as follows：

F in formula_zlAnd F_zrRespectively vehicle right and left vertical load, and understanding：

F_zr+F_zl=mg

M is automobile gross mass in formula；LTR value is [- 1, l], and as LTR=0, automobile is not rolled, and would not also occur side Turn over, when LTR=± 1, it is 0 to have single wheel vertical load, and rollover occurs for side tire liftoff dangerous；

Step 4.2), automobile is around the stress balance equation of roll center：

Automobile is around the stress balance equation for the wheelspan central point resting on the ground：

The lateral load rate of transform that automobile can be obtained is：

In formula：t_wFor wheelspan, h_RFor the distance of roll center to ground.

6. the control strategy of the anti-rollover system for automobiles according to claim 2 based on driver's input prediction, its feature It is：Step 5) specific method be：Control ECU compares the rollover evaluation of estimate received and default rollover threshold,

Step 5.1), if the rollover evaluation of estimate received is more than default rollover threshold；

Step 5.1.1), the rollover evaluation of estimate received and default rollover threshold are obtained into difference of turning on one's side as difference, according to rollover Mathematic interpolation obtains brake pressure difference signal；

Step 5.1.2), brake pressure difference signal is exported to the front-wheel differential braking module.

7. according to the control strategy of any described anti-rollover system for automobiles based on driver's input prediction of claim 2 to 6, It is characterized in that：Step 5) described in rollover threshold be 0.7-0.9.