Wheel anti-slip control method for driverless vehicle

Abstract

The present invention discloses a wheel anti-slip control method for a driverless vehicle, comprising: S1, starting and running the driverless vehicle; S2, system initialization; S3, system startup self-check; S4, collecting engine speed and wheel speed signals; S5, signal processing; S6, reading the wheel speed of the current driving wheel and reading the wheel speed of the driving wheel according to the wheel speed data obtained in S5; S7, calculating the current vehicle speed; S8, estimating wheel speed and vehicle speed; S9, slip detection, if slipping, respectively execute S10 and S14; if not slipping, return to S4; S10, tire model identification; S11, calculating the optimal slip ratio; S12, throttle fuzzy PID control; S13, adjusting the throttle opening, and then returning to S4; S14, brake pressure fuzzy control; S15, brake system anti-slip control, and then returning to S4. This method effectively suppresses the wheel slipping problem of the driverless vehicle and improves the driving safety of the vehicle.

Description

Technical Field

[0001] The present invention relates to the field of driverless, and particularly to a method for controlling wheel anti-slip of a driverless vehicle. Background Art

[0002] Driverless technology has been increasingly widely applied in the entire coal mine production system due to its advantages of being automatic, efficient, safe, and low-cost. However, different from the road surface in cities, with the deepening of the mine depth, the roadway environment in the coal mine is harsh, the lighting is insufficient, and the road conditions are complex. In particular, the wet ground underground may cause the wheels to slip. The control instability and braking instability caused by wheel slip are extremely likely to lead to safety accidents. Therefore, anti-slip control must be carried out in the driverless system of the mine car to improve the safety and stability of the whole vehicle.

[0003] In the prior art, many experts and scholars have proposed some methods for controlling wheel anti-slip of driverless vehicles. Some experts have proposed a speed control method for suppressing the wheel slip of a mine-used driverless electric locomotive, which reduces the influence of load mutation on the motor speed by adding a feedback term to the output of the motor control, thereby suppressing the wheel slip phenomenon caused by the change of adhesion force of the electric locomotive. However, this method is designed for the control of the motor and cannot be applied to fuel vehicles.

[0004] Some scholars have proposed a method and system for identifying and controlling wheel slip of a driverless vehicle. This method judges whether the wheel is in a slip state according to the EBS vehicle speed or wheel speed. If it is in a slip state, active anti-slip control is achieved by performing gear limit control, differential lock active control, and engine throttle active control. However, this solution only judges whether slip occurs and does not evaluate and quantify the slip situation, so high-precision control cannot be achieved. Summary of the Invention

[0005] Aiming at the problems existing in the prior art, the present invention provides a method for controlling wheel anti-slip of a driverless vehicle to improve the stability and safety of the driverless vehicle when driving on a wet ground.

[0006] For this reason, the present invention adopts the following technical solutions:

[0007] A method for controlling wheel anti-slip of a driverless vehicle includes the following steps:

[0008] S1, the driverless vehicle starts to run;

[0009] S2, system initialization: including reading the vehicle mass, front wheel wheelbase, rear wheel wheelbase, drive wheel centroid height, and initial parameters of the fuzzy PID controller;

[0010] S3, System startup self-check: Check whether each sensor, inertial navigation system, global positioning system, and lidar are working properly;

[0011] S4, Engine speed and wheel speed signal acquisition: The power domain controller reads the engine speed and wheel speed signals from the crankshaft position sensor and the wheel speed sensor;

[0012] S5, Signal processing: Use the EKF algorithm to filter the noise of the engine speed and wheel speed signals, and fit the filtered signals based on the cubic spline function to ensure that the engine speed and wheel speed do not mutate, and obtain smooth and continuous engine speed and wheel speed data;

[0013] S6, Read the wheel speed of the current driving wheel, and read the wheel speed of the driving wheel according to the wheel speed data obtained in step S5;

[0014] S7, Calculate the current vehicle speed. The driving speed of the vehicle is calculated by the following formula:

[0015]

[0016] Where: v is the driving speed of the vehicle, with the unit of km·h -1 ; r is the rolling radius of the wheel; n is the engine speed, with the unit of r·m -1 , the engine speed is measured by the crankshaft position sensor, and the detection signal is sent to the power domain controller; i g is the transmission ratio; i 0 is the final drive ratio;

[0017] S8, Wheel speed and vehicle speed estimation: Fuse the driving wheel speed and vehicle speed obtained in steps S6 and S7 with the inertial navigation system, global positioning system, and lidar to obtain the fused driving wheel speed ω 融 and the driving speed v_fusion of the vehicle;

[0018] S9, Slip detection: Calculate the current slip ratio s current according to the wheel speed and vehicle speed obtained in step S8. The formula is as follows:

[0019]

[0020] Among them, ω 融 is the fused driving wheel speed; r is the rolling radius of the driving wheel, and v 融 is the fused driving speed of the vehicle;

[0021] Judge the slip condition according to the current slip ratio. If the slip condition is met, execute steps S10 and S14 respectively; if not, return to step S4 to re-perform the slip detection;

[0022] S10, Tire model identification: The parameters of the tire model and the peak adhesion coefficient are estimated in real time by the least squares method;

[0023] S11, Calculate the optimal slip ratio. Substitute the peak adhesion coefficient μ obtained in step S10 peak into the Burckhardt tire model, and use the Newton method to iteratively solve the optimal slip ratio s target , The formula is as follows:

[0024]

[0025] S12, Throttle fuzzy PID control: The fuzzy PID controller dynamically adjusts the PID parameters according to the difference between the current slip ratio s current and the optimal slip ratio s target , and calculates the throttle opening control amount;

[0026] S13, Adjust the throttle opening: The ECU adjusts the throttle opening according to the throttle opening control amount, controls the engine torque output, and then returns to step S4;

[0027] S14, Brake pressure fuzzy control: Use the fuzzy algorithm to calculate the electronic hydraulic valve control amount for active braking control;

[0028] S15, Anti-skid control of the braking system: The hydraulic control unit outputs the electronic hydraulic valve control amount as a hydraulic control signal and sends it to the braking system. The braking system converts the hydraulic control signal into a braking pressure signal. After receiving the signal, the driverless vehicle performs active braking intervention on the slipping drive wheels to suppress the slipping of the single-side drive wheels, and then returns to step S4.

[0029] In the above step S9, the following three situations are determined as the vehicle being in a slipping state:

[0030] (1) The slip ratio of one or more drive wheels is higher than 50%, and this state is maintained for 15 s;

[0031] (2) The slip ratio of one or more drive wheels is higher than 70%, and this state is maintained for 10 s;

[0032] (3) The slip ratio of one or more drive wheels is higher than 90%, and this state is maintained for 5 s.

[0033] The above step S10 includes the following sub-steps:

[0034] (1) Calculate the normal load of the drive wheel:

[0035] F zf =(mgb - mah g ) / 2L

[0036] In the above formula, F zf is the normal load of the driving wheel; m is the vehicle mass; g is the acceleration due to gravity; b is the distance from the rear axle to the center of gravity; a is the vehicle longitudinal acceleration, calculated by the inertial measurement unit; h g is the height of the center of mass; L is the wheelbase;

[0037] (2) Substitute the current slip ratio s current described in step S9 into the Burckhardt tire model formula determined by parameters to obtain the road surface adhesion coefficient μ. The expression of the Burckhardt tire model is:

[0038]

[0039] In the formula, C 1 , C 2 , C 3 are all parameters in the Burckhardt tire model expression; the value range of the current slip ratio s current is 0 to 1;

[0040] Multiply the road surface adhesion coefficient by the normal load to obtain the ground driving forces of the left and right driving wheels of the driverless vehicle; take 2 times the smaller driving force of the two driving wheels as the total driving force of the current vehicle, and divide it by the vehicle body mass to obtain the estimated acceleration value of the vehicle The calculation formula is:

[0041]

[0042] (3) Preset multiple tire models with different peak adhesion coefficients, and select the tire model closest to the current tire from them:

[0043] Discretize the peak adhesion coefficient into multiple groups, calculate multiple groups of estimated acceleration values a * according to the tire model parameters of each group, and use the least squares method to calculate the sum of the squares of the errors between the acceleration reference value a and a * within 3 - 5 s. Select a set of C 1 , C 2 , C 3 parameters with the smallest sum of the squares of the errors as the tire model identification result. Among them, the formula of the least squares method is as follows:

[0044]

[0045] In the formula, is the acceleration estimated value of the driverless vehicle at the i-th moment; a i is the acceleration measured value of the driverless vehicle at the i-th moment;

[0046] (4) The tire model parameters C determined through step (3) 1 、C 2 、C 3 Look up the table to obtain the peak adhesion coefficient.

[0047] In the above step S12, the throttle fuzzy PID control includes the following sub-steps:

[0048] 1) Calculate the current slip ratio deviation e(k) of the driving wheel and the change rate of the current slip ratio deviation The formula is as follows:

[0049] e(k) = s current (k) - s target ,

[0050]

[0051] where k is the current moment; k - 1 is the previous moment;

[0052] 2) The input signals of the fuzzy control are the current slip ratio deviation e(k) and the change rate of the current slip ratio deviation The output signals are the proportional coefficient K of the PID controller p 、the integral coefficient K i 、the differential coefficient K d ,Through the current slip ratio deviation e(k) and the change rate of the current slip ratio deviation The fuzzy subsets are divided into {Negative Big (NB), Negative Medium (NM), Negative Small (NS), Zero (ZE), Positive Small (PS), Positive Medium (PM) and Positive Big (PB)}, and the membership function uses the triangular membership function type;

[0053] 3) Establish a fuzzy control rule table in the form of "if - then" to obtain the corresponding fuzzy control rules;

[0054] 4) In the fuzzy inference process, use Mamdani as the composition operation rule of the fuzzy relation and the fuzzy set to calculate the fuzzy output quantity;

[0055] 5) In the defuzzification link, use the maximum membership reading method and the average value method, select the output value corresponding to the maximum membership degree and take the average value of all output values, and convert the output fuzzy quantity into a clear value K p , K i , K d ;

[0056] 6) After the PID controller updates the parameters, calculate the throttle opening control quantity.

[0057] In the above step S13, the specific method for controlling the engine torque output is:

[0058] The fuzzy PID sends the control quantity signal to the electronic throttle control unit via the CAN bus; the electronic throttle control unit converts the control quantity signal into the starting angle of the throttle and sends the corresponding voltage signal to the drive motor; the drive motor controls the throttle to reach the target opening position.

[0059] Step S14 specifically includes the following sub-steps:

[0060] 1) Calculate the wheel speed difference E and the change rate of wheel speed difference EC between the two drive wheels. The formulas are as follows:

[0061] E = ω L - ω R

[0062] EC(k) = E(k) - E(k - 1)

[0063] Where ω L is the wheel speed of the left drive wheel, and ω R is the wheel speed of the right drive wheel;

[0064] 2) Divide the fuzzy subsets by the wheel speed difference E and its change rate EC into: {Negative Big (NB), Negative Medium (NM), Negative Small (NS), Zero (ZE), Positive Small (PS), Positive Medium (PM), and Positive Big (PB)}, and the membership function uses the triangular membership function type;

[0065] 3) Determine the fuzzy control rules based on manual experience. The fuzzy control rule table is established in the form of "if - then", and the corresponding fuzzy control rules can be obtained according to the look-up table;

[0066] 4) Use the maximum membership reading method to take out the output value corresponding to the maximum membership degree, and take the average value of all output values, so as to convert the output fuzzy quantity into the electronic hydraulic valve control quantity u.

[0067] Compared with the prior art, the present invention has the following beneficial effects:

[0068] 1. The present invention designs a wheel anti-skid control method for a driverless vehicle based on an autonomous driving domain controller and a power domain controller. The control method can identify the road surface condition and the trend of excessive wheel spin according to the wheel speed, and actively brake and intervene in the drive wheels by adjusting the throttle opening of the engine and the braking pressure of the braking system, so that the wheel slip rate is controlled within a reasonable range, thereby effectively suppressing the wheel slip problem of the driverless vehicle caused by wet ground and improving the driving safety of the vehicle.

[0069] 2. In the engine controller part of the present invention, a parameter adaptive PID controller based on a fuzzy control algorithm is used, which improves the algorithm convergence speed, can effectively suppress the sudden change of the slip ratio, and quickly controls the driving force of the driverless vehicle to the optimal value. In the brake system controller part of the present invention, a fuzzy control algorithm is used, which makes the adaptability of the braking effect stronger and effectively suppresses the slipping phenomenon of the driving wheels during the acceleration and starting processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 is a schematic structural diagram of the wheel anti-skid control system of the driverless vehicle in the present invention;

[0071] Figure 2 is a flowchart of the wheel anti-skid control method of the driverless vehicle in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0072] The technical solutions of the present invention will be described in detail below with reference to the drawings and embodiments.

[0073] See Figure 1 , in the present invention, the wheel anti-skid control system of the driverless vehicle includes: sensors (including wheel speed sensors, crankshaft position sensors), a power domain controller (Vehicle Control Unit, VCU), an autonomous driving domain controller (Automatic Driving Control Unit, ADCU), an electronic throttle control unit, an engine management system (Engine Management System, EMS), a hydraulic control unit, and a brake system. Data interaction and synchronization are achieved between each sensor and control unit through a Controller Area Network (CAN). The driverless vehicle is equipped with an inertial navigation system (INS), a global positioning system (GPS), a lidar (LiDAR), and an inertial measurement unit (IMU). Among them:

[0074] The crankshaft position sensor and the wheel speed sensor read the engine speed and wheel speed information of the driverless vehicle within the sampling period and send them to the power domain controller for processing in the form of digital signals;

[0075] The power domain controller is used to read the engine speed and wheel speed signals from the crankshaft position sensor and the wheel speed sensor; calculate the current vehicle speed and wheel speed, calculate the current slip ratio according to the vehicle speed and wheel speed, and perform a skid judgment. If the anti-skid mode is entered, the autonomous driving domain controller calculates the optimal slip ratio of the current wheel; the power domain controller sends the electronic hydraulic valve control quantity to the electronic throttle control unit and the hydraulic control unit according to the optimal slip ratio and the left and right wheel speed differences.

[0076] After receiving the control quantity of the electronic hydraulic valve, the electronic throttle control unit converts it into a throttle control signal according to the control quantity and sends the throttle control signal to the engine management system. The engine management system controls the throttle opening to change the engine power so that the wheels can be in the target state of the optimal slip ratio.

[0077] After receiving the control quantity of the electronic hydraulic valve, the hydraulic control unit converts it into a brake pressure control signal according to the control quantity and sends the hydraulic control signal to the braking system.

[0078] The braking system adjusts the wheel speed difference between the left and right wheels by regulating the braking torque of the electro-hydraulic valve to suppress the slipping of the driving wheels.

[0079] Figure 2 It is a flowchart of the wheel anti-skid control method for the driverless vehicle of the present invention. As Figure 2 shown, the control method includes the following steps:

[0080] S1, the driverless vehicle starts to run;

[0081] S2, system initialization: including reading vehicle mass, front wheel wheelbase, rear wheel wheelbase, driving wheel centroid height, and initial parameters of the fuzzy PID controller;

[0082] S3, system startup self-check: check whether each sensor, inertial navigation system (INS), global positioning system (GPS), and lidar are working properly;

[0083] S4, engine speed and wheel speed signal acquisition: the power domain controller reads the engine speed and wheel speed signals from the crankshaft position sensor and the wheel speed sensor;

[0084] S5, signal processing: use the EKF algorithm to filter the noise of the engine speed and wheel speed signals, and fit the filtered signals based on the cubic spline function to ensure that the engine speed and wheel speed do not mutate, and obtain smooth and continuous engine speed and wheel speed data;;

[0085] S6, read the wheel speed of the current driving wheel, and read the wheel speed of the driving wheel according to the wheel speed data obtained in step S5;

[0086] S7, calculate the current vehicle speed, and the driving speed of the driverless vehicle is calculated by the following formula:

[0087]

[0088] In the formula: v is the driving speed of the vehicle, with the unit of km·h -1 ; r is the rolling radius of the wheel; n is the engine speed, with the unit of r·m -1, the engine speed is measured by the crankshaft position sensor, and the detection signal is sent to the power domain controller; i g is the transmission gear ratio; i 0 is the final drive ratio;

[0089] S8, Wheel speed and vehicle speed estimation: In practical applications, to solve the problem of inaccurate measurement values of wheel speed and vehicle speed, the driving wheel speed and vehicle speed obtained in steps S6 and S7 are combined with an inertial navigation system (INS), a global positioning system (GPS), and a lidar (LiDAR) for information fusion to obtain the fused wheel speed ω 融 and vehicle speed v 融 ;

[0090] S9, Slip detection, Based on the wheel speed and vehicle speed obtained in step S8, calculate the current slip ratio s current , The formula is as follows:

[0091]

[0092] where: ω 融 is the wheel speed of the fused driving wheel; r is the rolling radius of the driving wheel, v 融 is the fused vehicle driving speed;

[0093] When the driverless vehicle travels on a wet and slippery road surface, the driving wheels are prone to slipping and the direction is prone to getting out of control, and at this time, safety accidents are extremely likely to occur. For wet and slippery road surfaces, the driverless vehicle must perform anti-slip control. First, it is necessary to make a slip judgment. This method determines the following three situations as the vehicle being in a slip state:

[0094] (1) The slip ratio of one or more driving wheels is higher than 50%, and this state is maintained for 15 s;

[0095] (2) The slip ratio of one or more driving wheels is higher than 70%, and this state is maintained for 10 s;

[0096] (3) The slip ratio of one or more driving wheels is higher than 90%, and this state is maintained for 5 s;

[0097] If the above conditions are met, steps S10 and S14 are executed respectively; if not, return to step S4 to re-perform slip detection.

[0098] S10, Tire model identification: Identify the parameters of the tire model based on the vehicle operation feedback data, and estimate the parameters of the tire model and the peak adhesion coefficient in real time by the least squares method. The specific steps are as follows:

[0099] (1) Calculate the normal load of the driving wheel:

[0100] F zf=(mgb - mah g ) / 2L

[0101] In the above formula, F zf is the normal load of the driving wheel; m is the vehicle mass; g is the acceleration due to gravity; b is the distance from the rear axle to the center of gravity; a is the longitudinal acceleration of the vehicle, calculated by the inertial measurement unit (IMU); h g is the height of the center of mass; L is the wheelbase.

[0102] (2) Substitute the current slip ratio s current described in step S9 into the Burckhardt tire model formula determined by parameters to obtain the road surface adhesion coefficient μ. The expression of the Burckhardt tire model is:

[0103]

[0104] In the formula, C 1 , C 2 , C 3 are the parameters in the expression of the Burckhardt tire model; the value range of the current slip ratio s current is 0 to 1;

[0105] Multiply the road surface adhesion coefficient by the normal load to obtain the ground driving forces of the left and right driving wheels of the driverless vehicle; take twice the smaller driving force among the two driving wheels as the total driving force of the current vehicle, and divide it by the vehicle body mass to obtain the estimated acceleration value of the current vehicle The calculation formula is:

[0106]

[0107] (3) Preset multiple tire models with different peak adhesion coefficients, and select the tire model closest to the current tire from them.

[0108] In an embodiment of the present invention, the peak adhesion coefficient is discretized into 9 groups. The corresponding relationships between the peak adhesion coefficient and the tire model parameters C 1 , C 2 , C 3 are as follows:

[0109] Table 1

[0110] Peak adhesion coefficient <![CDATA[C 1 > <![CDATA[C 2 > <![CDATA[C 3 > 0.1 0.135 47 0.01 0.2 0.204 38 0.02 0.3 0.317 35 0.03 0.4 0.409 31 0.04 0.5 0.512 26 0.05 0.6 0.608 23 0.06 0.7 0.711 15 0.07 0.8 0.805 12 0.08 0.9 0.985 0. 0.09

[0111] Calculate 9 groups of estimated acceleration values a * according to the 9 groups of tire model parameters in the table, and use the least squares method to calculate the sum of the squares of the errors between the acceleration reference value a and the estimated acceleration value a * within 3 - 5 s, and take the group of parameters with the smallest sum of the squares of the errors as the tire model identification result.

[0112] The least squares formula is as follows:

[0113]

[0114] In the formula, is the estimated value of the acceleration of the driverless vehicle at the i-th moment; a i is the measured value of the acceleration of the driverless vehicle at the i-th moment.

[0115] (4) Through the tire model parameters C 1 , C 2 , C 3 determined in step (3), look up Table 1 to obtain the peak adhesion coefficient.

[0116] S11, calculate the optimal slip ratio. Substitute the peak adhesion coefficient μ peak obtained in step S10 into the Burckhardt tire model, and use the Newton method to iteratively solve the optimal slip ratio s target , the formula is as follows:

[0117]

[0118] S12, throttle fuzzy PID control: Use the difference between the current slip ratio s current and the optimal slip ratio s target as the error to input into the PID controller, and use the fuzzy control algorithm to online tune the PID parameters to make the error converge quickly, and then calculate the throttle opening control amount. The specific steps are as follows:

[0119] 1) Calculate the current slip ratio deviation e(k) and the current slip ratio deviation change rate of the driving wheel. The formula is as follows:

[0120] e(k) = s current (k) - s target ,

[0121]

[0122] where k is the current moment; k - 1 is the previous moment;

[0123] 2) The input signals of the fuzzy control are the current slip ratio deviation e(k) and the current slip ratio deviation change rate , and the output signals are the proportional coefficient K p , integral coefficient K i , and derivative coefficient K d of the PID controller. Through the current slip ratio deviation e(k) and the current slip ratio deviation change rate The fuzzy subsets can be divided into: {Negative Big (NB), Negative Medium (NM), Negative Small (NS), Zero (ZE), Positive Small (PS), Positive Medium (PM), Positive Big (PB)}. The membership function uses the triangular membership function type.

[0124] 3) The fuzzy control rule table is established in the form of "if-then", and the fuzzy control rule tables are shown in Tables 2, 3, and 4.

[0125] The corresponding fuzzy control rules can be obtained according to the look-up table.

[0126]

[0127] Table 2

[0128]

[0129] Table 3

[0130]

[0131]

[0132] Table 4

[0133] 4) In the fuzzy inference process, Mamdani is used as the compositional operation rule for the fuzzy relation and the fuzzy set to calculate the fuzzy output quantity.

[0134] 5) In the defuzzification link, the maximum membership reading method and the average value method are used to select the output value corresponding to the maximum membership degree and take the average value of all output values, so as to convert the output fuzzy quantity into a clear value K p ,K i ,K d .

[0135] 6) After the PID controller updates the parameters, the throttle opening control quantity is calculated.

[0136] S13, adjust the throttle opening to control the engine torque output:

[0137] The fuzzy PID sends the control quantity signal to the electronic throttle control unit through the CAN bus. The electronic throttle control unit converts the control quantity signal into the starting angle of the throttle and sends the corresponding voltage signal to the drive motor. The drive motor controls the throttle to reach the target opening position, thereby changing the engine output torque, making the driving wheels travel at the optimal slip ratio, improving the tire grip, and thus achieving the purpose of anti-skid.

[0138] After completing the above control, return to step S4 to re-collect the wheel speed and engine speed signals, detect whether there is skidding and perform anti-skid control.

[0139] S14, Brake Pressure Fuzzy Control:

[0140] When the driverless vehicle is accelerating and starting, a single driving wheel may come into contact with the standing water on the ground area and slip. Since the ground adhesion coefficients in contact with the two driving wheels are not equal, unilateral slipping may occur. Therefore, a fuzzy algorithm is used to calculate the control quantity of the electro-hydraulic valve for active braking control, adjust the driving force of the unilateral wheel, and improve the stability of the vehicle during starting and accelerating. The specific control method is as follows:

[0141] Calculate the wheel speed difference E and the change rate of wheel speed difference EC between the two driving wheels. The formulas are as follows:

[0142] E = ω L - ω R

[0143] EC(k) = E(k) - E(k - 1)

[0144] Among them, ω L is the wheel speed of the left driving wheel, and ω R is the wheel speed of the right driving wheel.

[0145] The fuzzy subsets are divided by the wheel speed difference E and its change rate EC into: {Negative Big (NB), Negative Medium (NM), Negative Small (NS), Zero (ZE), Positive Small (PS), Positive Medium (PM), Positive Big (PB)}. The membership function uses the triangular membership function type.

[0146] The fuzzy rules are determined based on manual experience. The fuzzy control rule table is established in the form of "if - then". The fuzzy control rules are shown in Table 5. The corresponding fuzzy control rules can be obtained according to the table lookup.

[0147]

[0148]

[0149] Table 5

[0150] Using the maximum membership reading method, take out the output value corresponding to the maximum membership degree, and average all the output values, so as to convert the output fuzzy quantity into the control quantity u of the electro - hydraulic valve.

[0151] S15, Anti - slip Control of the Braking System:

[0152] The hydraulic control unit outputs the control quantity of the electro - hydraulic valve obtained in step S14 into a hydraulic control signal and sends it to the braking system. The braking system converts it into a braking pressure signal. When the driverless vehicle receives the signal, it will perform active braking intervention on the slipping driving wheel to ensure the stability of the left and right driving wheels, inhibit the slipping of the unilateral driving wheel, enable the vehicle to obtain better ground friction, and improve the vehicle stability and anti - slip performance.

[0153] After the above control is completed, return to step S4 to collect the wheel speed and engine speed signals again, detect whether there is skidding and perform anti-skid control.

[0154] In the test experiment of wheel anti-skid control using the method of the present invention, the driverless mining truck (fuel truck) completed more than 6,000 hours of work tasks underground. During this period, no safety accidents caused by wheel skidding occurred. When driving into a wet and slippery road surface, the anti-skid control system will intervene in a timely manner to guide the vehicle to pass safely, thereby improving production efficiency and safety.

Claims

1. A wheel anti - slip control method for a driverless vehicle, comprising the following steps: S1, the driverless vehicle starts to run; S2, system initialization: including reading the vehicle mass, front wheel wheelbase, rear wheel wheelbase, drive wheel centroid height, and initial parameters of the fuzzy PID controller; S3, system startup self - inspection: checking whether each sensor, inertial navigation system, global positioning system, and lidar are working properly; S4, engine speed and wheel speed signal acquisition: the power domain controller reads the engine speed and wheel speed signals from the crankshaft position sensor and the wheel speed sensor; S5, signal processing: using the EKF algorithm to filter the noise of the engine speed and wheel speed signals, and fitting the filtered signals based on the cubic spline function to ensure that the engine speed and wheel speed do not have sudden changes, obtaining smooth and continuous engine speed and wheel speed data; S6, reading the wheel speed of the current drive wheel, and reading the wheel speed of the drive wheel according to the wheel speed data obtained in step S5; S7, calculating the current vehicle speed, and the driving speed of the vehicle is calculated by the following formula: Where: v is the driving speed of the vehicle, with the unit of km·h -1 ; r is the rolling radius of the wheel; n is the engine speed, with the unit of r·m -1 , and the engine speed is measured by the crankshaft position sensor, and the detection signal is sent to the power domain controller; i g is the transmission ratio; i 0 is the final drive ratio; S8, Estimation of wheel speed and vehicle speed: The driving wheel speed and vehicle speed obtained in steps S6 and S7 are fused with the inertial navigation system, global positioning system, and lidar to obtain the fused driving wheel speed ω 融 and the driving speed v of the vehicle 融 ; S9, Slip Detection: Calculate the current slip ratio s based on the wheel speed and vehicle speed obtained in step S8. current , and the formula is as follows: where ω 融 is the wheel speed of the fused driving wheel; r is the rolling radius of the driving wheel, and v 融 is the vehicle driving speed after fusion; Judging the skidding condition according to the current slip ratio. If the skidding condition is met, steps S10 and S14 are respectively executed; if not, return to step S4 to re - perform the skidding detection; S10, tire model identification: real - time estimating the parameters and peak adhesion coefficient of the tire model by the least - squares method; S11. Calculate the optimal slip ratio. Substitute the peak adhesion coefficient μ obtained in step S10 peak into the Burckhardt tire model, and use the Newton method to iteratively solve for the optimal slip ratio s target . The formula is as follows: S12, Throttle fuzzy PID control: The fuzzy PID controller dynamically adjusts the PID parameters according to the difference between the current slip ratio s current and the optimal slip ratio s target to calculate the throttle opening control amount; S13, adjusting the throttle opening: the ECU adjusts the throttle opening according to the throttle opening control amount, controls the engine torque output, and then returns to step S4; S14, brake pressure fuzzy control: using the fuzzy algorithm to calculate the electronic hydraulic valve control amount for active braking control; S15, brake system anti - slip control: the hydraulic control unit outputs the electronic hydraulic valve control amount as a hydraulic control signal and sends it to the brake system. The brake system converts the hydraulic control signal into a brake pressure signal. After receiving the signal, the driverless vehicle performs active braking intervention on the skidding drive wheel to inhibit the skidding of the single - side drive wheel, and then returns to step S4.

2. The wheel anti - slip control method for a driverless vehicle according to claim 1, characterized in that, in step S9, the following three situations are determined as the vehicle being in a skidding state: (1) The slip ratio of one or more drive wheels is higher than 50%, and this state is maintained for 15 s; (2) The slip ratio of one or more drive wheels is higher than 70%, and this state is maintained for 10 s; (3) The slip ratio of one or more drive wheels is higher than 90%, and this state is maintained for 5 s.

3. The wheel anti - slip control method for a driverless vehicle according to claim 1, characterized in that, step S10 includes the following sub - steps: (1) Calculating the normal load of the drive wheel: F zf = (mgb - mah g ) / 2L In the above formula, F zf is the normal load of the driving wheel; m is the vehicle mass; g is the acceleration due to gravity; b is the distance from the rear axle to the center of gravity; a is the longitudinal acceleration of the vehicle, which is calculated by the inertial measurement unit; h g is the height of the center of mass; L is the wheelbase; (2) Substitute the current slip ratio s described in step S9 current into the Burckhardt tire model formula determined by the parameters to obtain the road surface adhesion coefficient μ. The expression of the Burckhardt tire model is: where C 1 , C 2 , and C 3 are all parameters in the Burckhardt tire model expression; the value range of the current slip ratio s current is from 0 to 1; Multiply the road surface adhesion coefficient by the normal load to obtain the ground driving forces of the left and right drive wheels of the driverless vehicle; take twice the smaller driving force among the two drive wheels as the total driving force of the current vehicle, and divide it by the vehicle body mass to obtain the estimated acceleration value of the vehicle The calculation formula is as follows: (3) Presetting multiple tire models with different peak adhesion coefficients, and selecting the tire model closest to the current tire from them: The peak adhesion coefficient is discretized into multiple groups, and multiple groups of estimated acceleration values a are calculated according to the tire model parameters of each group * , and the least squares method is used to calculate the sum of the squared errors between the acceleration reference value a and a * within 3 - 5 s, and a group C 1 , C 2 , C 3 with the smallest sum of squared errors is taken as the tire model identification result. Among them, the formula of the least squares method is as follows: Wherein, is the estimated acceleration value of the driverless vehicle at the i-th moment; a i is the measured acceleration value of the driverless vehicle at the i-th moment; (4) The tire model parameters C determined by step (3) 1 , C 2 , C 3 Look up the table to obtain the peak adhesion coefficient.

4. The wheel anti - slip control method for a driverless vehicle according to claim 1, characterized in that, the throttle fuzzy PID control in step S12 includes the following sub - steps: 1) Calculate the current slip rate deviation e(k) of the driving wheel and the change rate of the current slip rate deviation The formula is as follows: e(k) = s current (k) - s target , where k is the current moment; k - 1 is the previous moment; 2) The input signals of the fuzzy control are the current slip ratio deviation e(k) and the change rate of the current slip ratio deviation The output signals are the proportional coefficient K p of the PID controller, the integral coefficient K i and the derivative coefficient K d . Through the current slip ratio deviation e(k) and the change rate of the current slip ratio deviation , the fuzzy subsets are divided into {Negative Big (NB), Negative Medium (NM), Negative Small (NS), Zero (ZE), Positive Small (PS), Positive Medium (PM) and Positive Big (PB)}, and the membership function uses the triangular membership function type; 3) Establish a fuzzy control rule table in the form of "if-then" to obtain the corresponding fuzzy control rules; 4) In the process of fuzzy inference, use Mamdani as the composition algorithm for fuzzy relations and fuzzy sets to calculate the fuzzy output quantity; 5) In the defuzzification step, the maximum membership reading method and the average value method are used to select the output value corresponding to the maximum membership degree and take the average value of all output values, converting the output fuzzy quantity into a clear value K p ,K i ,K d ; 6) After the PID controller updates the parameters, calculate the throttle opening control quantity.

5. The wheel anti-skid control method for an autonomous vehicle according to claim 1, characterized in that The specific method for controlling the engine torque output in step S13 is: The fuzzy PID sends the control quantity signal to the electronic throttle control unit through the CAN bus; the electronic throttle control unit converts the control quantity signal into the starting angle of the throttle and sends the corresponding voltage signal to the drive motor; the drive motor controls the throttle to reach the target opening position.

6. The wheel anti-skid control method for an autonomous vehicle according to claim 1, characterized in that Step S14 is specifically: 1) Calculate the wheel speed difference E and the wheel speed difference change rate EC between the two driving wheels. The formulas are as follows: E = ω L -ω R EC(k) = E(k) - E(k - 1) where ω L is the wheel speed of the left drive wheel, and ω R is the wheel speed of the right drive wheel; 2) Divide the fuzzy subsets by the wheel speed difference E and its change rate EC into: {Negative Big (NB), Negative Medium (NM), Negative Small (NS), Zero (ZE), Positive Small (PS), Positive Medium (PM), and Positive Big (PB)}, and the membership function uses the triangular membership function type; 3) Determine the fuzzy control rules based on manual experience. The fuzzy control rule table is established in the form of "if-then", and the corresponding fuzzy control rules can be obtained by looking up the table; 4) Use the maximum membership reading method to extract the output value corresponding to the maximum membership degree, and take the average value of all output values, so as to convert the output fuzzy quantity into the electronic hydraulic valve control quantity u.

Images