Development And Test Of Intention Recognition And Braking Force Distribution Control Strategies For Tractor-Semitrailer

Electronic braking system (EBS) of commercial heavy vehicle is a very advanced brakingsystem. It compensates for brake delay of conventional brake system greatly, and increase theresponse time of the braking system. Through braking by wire, EBS improves brake comfort andsafety significantly, and provides a huge space for commercial vehicle braking performanceenhancement. EBS is the development direction of the coming development of commercial vehiclebraking system.Base on the research of commercial vehicle electric braking systems theory and applicationfrom domestic to foreign, this Ph.D. dissertation aims at the improvement of braking comfort andsecurity. This dissertation proposed a braking intention recognition method, which was used as thebasis of brake assist and multi-axis brake force distribution control method. This provided atheoretical support for independent research and development of China’s commercial vehicleelectrical braking system. The main research work is summarized as follows:(1)Studied the structural characteristics of commercial vehicle, and develop a26-DOFcommercial vehicle dynamics modelThis model was simplified from the vehicle with three-axis: steering shaft, drive shaft and traileraxle, and tires of the axis are simplified to single tire. The established model is amulti-degree-of-freedom nonlinear model, including tractor longitudinal, lateral, yaw and sprungmass roll and vertical freedom; semitrailer longitudinal, lateral, yaw and the roll of sprung mass, pitchand vertical, vertical of the non-sprung mass, pitch and roll of the non-sprung mass;6wheel rotation,aggregately26degrees of freedom model. The model was divided into driver model, steering model,engine and transmission model, brake model, wheel model, tire model, suspension model,non-sprung model, body dynamics model, the road model and aerodynamic model.(2)Electric performance testing of the core components of the system and the closed-loop control was designed and constructed. The electric control actuation system hardware-in-the-loop testbench was conducted, as well as the proportional relay valve performance testing and controlalgorithms.Both proportional relay valve and axle modulator contain proportional solenoid valve, whichhas a large hysteresis characteristics. In order to improve the accuracy of the control, a hysteresischaracteristic curve of the relay valve was gained through bench testing. A method was proposed,which used the feed-forward control combined with the PID control method to compensate controlerror caused by the hysteresis of the control valve. This achieves the aiming cylinder pressureimproves control accuracy greatly.In this dissertation, electric control actuation system hardware in the loop test bed was consistedof a wheel speed simulation system, measurement and control system, electrical control actuationsystem, the wheel cylinder pressure analog system and dynamics real-time interactive system.Proportional relay valve, axle modulator, trailer valve was tested to gain the control method. Aplatform was built with Matlab/xPC Target. This could be used as the basis of thehardware-in-the-loop experiment to verify the driver’s braking intention identification strategy, brakeassist strategy and brake force distribution strategy.(3)A braking intention identification strategy was proposed based on the displacement of thebrake pedal, the pedal displacement rate of change and the braking intensity of the three-dimensionalmodel. The relationship between the operation of the driver and his intention was recognized throughneural network method.Braking intention reorganization will enhance the safety of the vehicle and the responsecharacteristics of the braking system greatly. This dissertation used the structural characteristics of theelectric control actuation system, selected the appropriate parameters, and established the brake pedalfor the driver’s braking intention recognition model. A three-dimensional model, including pedaldisplacement, the pedal displacement rate of change and the brake intensity, was selected torecognize the braking intention.The relationship between the driver’s operation and the intention was obtained through severaldrivers’ test on the platform. The data were used to establish the neural network model. (4)A brake assist strategy was proposed.When the driver quickly steps the brake pedal, the brake system should identify whether thevehicle is in need of emergency braking, regardless pedal position, the brake assist function willrelease the maximum braking pressure to the brake chamber full braking. In this dissertation, theestablished brake intention identification three-dimensional model could identify the driver’s brakingintention. When the change rate of the brake pedal displacement is greater than the experimentallydetermined threshold, the system entered the emergency brake assist state, the strategy releasesmaximum value of the brake pedal. The dissertation collected several driver emergency brakingbehaviors using test bench, and the corresponding brake strength processed data were used todetermine emergency brake threshold.(5)A multi-axial vehicle brake force distribution strategy was established.Through the multi-axial vehicle braking mechanics analysis, the state of each axis has wasobtained. When the tractor front axle wheel brake locked, the front wheel would not generate thelateral force, and the train steering was uncontrolled, which would bring loss of steering capability. Inthe disturbance-free case, the vehicle could continue traveling straight until stopping; in lateral forceinterference case, the front wheel would skid, and off tracking occurred. Through the effect of inertialforce and the adjustment of the driver, the vehicle could resume operations under the steering controland straight track; if the semitrailer wheel locked before the tractor wheel, semi-trailer axle wouldalso slide. If the drift was not obvious and the swinging angle was not big, this instability could beovercome by relaxing brake pedal properly while slightly accelerated. The front and rear wheels ofthe car did not lock, and the ground could produce lateral forces; if the tractor rear axle locked first,the train would continue driving straight until stopping when there was no outside interference; whenthere was right lateral force interference (left lateral force interference can also be similar analyzed),the tractor had to rotate counter-clockwise around the saddle, the semi-trailer is a clockwise rotationaround the saddle Because Semitrailer was generally driven by rear axle, so neither to manipulate thesteering wheel or manipulating the accelerator pedal could made the vehicle resume drivingstraight, meanwhile semi-trailer could also exacerbate the folding of the vehicle due to the inertiapushing force generated by the brake on the tractor. Therefore, this braking condition was extremely dangerous condition. To sum up, it is stable and controllable when the front axle locked first, vehicles;it is uncontrollable when the semitrailer axle locked first; it is extremely dangerous when the rear axlelocked first. According to the results of the mechanical analysis, a braking force distributionalgorithm of multi-axial vehicle was developed. Currently, the automotive braking control systemmainly uses the control algorithm based on the threshold logic. This method is more complex, andcontrol process fluctuates heavily. To solve these problems, this dissertation used a method based onslip ratio control method. The PID control would control the slip ratio around the target slip ratio,which was determined through the related control algorithm. The control strategy was simulatedusing the26degrees of freedom vehicle dynamics model.In summary, this dissertation made innovative achievements as follows：(1)A26-DOF commercial vehicle dynamics model was developed after the analysis ofcommercial semitrailer’s structure. Electric performance testing of the core components of the systemand the closed-loop control was designed and constructed. The electric control actuation systemhardware-in-the-loop test bench was conducted, as well as the proportional relay valve performancetesting and control algorithms.(2)A three-dimensional model, including pedal displacement, the pedal displacement rate ofchange and the brake intensity, was selected to recognize the braking intention. The relationshipbetween the driver’s operation and the intention was obtained through several drivers’ test on theplatform. The data were used to establish the neural network model.(3)The force state of the tractor and the semi-trailer was analyzed, and the effect of the lockingsequence on the vehicle was discussed. This dissertation used a method based on slip ratio controlmethod. The PID control would control the slip ratio around the target slip ratio, which wasdetermined through the related control algorithm. The control strategy was simulated using the26degrees of freedom vehicle dynamics model.