Data-driven model-free slip control of anti-lock braking systems using reinforcement Q-learning

Independently developing anti-lock braking system(ABS) is the basis of mastering the active dynamic control technology of vehicle chassis.According to the experimental data and the logic threshold control method widely used in the industrial ABS,the impact on ABS control of every logic threshold is analyzed and the basic rules of calibrating the logic thresholds is established.By borrowing the idea from the self-tuning regulator in adaptive control,a logic threshold self-adjusting algorithm is presented.The algorithm identifies the optimal brake slip ratio of the road and calculates the reference speed on-line.The main logic thresholds are adjusted automatically according to the error between the wheel slip ratio and the optimal slip ratio in order to improve the control performance.The control parameters are calibrated and the control performance of the self-adjusting algorithm is verified by road tests under various conditions.The test results indicate that the presented logic threshold self-adjusting algorithm overcomes the influence of different factors,such as road adhesion coefficient,speed and load,on the ABS,thereby the optimal ABS control under every condition is realized.The developed ABS has good control effect and meets the development requirement.It provides a foundation for ABS industrialization.

Braking system plays a major safety feature in any vehicle; it is considered as one of the key points in safe motoring. The main objective of a good braking system is to provide safe stopping, in addition to the fact that it provides repeatable stopping. Anti-lock Braking System (ABS) is one of the recently developed safety feature used in a vehicle in order to prevent wheels from locking. This article proposes an intelligent adaptive fuzzy logic (FL) genetic algorithm (GA) controller for the Anti-Lock Braking System (ABS) to control the braking of a vehicle by controlling the slip ratio based on quarter car model. The main objective of the proposed controller is to prevent wheels from locking, which is a major safety issue in any braking system for different road conditions. The proposed controller consists of a Fuzzy Logic (FL) Genetic Algorithm (GA) based controller. The Genetic Algorithm is utilized in tuning the membership functions of the Fuzzy logic controller, which is developed to calculate the amount of reduction of the applied torque due to applying brake based on the error of the measured slip ratio signal compared with the desired one. Then a Genetic Algorithm (GA) is utilized in tuning the Fuzzy Logic parameters, which are the center of the Gaussian membership function, and the spread (б). The simulation results of the proposed fuzzy logic genetic algorithm controller are compared with both the results of the Fuzzy logic controller and the PD controller developed in previous work of the authors. The simulation results present many advantages of using Fuzzy Logic Genetic Algorithm based controller over both the Fuzzy Logic and the PD controllers, and the simulation is carried out at different road conditions, namely dry, wet and icy road surface.

Hierarchical [formula omitted] fuzzy-neural control of anti-lock braking system and active suspension in a vehicle

This paper investigates the use of slip control on heavy vehicles to reduce stopping distances and compressed air usage in an emergency stop. A hardware-in-the-loop (HiL) rig has been developed to investigate slip control. The hardware includes an air receiver, anti-lock braking system (ABS) or slip-control modulator, brake actuator and brake calliper. The braked wheel and vehicle dynamics are simulated on a micro-processor. A gain-scheduled slip controller using the same pneumatic actuation as used by conventional ABS is developed. The slip control strategy increases the mean fully developed deceleration by 6% and halves the air usage compared to the ABS strategy on the HiL rig. The transient performance of the slip controller is poor due to the sluggish response of the ABS valves, causing a 10% increase in stopping distance. A novel fast-acting brake actuator design is described. Simulations of the new actuator implementing slip control show an improvement of 10% in stopping distance performance and 30% in air usage.

Electric vehicles (EVs) have the advantage of using both regenerative braking and mechanical braking to produce the necessary braking torque. Currently, vehicles use Anti-lock Braking Systems (ABS) which provide intermittent braking torque to the vehicle and thereby prevents skidding or slipping of the vehicle by ensuring pure rolling of the wheels. Regenerative braking on the other hand generates the necessary braking torque by converting the kinetic energy of the vehicle to electrical energy via the motor(s). To coordinate both braking systems, the conventional method involves the use of crisp logic controllers that depend on lookup tables containing the data about various aspects of the vehicle such as battery state of charge (SoC), brake pedal input etc., to enable co-operative control of both systems. This paper presents a combined implementation of both braking systems in electric vehicles using separate fuzzy controllers for both ABS and regenerative braking. The control scheme is implemented such that when regenerative braking is not capable of providing enough braking torque, braking will be one by ABS. Two fuzzy logic controllers have been used, one to implement regenerative braking and the other to implement mechanical braking via ABS. If such a control algorithm is used, the driver will not have to manually adjust the level of regenerative braking in the vehicle and such an approach ensures that the braking process of an EV is smooth, safe and efficient. It also reduces the wear and tear of the brake pads thereby extending their life and reducing the cost of maintenance and by extension the total cost of ownership of the vehicle. The results demonstrate that the controllers for both ABS and regenerative braking can coordinate efficiently and reliably to produce proper and safe braking of the vehicle.

<div>As the basic function of the active safety configuration of a vehicle, the anti-lock braking system will compromise the driving safety if it fails. Based on the self-designed electro-hydraulic braking system, this article proposes an anti-lock brake redundant control architecture. The electro-hydraulic braking system is mainly composed of four parts: a brake pedal unit, a hydraulic drive unit, a brake execution unit, and a control unit. The mechanical structure is compact and exquisite, and the system has the function of precise and adjustable hydraulic pressure. The control architecture adopts a hierarchical control design, which is mainly composed of an upper wheel slip rate controller and a lower hydraulic pressure controller. Both the upper and lower controllers use a sliding mode variable structure control to improve the robustness and accuracy of the control. The upper slip rate controller outputs the desired master cylinder hydraulic pressure with the optimum slip rate of the rear wheels of the vehicle as the control target. The lower hydraulic pressure controller outputs the desired torque of the motor with the desired master cylinder hydraulic pressure as the target and achieves hydraulic pressure regulation of the master cylinder by controlling the motor motion. To verify the effectiveness of the algorithm, co-simulation and a hardware-in-the-loop test platform are built. Anti-lock braking tests are carried out under different typical working conditions: low-adhesion road, high-adhesion road, butt road, and split road. The results show that when the anti-lock braking system fails, the redundant control algorithm can achieve effective slip rate control under different driving road conditions and meets the anti-lock brake redundant control requirements. This study provides a reference for the development of low-cost anti-lock braking redundancy functions for vehicles equipped with electro-hydraulic braking systems.</div>

This paper presents the development of a PID controller for an Antilock Braking System (ABS) using vehicle longitudinal model. A Five Degree of Freedom (5-DOF) vehicle longitudinal dynamic model was derived and integrated with an analytical tyre dynamics, the Magic Tyre model. Several transient handling tests are performed such as sudden acceleration and sudden braking test to validate the vehicle model. The model is used as a plant to develop an antilock braking system to control longitudinal slip and reduce the stopping distance. A hydraulic brake model was developed as the brake actuator to produce brake torque. A conventional PID controller has been implemented to deal with the strong nonlinearity in the design of ABS controller. The proposed ABS control structure is shown able to significantly reduce stopping distance and control the longitudinal slip during heavy braking.

To research the influence of the proportional relay valve on the economy and safety of the electric bus through the braking recovery system, a proportional relay valve model was proposed. Based on the analysis of the structure and working principle of the solenoid on/off valve, a dynamic model of “electro-magnetic-machine-air” coupling was simulated and analyzed. Then, based on the solenoid on/off valve and proportional relay valve, two pressure regulator modules were designed, the braking energy recovery system schemes were designed, respectively. Fuel economy tests were carried out under China city bus typical driving cycle based on two kinds of air pressure regulating module to research the influence of the proportional relay valve on the economy of the electric bus. An integrated control system of braking energy recovery and anti-lock braking system based on the proportional relay valve was proposed, hardware-in-the-loop braking energy recovery system contrast tests of the anti-lock braking system were also carried out based on two kinds of air pressure regulating module. The test results indicated that the proposed braking energy recovery system schemes were effective. The test results also demonstrated the economic contribution rate of the braking energy recovery of the proportional relay valve pneumatic control module was slightly higher than that of the proportional relay valve. Meanwhile, the proportion relay valve pressure regulating module achieved a good target speed follow and the accurate adjustment of air pressure in anti-lock braking system control.

<div class="section abstract"><div class="htmlview paragraph">This paper presents a new method for finding slip control law of Anti-lock Braking System (ABS), based on sliding mode control method . A four wheel car model with seven degrees of freedom is considered. Slip of each wheel is controlled separately so that it remains in the desired range for every kind of road condition, and by tuning desired slip undesired yaw on miu-split surfaces is prevented.</div></div>

With the development of the global automotive industry, environmental pollution and driving safety have been major problems. This paper focuses on a novel integrated cooperative antilock braking system (ABS) controller, which is used for antilock regenerative braking systems (ARBS) and antilock mechanical braking systems (AMBS) for hybrid electric vehicles (HEV) driving on different road surfaces. An intelligent tire system is utilized to detect varying road surfaces to obtain friction information and optimal operation slip ratio. In addition, the HEV eight‐degree‐of‐freedom dynamics model is developed for ABS control, which includes the LuGre tire model. Adaptive backstepping and finite state machine controllers are explored to maintain optimal wheel slip and regenerate energy, based on wheel slip, battery state of charge (SOC), motor capability torque, and vehicle velocity, and to develop the switching rules of regenerative and mechanical braking systems. In order to evaluate the performance of regenerative and mechanical braking systems on harvesting energy and driving safety, the HEV and LuGre dynamic tire models are developed in MATLAB SIMULINK /SimDriveline. The estimation results of different road surfaces and slip controller effects are tested via the experiments and simulations, finding good braking performances and high regenerative efficiency in different maneuvers.

A laboratory Anti-Lock Brake System (ABS) is examined. The architecture of the ABS system is shown. The real-time experiments related to a control action to stabilize the slip at a certain level are taken into consideration. A new slip control algorithm differs from the previously used approaches. It is based on the measured current of the braking DC motor. The experimental results collected in the real-time for an old (relay) and new (current based) slip control are compared.

This paper focuses on the design of a fuzzy fractional Proportional-Integral-Derivative (PID) controller. The controller which is based on the conventional PID controller is modified to be P1–α I1–β D1–γ with the objective to make the controller more flexible and robust to the changes in input and the parameters. In this control structure, the fuzzy logic control acts as the tuner for the parameters α, β and γ, so that the overall controller parameter can be varied and converge better to the changes in a system. The designed fuzzy fractional PID controller is applied to a validated antilock braking system model using an electronic wedge brake system as the actuator. Numerous numerical simulations and comparisons with other fractional PID/conventional PID controllers show that the fuzzy fractional PID controller can not only ensure good control performance with respect to reference input but also improve the system robustness with respect to model uncertainties.

"The anti-lock braking system is a safety equipment that prevents the vehicle wheels from locking up under emergency, panic, or harsh braking conditions, ensuring steering control during braking and reducing braking distance. This system is studied by automotive engineers for performance comparisons and for determining optimal control methods. For this reason, this paper presents a method to mathematically model the anti-lock braking system. The obtained model is afterwards used inside a Simulink simulation to investigate the results with and without the anti-lock braking system and to propose a three-position regulator instead of the generally used bang-bang controller. The simulation results show the improvement of the vehicle braking time and braking distance."

The aim of this research is to apply vehicle dynamics control systems to mitigate the vehicle collision and study the effects of these systems on the kinematic behaviour of the vehicle’s occupant. An unique three-degree-of-freedom vehicle dynamics/crash mathematical model and a simplified lumped mass occupant model are developed. The first model is used to define the vehicle body crash parameters and it integrates a vehicle dynamics model with a vehicle front-end structure model. In this model, the anti-lock braking system and active suspension control system are co-simulated, and its associated equations of motion are developed. The second model aims to predict the effect of vehicle dynamics control systems on the kinematics of the occupant. The Lagrange’s equations are used to solve that model due to the complexity of the obtained equations of motion. It is shown from the numerical simulations that the vehicle dynamics/crash response and occupant behaviour can be captured and analysed quickly and accurately. Furthermore, it is shown that the vehicle dynamics control systems can affect the crash characteristics positively and the occupant behaviour is improved.

This study employs a dual deep neural network (D-DNN) to accurately estimate the absolute longitudinal speed of a vehicle. Accuracy in speed estimation is crucial for vehicle safety, because longitudinal speed is a common parameter employed as a state variable in active safety systems such as anti-lock braking system and traction control system. In this study, DNNs are applied to determine the gain of an adaptive filter to estimate vehicle speed. The used data consists of longitudinal acceleration, wheel speed, filter gain, and estimated vehicle speed. The data generated from Carsim software are collected and preprocessed using a Simulink model. To acquire data with numerous wheel slip patterns, various acceleration and deceleration conditions are applied to four different road conditions. Though, it is challenging to achieve a single DNN model that is optimally cope with the various driving situations. Thus, we adopt two DNN models that were individually trained in low and high acceleration regions. The dual DNN model results in error reductions of 74% and 65%, compared with a single DNN and classical adaptive Kalman filter approaches, respectively.