Research On The Adaptive Cruise Control Tracking System Applied For Motor Vehicles

Adaptive Cruise Control (ACC), as an essential part of the Intelligent TransportationSystem (ITS), provides a feasible approach for addressing the public concerns for vehiclesafety, driving comfort and traffic congestion which are incited by the surge of on roadvehicles. Aiming to broaden the traditional ACC’s functionality and usability, thisdissertation conducts research on some key technologies of ACC system through theoreticalanalysis, actuators in the loop simulations and road test experiments.Consisting of simulation model parts and hardware actuator parts, the Actuators in theLoop Simulation (ALS) platform is designed. Furthermore, additional sensors and actuatorsare equipped in the demo vehicle so as to facilitate it with the capability of longitudinalautomated control. Actuators control algorithms for throttle openness and braking pressuretracking control are developed and the satisfying results could be obtained in ALS and invehicle experiments for actuators’ control.Multi-lane objects identification method is devised for on car radar. With thetransformed radar coordinates, radar raw data is sifted with respect to the valid processingregion and object lane information. For best identification performance, threshold, slidingsample window and nearest object selection are used for object tracking in different phaseof calculation. On the curve road, norm central angle method is employed for objectsposition compensation. Road experiments indicate both the identification of multi-laneobjects and curve road compensation could be realized in a satisfied way.For control simplicity, sub functions of ACC system are defined and their controlpriorities are allocated. Using the inquiry map, cruise control of the demo vehicle isrealized. For better control robustness, preceding vehicle following control for the hostvehicle is based on the motion planning and the acceleration regulation. Accelerationregulation is achieved by two inquiry maps calculating throttle openness and brakingpressure from the desired acceleration. Meanwhile, the disturbance of environmentalchanges is offset by the adaptive control method. The tracking error model and accumulatedtracking error model of the host and preceding vehicle are established based on the motionanalysis of the two vehicles, thus enabling the Linear Quadratic Regulator (LQR) foroptimal desired acceleration calculation for the host vehicle.The cut-in maneuver identifier, which is trained by the Support Vector Machine (SVM)utilizing on road side lane vehicle cut-in data, is designed for side lane vehicle cut-in control of the host vehicle. For SVM training, each attribute of the training sample isnormalized and a fuzzy membership coefficient is applied in order to improve the trainingaccuracy of the identifier. Meanwhile, other coefficients’ optimal values for SVM trainingare determined by grid search based on cross validation. Using longitudinal distance andtime to collision of the host and side lane vehicle as inputs, fuzzy controller outputs thedesired acceleration for the host vehicle’s control. Simulation comparison and on roadexperiments confirm that, with improving safety potentials for the ACC vehicles, both thepreceding vehicle following control and the side lane vehicle cut-in control could achieveeffective speed and distance control of the host vehicle.For longitudinal and lateral motion stability analysis of the ACC vehicle, a dedicatedvehicle model is devised. The transformation process of the phase plot with regard to thetangent value of vehicle side slip angle and vehicle yaw rate, and the presence of thesteering equilibrium point are analyzed under different maneuvers. The front and rearwheels’ slip ratio behavior under different wheel torque are discussed, and it is concludedthat, under a fixed wheel torque, front and rear wheels could achieve specific slip ratio orlock up. Furthermore, the presence of steering equilibrium point regarding to different slipratio of the front and rear wheels is discussed. As a result, the braking strategy for all wheelbraking or front wheel braking only in collision avoidance control is obtained, and itsnecessity is confirmed by vehicle simulation.