Design Of Hardware-in-the-loop Simulation System For Driverless Vehicle

Intelligent decision-making system is the most important component in the driverless vehicle, and its level of decision-making will directly dominate the safety and reliability of driverless vehicle in running. In order to create an intelligent decision-making system owning reliable performance, fully testing to uncover its potential design flaws during the process of research is a very important way. Because of the limitation of field testing, hardware-in-the-loop (HIL) simulation technology will become an important means of analyzing, designing and validating intelligent decision-making system of driverless vehicle in light of its many advantages.At present, HIL simulation technology has been widely used in research and design activities of automotive electronic products, but few application cases in the validation of the level of driverless vehicle intelligent decision-making is published. Driverless vehicle HIL simulation is a novel application direction of simulation theory, which can be used to evaluate the level of driverless vehicle decision-making. Before applying the driverless vehicle HIL simulation system to experiments, a thoroughly research of its working principles, system structure and simulation credibility is an important means to ensure the high credibility of simulation system.First, the development history of automobile is reviewed and widely existing safety problems of automobile are analyzed, and then the concept of driverless vehicle is leaded. The development status of simulation theory and simulation technology in home and abroad are outlined, and the focus is application status of HIL technology in automotive engineering, especially in automotive electronics. The basic concepts of simulation credibility theory and the development of various specifications are introduced, and the research approach of the simulation system credibility is detailed; meanwhile, the application status of simulation technology is also presented.Next, from the perspective of technical requirements of the simulation system, the whole structure of the HIL simulation system is developed, and the working principles of the whole simulation system are detailed. In view of the test content of the intelligent decision-making level of driverless vehicle, the overall design frame of the virtual traffic environment and some typical test road structures are provided. In order to deal with the problem of acquiring information from virsual environment, the image processing algorithm to recover the3D information from2D images is established.Under the specific application requirements of HIL simulation, the solution to the inverse kinematics of parallel manipulator is worked out. The whole parallel manipulator is divided into three tetrahedrons, and then the Taylor series expansion and Newton-Raphson numerical algorithm are used to obtain the solution for the forward kinematics. Similarly, under the requirement aforementioned, based on the velocity, angular velocity, acceleration, and angular acceleration of the moving platform, the velocity, angular velocity, acceleration, and angular acceleration of every actuator are worked out, and then the dynamic equations of the system are established according to the Newton-Euler method. Based on the model control method, the PD controller is designed to control the motion track of the parallel manipulator. In light of the inconvenience of computing the solution of the forward kinematics in the general calibration algorithm, the calibration algorithm in this paper has been converted into computing the residual errors between the measured length and calculated length of the actuator under the condition of known pose of the moving platform, and the least squares iterative calculation method is used to calculate the kinematic parameters of parallel manipulator. The natural frequencies of the Stewart Platform is also given out.As an important component of the HIL simulation system, a vehicle system dynamics equations including the car body, tires, and suspensions are established. Considering to the fact that the driverless vehicle is usually tested in flat terrain, and the forward, lateral and yaw velocities are important and can sufficiently examine the behavior of a driverless vehicle, so the simplified vehicle dynamics equations are also built to simulate the dynamic behaviors.With credibility of driverless vehicle HIL simulation system as the object of the research, the development process of driverless vehicle HIL simulation system is built based on the past test methods. Two typical data preprocessing statistical methods are given out. In order to calculate the similarity between simulation data and experiment data, the concept of statistical similarity analysis is detailed, and similarity calculation method is highlighted. The hierarchical structure graph for driverless vehicle hardware-in-the-loop simulation system credibility analysis is established according to analytic hierarchy process (AHP). As an example of the present HIL simulation system, the computing methods of credibility of vehicle dynamic model, the control model of parallel manipulator and the whole simulation system are detailed.Generally speaking, this paper is organized on the specific application for the driverless vehicle HIL simulation system engineering. It is wished that the research works in this paper have some significances of reference for further research in this field; meanwhile, it is also wished that the research works in this paper can promote the application of the HIL simulation technology in the research and development of driverless vehicles.