Study On Intelligent Control Strategy And Simulation Of Vehicle Based On Path Recognition

The research on intelligent vehicle involves many fields, such as computer measurement and control, computer vision, sensor data fusion and vehicle engineering, etc. It mainly includes safety monitoring, intelligent anticollision, aided driving, auto driving, behavior planning decision-making, system structure, and synthetical integration, etc. The sensor and control algorithm are major factors influencing the development of intelligent vehicle. This paper chooses a motor model for game as hardware platform, and studies the intelligent vehicle control strategy and algorithm based on path identification. The driving simulation based on path identification is performed on the computer. At last, the motor model can identify path by itself and drive on the appointed path.Firstly, this paper describes the research directions and involved fields. Then, three control algorithms realizing intelligent control are presented:①intelligent control based on fuzzy control theory;②intelligent control based on artificial neural network;③intelligent control based on fuzzy neural network. The works in this paper are as follows:(1) Study the technologies relevant to control strategy. The precise identification based on infrared sensor linear path, such as judgement and return control on divorce path, valid identification on cross path, and typical path control, is solved. The algorithm designing continuous path offset, rudder servo rotation angle and traffic speed controller are studied. The relations between parameters of speed PID controller and speed error, objective speed and path offset, are established on the basis of experiments and experience.(2) Perform the practice, calibration and matching of software and hardware in control system. Combining simulation with bench test and road test, the parameters of simulation object are calibrated and the parameters of PID controller are set on the basis of BP neural network. Then, the whole system is tested, and the intelligent vehicle model navigation system is matched. (3) Off-line simulation platform based on PLASTID simulation software, experimental platform based on Real Time Debugger and vehicle load development platform are established to ensure the project to be performed successfully.After many experiments, the following aspects have been achieved in this paper:(1) The project to develop intelligent vehicle, power-supply module, sensor module, rudder servo, DC motor, other parts and vehicle model platform were designed. To verify the design validity of the intelligent vehicle model, a series of path was implemented, on which the system test was performed.(2) According to the flow chart of control system designed, the control algorithm was programmed with C compiler of METROWERKS CODEWARRIOR. By BP neural network, the parameters of PID control algorithm were set and the rotation angle PID controller and speed PID controller of the vehicle were designed. Also, the synthetical harmonization control system of rotation angle control and speed control was designed. After the compute module of path offset and path identification module were designed, the intelligent vehicle can identify the cross path and the situation of divorcing from the path. The feasibility of the algorithm was validated by real road test. Furthermore, the distribution of"―"type was made to identify the path taking into account the influence of infrared sensor distribution on path identification.(3) The paper describes the simulation software and control algorithm programme and gives the simulation results, then shows the experimental results of the intelligent vehicle on real road. The following situations were taken into consideration:The vehicle runs out of the road and return. In this case, the highest speed is 1.3 m/s.The vehicle runs according to the assigned path when it passes the cross road. In this case, the highest speed is 1.4 m/s.The vehicle runs on the"snake"type road according to the assigned path. In this case, the highest speed is 1.2 m/s.The vehicle makes a turn according to the assigned path when it runs on the straight road. In this case, the highest speed is 1.4 m/s. The simulation and experimental results show that the intelligent vehicle model designed in this paper can run independently according to the assigned path and the control algorithm is stable and feasible.