Based On Data Mining TCS Fuzzy Control For Electric Vehicle

Traction control system (TCS) is a kind of active safety control system. It is an extension of the antilock brake systems (ABS). Through controlling automotive drives and roller-blading turning rate, TCS can prevent the driving wheels of cars from excessive spin at the beginning, accelerating and hill-climbing maneuvers. TCS can make sure the cars get the maximum traction and the best driving stability through using full of vertical adhesion and lateral adhesion from the ground as to inhance stability of the vehicle. Therefore, further study on traction control system car (TCS) to solve the stability and safety is of great significance.This study is designed to get the knowledge of the traction control system for electric vehicles. We choose the hub motor as the leading program. We can adjust the voltage of the motor to control the output torquethe which can keep the slip ratio of the wheels at16%. If the slip ratio of the wheels up to16%, the ground can supply the maximum longitudinal adhesion coefficient and the perfect longitudinal adhesion coefficient. The main research contains the following several parts:(1) Selection and modelingFirst, we researched the structure of the electric vehicles in the program. Then we establish the dynamic model、motor model、wheel model and the slip ratio calculate medel.(2) The research of fuzzy controlFirst, we design a PID controller. It can simulate the electric cars and analyze the ability of self-control and robustness of cars on three kinds of different grounds. We have found that PID control have some features such as smooth trasition, none of steady-state error and adaptive capacity. The most important feature for electric cars is that it is very comfortable for the cars to work on the road with low friction coefficient. At the same tiome, fuzzy control has the features that shorter transition time, and this control is more comfortable of working on the ground with high friction coefficient.(3)The research of PID and fuzzy compensationAs the PID control and fuzzy control has its each advantage, we want to imagine making the two parts together like the style of compensation controlt. Through simulation we can find that:compensation control perform well on the transition time, steady-state error and adaptive capacity. These can express that compensation is possible.(4)The detail of fuzzy rulesFuzzy rule acquisition is the key to the successful fuzzy modeling. Usually the fuzzy rules can be created in three ways. The first is from the experience of the experts and the rules of knowledge generation. The second is from the rules of sample data. The last one is from the mixture of sample data of the knowledge of experts. The simple system can be ruled based on the knowledge of experts, but it is difficult of getting rules just depending on experts. Compared with the fuzzy rules from the experience,data mining technology can automatically extract fuzzy rules from the data.It’s more objective and accurate. We proposed a algorithm with better robustness base on WM and iWM algorithm,and realized it in computer with matlab language. The simulation results demonstrate that the proposed algorithm can effectively extract rules from the sample data.