ABS Road Surface Fuzzy Discrimination Based on Normal and Exponential Distribution
Tie Wang, Jing Yuan Wu, Xin Zhang, Jing Shang · Advanced materials research · 2012
The methods of work this identification system is that distill wheel speed characteristics to standard signal then stored in the system. When the first anti-lock loop begin to input the initial velocity to identification model then extract theoretical value of the system and input the actual speed then calculate the similar with theoretical speed. The maximum similarity value is the identification when the pavement is true. After that every cycle identifies monitoring (again identify based on new data) that is establishing state observer. Though it is inaccurate due to various interference identifications or when the adhesion coefficient of pavement has changed ABS will determine quickly make the work more reasonable. Choose normal distribution and exponential distribution of statistical principles to identify the pavement and modify the identification model with gamma membership function. Tire model choose H.B.Pacejka model suitable for programming. It will use Gauss numerical integration and rational interpolation to increase the timeliness of identification. The method this paper described is suitable for the identification of low road.