Crawler mechanisms have the advantage of high-speeding and stable locomotion on uneven terrain. Therefore, such vehicles have been applied in many areas, including those used for search and rescue. Tracked mechanism is a high-tech product. One of the difficulties in the vehicle researching is the issue of path tracking. Numerous of control methodologies have been proposed before, yet limits to some simple paths, such as circle or straight line. In this paper, a control scheme involving complicated curve following is proposed to adapt vehicle to more intricate environment. Within the control structure, a novel method named curve fitting method (CFM) is issued to match the targeted route. Then, a multi-parameter adaptive control algorithm (MPACA) is further constructed based on conventional genetic algorithm (GA). This intends to automatically adjust the vehicle's speed parameters according to the changing route. However, some surveys show that GA is hard to be used to follow the transition area (TA) perfectly and this situation becomes more serious if many transition areas (TA) exist in the curve. Due to this phenomenon, multiple-populations genetic algorithm (MPGA) is applied to realize MPACA instead. Two simulations tested on typical routes which contain TA demonstrate the performance of CFM, MPACA and MPGA in the adaptive tracking sphere and the simulation tests are displayed in Section 4.1. At the end of this paper, a practical validation has been done in an open space of Shanghai University. This will be detailed in part 4-2.
Zhu, Hualin; 罗均; 谢少荣; 李恒宇.New improvement on adaptive path following control for multiple-populations genetic algorithm adopted in tracked vehicle,International Journal of Innovative Computing, Information and Control,2014,10(2):783-795