Mobile Robot’s Path Planning Based On Immune Genetic Algorithm

Since mobile robot has been playing an important role in many areas, it is natural to say thatits related technology research has been paid attention by people. As one of the key issues inrobotics research, path planning problem is always an important topic for the researchers whocome from all over the world. Especially when the precision of robot hardware can not beexceeded in a short time, the research on path planning algorithm is very important. Based on these,the problem of Immune Genetic Algorithm and its application on robot path planning problem isstudied in this thesis.Immune Genetic Algorithm can be seen as an improved genetic algorithm which has immunefunction. It has been widely used in mechanical design optimization, control optimization, channeloptimization, and so on. An improved immune genetic algorithm-adaptive immune geneticalgorithm (AIGA) which is based on the actual mobile robot path planning problem is proposed inthis thesis. In the improved algorithm, a new definition of the antibody concentration, strategy ofadaptive crossover probability and strategy of adaptive mutation probability are used.The mobile robot is assumed to work in2D environment and grid method is utilized to makeenvironment modeling in this thesis. To save memory and make operations of the improvedimmune genetic operator easy, both coordinate method and serial method are used for coding.To make the simulation more realistic, a new fitness function which take path length and collisiondepth into account is used. According to the actual situation, the MMB (Min-Max Box) strategy isproposed to speed up the computing speed and make collision detection more effective.Appropriate crossover and mutation operators, insert operator, delete operator and improveoperator is also used for the special path planning problem.Finally, the mobile robot path planning problem is simulated for static and dynamicenvironment on VC++6.0platform. In static complex environment, we find that the more complexenvironment, the more significant advantages of AIGA while comparing AIGA with IGA andEGA; Three cases which include increasing the number of obstacles, moving obstacles andmoving target is designed for mobile path planning in dynamic environment. The simulationresults show that AIGA is capable of performing robot path planning in dynamic environment andhave strong real time Capability.