Environmental Modeling And Path Planning Of Mobile Robot

Mobile robot is currently one kind of high intellectualized machine system. It hasmany functions, such as environment sensation and forecast, behavior control andexecution, dynamic decision-making and planning, etc. However, environment modelingand path planning of the mobile robot are the most basic and important branch of themobile robot. So in this dissertation, the main assignment is environment modeling andpath planning of the mobile robot. And this paper studies on improved algorithm relatedon the path planning. The main work and innovation points include:Firstly, this paper builds the grid method for the environment modeling. It isadvantageous for the interpretation of the environment, and good to express and coding;According to inspire function, this paper use the algorithm ofA *to search path, to savethe search time of the follow-up of genetic operator; by marking the feasibility of gridpoints, to avoid program into a "dead end" repeat. This paper uses the Maklink chartmethod to build the environment model. It is not restricted by shape or size restrictions,and more tally with the actual situation; and then to improve the Maklink chart, to makethe application area more widely; and then in the Maklink chart this paper uses theadvantages of Dijkstra algorithm to plan out the initial path.Secondly, in path planning of improved genetic algorithm, through the improvementof fitness function, this paper avoids the weighted summation from instability problems;this paper introduces the improved operation of path selection, to solve the problem of theinviable path, when the initial population genetic operation and in the process of genetic;In order to prevent the traditional genetic algorithm premature convergence, this paperuses the pure population evolution method, and uses many times the simulation results toverify the feasibility and effectiveness of the algorithm.Finally, in the path planning of the improved ant colony algorithm, this paper uses theimproved path encoding method, to reduce the computation; when select the path node,this paper introduces the conception of angle, when choosing the next node, priorityselection and target point angle smaller point, to avoid the ants from detours; this paper uses the improved pheromone update strategy, join the adapt to adjust the parameter in theexisting elite ants, effectively improve the speed of convergence.