Autonomous Robot Obstacle Avoidance And Path Planning Studies

Autonomous robot is an important branch of robotics. And the real-time obstacle avoidance and path planning is also one important research aspect of the autonomous robot. The main task of obstacle avoidance and path planning is figuring out a collision-free path from the known start position to known goal position in the robot environment at the least cost, and also requested the path smooth as much as possible to meet certain safety standards. In this dissertation, taking Grandar AS-R automobiles robot as research objects, and deeply research in autonomous robot's obstacle avoidance and path planning. The main contents and results are as follows:1. Detail described the traditional control methods of obstacle avoidance and path planning, and summarized their respective advantages and disadvantages. On this basis, the introduction of the obstacle avoidance and path planning based on ant colony algorithm.2. Detail described the biological model of ant colony algorithm, the algorithm features, algorithm theory and algorithm model. Then made a description of robot's obstacle avoidance and path planning environment, and elaborated the realization of algorithm steps.3. For the weight coefficientsαandβwere dynamically determined.αandβare constants in general ant colony algorithm,and remain unchanged in the algorithm implementation process. However, since the ants may be too much after a point in the obstacle avoidance and path planning, it is difficult to ensure that each of these points could be pheromones, which would bring about a serious problem, that is, if the prime point of access to information, the less, then the the results fall into a local solution more likely. Simulation results found that the constantsαandβto set to normal circumstances, can not find better solution. Therefore, in this design of ant colony algorithm,αandβwill change over time to adjust.4. Combined use of the heuristic information probability and transition probability from around the the current point, resulting from the beginning to the end point feasible way. By proportional selection can be used interchangeably three kinds of probabilities:based on the pheromone track transition probability;From the probability of heuristic information and considers the above-mentioned two kinds of probabilities arising from the transfer of probability.5. The ant colony algorithm and genetic algorithm were compared. By a large number of simulation experiments verify the ant colony algorithm for obstacle avoidance and path planning in the superiority and effectiveness.Finally, verified its correctness by experiments.