Path Planning And Location For Mobile Robot

Mobile robot technology involves multiple research areas and represents the frontier of high technology. Mobile robot technology has a wide range of applications in different walks of life. Mobile robot is capable of acquiring the information of environment and its own state to achieve the scheduled mission in the environment with obstacles. Navigation of mobile robot is the process of moving toward the target autonomously in the environment with obstacles. In the process of navigation, mobile robot has to carry out the accurate modeling of the environment, achieving the location of the pose and planning the optimal path from the starting point to the target point. Therefore, the research of environment modeling method, path planning algorithm and locationing method for mobile robot has theoretical and realistic significance.In this dissertation, the path planning approach and location technology for mobile robot are focused on. The main contents of this dissertation are summarized as follows:For global path planning of mobile robot, simplified visibility graph suitable for path planning algorithm of mobile robot is proposed to solve the problem of environment modeling. Considering the position of the obstacles in the environment and the relationship between starting point and end point of the mobile robot, the redundant obstacles which does not affect the result of path planning are removed. The representation of environment model is simplified. The purpose of reducing the number of alternative paths in the process of path planning is achieved, which improves the efficiency of the follow-up path planning algorithm.To solve the contradictory between the convergence speed and the local optimum in ant colony algorithm, an improved ant colony optimization algorithm is proposed for path planning. The local path information is integrated with the initialization of pheromone and the selected probabilities of the paths, resulting in improving the convergence speed. For overcoming the stagnation phenomenon, crossover operation is drawn into the proposed algorithm, which enhances the capability of escaping stagnation phenomenon.The proposed algorithm improves the search efficiency of optimum path for mobile robot.To solve the local minimum problem in the complex environment of local path planning for mobile robot, a multi-behaviors coordination approach is proposed. The proposed approach defines three kinds of basic behaviors. The mission of path planning is completed by switching amoung three kinds of basic behaviors. The trial-angle compensation method in escaping from the local minimum behavior is designed to solve the local minimum problem existing in the environment with the U-shaped obstacles. The proposed approach improves the reliability of results of local path planning for mobile robot in the environment with the U-shaped obstacle.To solve the “particles degeneracy” phenomenon of the Rao-Blackwellized particle filter, a new approach based on Particle Swarm Optimization is presented to solve SLAM problem for mobile robot. During the particle re-sampling process, the proposal distribution of mobile robot’s pose is acquired by Particle Swarm Optimization. The attraction of energy efficiency is applied to optimize and adjust the obtained particle sets, which the diversity of the particles is enrished. The proposed algorithm eases the “particles degeneracy” problem and ensures the accuracy of SLAM results.Due to the drawback of FastSLAM2.0about the noise assumption being limited by the statistical characteristics, an improved FastSLAM2.0algorithm is proposed. H∞filter is used in the improved FastSLAM2.0algorithm instead of EKF, which reduces the influnce of error of robot pose estimation. For the “particles degeneracy” problem, the particle re-sampling strategy based on genetic algorithm and particle swarm optimization is proposed in improved FastSLAM2.0algorithm. The consistency of mobile robot pose estimation is improved effectively. The proposed approach overcomes the drawbacks of standard FastSLAM2.0algorithm which the inaccurate pose estimation of mobile robot is caused by map estimation error accumulation.