Font Size: a A A

Reseach On Intelligent Algorithm Based Mobile Robot Path Planning

Posted on:2015-03-12Degree:MasterType:Thesis
Country:ChinaCandidate:N LiuFull Text:PDF
GTID:2348330518472137Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
With the proposed bionic intelligent optimization algorithms and the extensive applications, these new optimization algorithms have become much attention from researchers and an important research direction optimization algorithm.Mobile robot path planning technology is an important research field of robotics. In this paper, the global path planning based on environmental model is researched, which is based on known environmental information and use a variety of algorithms to plan an optimal path;then according to this path, robots travel along this path as possible and avoid the unknown obstacles at the same time. If the deviation from the original path reaches a threshold value, a new request may be re- routing.The main contents are as follows:1. This article focuses on this paper path planning based on environmental information,to amend two aspects of the environment on the basis of the existing grid. Firstly, robot is no longer seen as a particle, which considers the size of the robot and safe distance when the grid environment generates obstacle grids; Secondly, a grid environment model based on effective vertices is used, and the detection algorithm process is given in detail. The introduction of this method makes the path planning avoid the U-groove type problems from the model and reduce the probability of the local convergence of the algorithm.2. In the part of the path planning algorithm, the effects of these parameters of ant colony algorithm (ACO) algorithm (BBO) are studied and their application methods in the optimal path planning are researched. In this paper, the basic algorithm is improved to make its stability, error rates and timeliness on all outstanding in path planning. Improvements on the ACO include: pheromones restriction policies, elite ant strategy, dynamic parameter adjustment,local path detection optimization and bi-directional searching mechanism;Improvements on the BBO include in addition to the local path detection optimization and bidirectional search mechanism used in ACO: elite island strategy and dimensionality reduction mechanism.3. BBO algorithm, PSO algorithm, AFSO algorithm and the ABC algorithm is largely similar. The improved methods and ideas for BBO algorithm can be easily used on the PSO,AFSO and ABC. Via comparing horizontal the five improved algorithm, the stability, solving ability (error rate), timeliness and robustness of the five improved algorithms in different environment models are analyzed by using statistical methods.4. An evaluation criteria matrix based on the needs of User-oriented perspective is proposed for user’s choosing a correct intelligent algorithm to path planning and is used to a real environment to test its performance.The innovations in this paper are as follows:1. A new search program is proposed in bi-directional searching mechanism which is used to five alforithms, and accelerate the searching speed of path planning.2. Dimensionality reduction mechanism is proposed in BBO, AFSA, ABC and PSO algorithm, and effectively reduce the time-consuming.3. A selecting algorithm method for users is proposed. The robot only need to provide requirements of planning algorithm performance and then the method can select the most suitable algorithm for the robot.
Keywords/Search Tags:ant colony algorithm, biogeography optimization algorithm, path planning, grid method
PDF Full Text Request
Related items