Path Planning Research For Mobile Robot Based On Improved Ant Colony Algorithm

The problem on path planning for Mobile robot is an important branch in robotics research area, has always been concerned about the problem by many scholars at home and abroad, So far, there have been significant breakthroughs in the field and made a series of achievements, there are many optimization algorithm to be applied on how to solve the path planning problem now, but there is a big flaw in most of the algorithms themselves. It is difficult to work out an effective path.In the environment where barrier is very complex, because a mobile robot only have a few priori knowledge.in an unknown environment. Demands are often able to inspire a spark of discovery and creataion. Since the bionics is created the mid-20th century, people are inspired from the mechanism of biological evolution, and put forward many new ways to solve complex optimization problems, ant colony algorithm is one of them. This method caused a great sensation in the academic since its inception, and has been successfully applied to the traveling salesman problem, production scheduling problems and other fields. In the trend of biomimetic, intelligent of robot path planning algorithm, path planning algorithm for mobile robot based on ant colony algorithm is proposed in this paper, and the content in this paper includes the following sections:1. First,mobile robot path planning research situation between At home and abroad and the advantages and disadvantages of a variety of path planning methods were compared, and a brief overview is made on the progress of the ant colony algorithm.2. Environmental model established by using the grid method provides a physical environment for path search of the ant colony algorithm. Robot's working environment is assumed to be the same size of the squares (grid), the grid size is Accordance with the guidelines in which the robot can. move freely。Suppose the robot working environment consists of black and white two kinds of grid. White means that the robot can move freely in this grid, and the black grid means barriers which robots can not cross. This article aims to get around a valid path to makes the robot avoid obstacles according to sensor information and effective control of the controller, when the robot encounter barriers grid。An to obtain effective path relys on repeatedly calculation and comparison under the control cycle times which is preseted in the program.3. This article intends to use ant colony algorithm, but the traditional ant colony algorithm is easy to deadlock, stagnation and other defects. This paper try crossover and mutation strategies is introduced In the ant colony algorithm to improve the traditional ant colony algorithm. comparison,analysis, observation is done between Improved algorithm and traditional algorithms by Matlab. Results of a large number of simulation experiments show that the consequent of the algorithm is obvious in robot path planning, especially in the shortest path planning. The simulation results and experimental results Have a high degree of moderate with the actual results anticipated.4. Analysis algorithm convergence using Markov process. Markov process is an important class of random process, the process has a number of important properties, such as ergodicity of the Markov process, monotonicity of Markov chain and so on. Ant activities is seen as just a natural phenomenon initially, is a complex random stochastic process, but as the scientist's observation and research, it is found that the activities of ant colony is not a random stochastic process ts activities, rules has many similarities with Markov process. After careful scrutiny, ant colony algorithm can be approximated as a Markov process to deal with when it does not affect the actual situation. iterative process of the ant colony algorithm the (Markov process) has benn done a more detail proof In some cases assumptions.5. Completed control process design based on path planning of the ant colony algorithm. In this paper, MATLAB7.1 is designed as a Programming Platform Complete programming of the shortest path planning of robot based on the ant colony algorithm. By MATLAB7.1 software emulation,we are able to see the shortest path to be planned out in the case of combinations of different parameters, but also we are able to read out the distance of the shortest path for each simulation By analyzing the parameters to determine the best combination to achieve better planning of the path.6. Experimental design. Due to limited conditions, subjects used in this experiment is based on the original wheeled mobile robot and do some partial structures Environment for experiments is the laboratory ground.。Experimental robot is wheeled robots, the rear is the wheel, wheel Track left is 140mm, Front and rear Wheelbase is 180mm and the rear height is about 80mm. The system control center use Hitachi's 16-bit SCM of H8/3048F series. Experiment measure mobile robot displacement and determine the location of obstructions by infrared ranging sensors In this study, using GP2D12 sensors produced by Sharp, the range of sensor measurement is 10cm-80cm, the maximum allowable angle> 40°, the voltage is 5V or so.