Optimization technology is a technology based on mathematics and applied to all kinds ofoptimization solution to engineering problems. It also plays an important role in fields of systemcontrol, vehicle Adaptive Routing, production adjusting, complex information system andelectronic commerce, and so on.Path planning of robot is a very typical Optimization problem which has many typicalcharacters such as complexity, restriction, non-linear, normative of modeling. Therefore, doingresearch to improve on an intelligent and general-purpose global optimization for large-scaleparallelization is a feasible researching direction.At present, there have been a lot of Optimization methods to solve the problem of pathplanning of robot, but most of these methods have some limitation. As a novel simulatedevolutionary algorithm, Ant Colony Optimization (ACO) has many merits as parallel essence,positive feedback and coordination, so it accords with the tendency that path planning algorithmevolves into intelligent and simulated. In my paper, I have done some research on the problem ofpath planning of robot based-on improved ACO. What I have done went into details aredescribed as follows:1. Establish an environment modeling to provide an abstract space of physical space, onwhich ACO can search for path.2. Put forward 3 improvements based on analysis of advantage and disadvantage:Firstly, improve the description method of barrier in grid model, and also take retracingstrategy to avoid un-mature convergence. Secondly, improve pheromone updated strategy, that tosay, join local pheromone updating with global pheromone updating to strengthen pheromoneintensity of iterative best ant. Lastly, bring path-crossing strategy of genetic algorithm into ACOto improve the converge speed of the new method.3. Use Markov to analysis the convergence of ACO.4. Use Matlab-tool for numerical value analysis-to do experiment, and the results of theexperiments show that the new method of this paper has the characteristics of simple, highefficiency and robust. |