Research On Intelligent Algorithms Based On Mobile Robot Path Planning

With the rapid development of computer technology and robotics,society is becoming more and more intelligent.Autonomous mobile robots are not only used in industrial production to assist people in some dangerous operations,but also used in home life to improve people's quality of life.Path planning is the prerequisite for robots to realize autonomous movement.As the environment becomes more and more complex,the completed tasks become more and more efficient,and highquality path planning algorithms are very important.Traditional path planning algorithm show great disadvantages in complex environments,such as poor adaptability,large amount of calculation,and low efficiency in finding the best.The emergence of intelligent algorithms represented by genetic algorithm and particle swarm algorithm provides a new idea and direction for solving the problem of mobile robot path planning.Intelligent algorithms use relatively simple mathematical models,which are easy to implement and can solve complex optimization problems.But the algorithm itself still has shortcomings,such as slow convergence speed,easy to fall into the local optimal value,etc,so some improvements to the algorithm are needed.After analyzing the two classic algorithms in detail,this paper starts from the problem to be solved,improves them respectively,and verifies them by experiments.The main research work is as follows:(1)Due to the disadvantages of the basic genetic algorithm in the path planning of mobile robots,such as slow convergence speed,excessive path turns and high energy consumption,the following improvements have been made.The same-neighbor crossover is proposed.The crossover operation is performed by selecting crossover pairs with the same feasible neighbor nodes to reduce the amount of calculation and avoid premature convergence.In the mutation operation,an initial mutation node is randomly selected and its feasible neighbor nodes are calculated.The fitness value of the path where the fitness value is the best is the mutation node.The new fitness function takes into account factors such as path distance,safety and energy consumption.In order to verify the effectiveness of this method,it is applied to a variety of different environments.The simulation results show that,compared with other methods,the combination of the improved crossover operator and fitness function helps to obtain the required optimal path.(2)Aiming at the shortcomings of standard particle swarm algorithm in solving the problem of mobile robot path planning,such as slow convergence speed,easy to fall into local optimal phenomenon and unsmooth path,the particle swarm algorithm is improved.This method slightly interferes with the speed of the global optimal particle when the particle falls into the local optimal value,so that the particle can jump out of the local maximum value and continue to search around;in order to balance the local and global search capabilities,nonlinear decreasing inertia weight is proposed.The inertia weight allows the particles to have strong local search ability in the later stage of the algorithm;finally,a fitness function that considers the path length and smoothness is proposed.After multiple experiments,the improved particle swarm algorithm converges quickly,the planned path can avoid obstacles,and the quality is better.The above two intelligent algorithms have been proved by experiments to be effective in solving the problem of mobile robot path planning.Future research work is how to make intelligent algorithms play the greatest advantages to achieve path planning.