Research On Improved Wolf Pack Algorithm And Its Application In Robot Path Planning

Wolf pack algorithm is a kind of swarm intelligence algorithm which has emerged in recent years.It is an optimization algorithm generated by simulating a series of behaviors and characteristics generated during the hunting process of wolves.Wolf pack algorithm is an iterative way of group random optimization.It is optimized by simulating the various behaviors performed during hunting by the wolves.Because the Wolf pack algorithm's optimization of the problem shows high accuracy,fast convergence,easy implementation,and good stability,it has been applied by some scholars to actual production practice problems such as neural network,production control,sensor optimization,and path planning.And get a good optimization result.However,the theoretical system of wolf pack algorithm is not mature enough,and further research and development are still needed to solve complex practical problems.The study found that the main problems existing in wolf pack algorithm are: for the optimization of most test functions,the accuracy of the algorithm needs to be improved;the algorithm is very random when the population is initialized;the algorithm is very easy in the later iteration Into the problem of local optimization,the algorithm is studied and improved from three aspects.They are single-objective optimization problems,multi-objective optimization problems and practical applications of algorithms.Here are three aspects of the study and improvement of this article:For the single-objective optimization problem,the problems such as low convergence precision,slow optimization speed and easy to fall into local optimum are studied.The analysis considers that there is a problem of too much randomness in the initialization process of the wolves,resulting in uneven distribution of the initial population and reducing the optimization performance of the algorithm.In order to improve the quality of the wolves in the initial population,Tent chaotic mapping strategy was introduced in the wolves initialization.According to the analysis,because the siege behavior exists because the convergence is too fast and the algorithm is premature,the introduction of Levy flight strategy in the siege behavior can effectively help the algorithm to escape from the local optimum.A modified wolf pack algorithm based on Tent chaotic mapping and Levy flight(TLWPA)was proposed based on the Tent chaotic map and Levy flight.The experiments were compared with other algorithms on the test function,and the TLWPA algorithm was verified.Good optimization performance on single-objective optimization problems.For the multi-objective optimization problems,the following improvements have been made: the elite individual search behavior based on improved differential evolution has been added to the algorithm;combining the sinusoidal search behavior of the search wolf with the elite individual search behavior based on improved differential evolution,the search behavior is Improve the search efficiency by the role of individual individual information;introduce pareto's difference entropy and individualized chaos factor into running behavior,and use pareto's difference entropy to understand the distribution uniformity of solutions in Pareto's Pareto optimal solution set to understand population convergence status and personalize The chaotic factor enhances its ergodicity,enables the wolf to dynamically adjust the wolf behavior behavior according to the state of population convergence in the running behavior,continues to introduce the Levy flight strategy in the siege behavior,and uses the individual mass comparison method based on the aggregation density,which can Effectively improve the uniformity of the optimal solution.An improved Multi-objective Optimization based on Wolf Pack Algorithm(MO-IWPA)is proposed.Through comparison experiments with other algorithms on the test function,the MO-IWPA algorithm is proved to have good performance on multi-objective optimization problems.In the robot path planning,the obstacle movement environment of robot movement is modeled by grid method.This is the case where the starting and ending points are known,the distribution of obstacles is clear,and the obstacles are fixed.The TLWPA algorithm is applied to it,and through the comparison experiment with the genetic algorithm,the TLWPA algorithm is verified to have a good ability to optimize the robot path planning problem.In summary,based on the MWPA algorithm,Tent chaos mapping strategy is introduced in the initialization phase of the wolves,and Levy flight strategy is introduced in the siege behavior.The proposed TLWPA algorithm improves the convergence speed of the algorithm in solving single-objective optimization problems.And optimization accuracy;in the multi-objective optimization problem,the concept of Pareto difference entropy and individualized chaos factor was introduced into the running behavior by combining the sinusoidal search behavior with the elite individual search behavior that improved differential evolution,and Levy flight was introduced into the siege behavior.The strategy,combined with the individual-mass comparison method based on aggregate density,presents the MO-IWPA algorithm with better convergence and diversity in multi-objective optimization problems.In the application,the TLWPA algorithm is applied to the path planning problem of the robot.The planning path obtained through the simulation experiment is shorter than the planned path obtained by the genetic algorithm,and the route is smoother and less time-consuming.The TLWPA algorithm shows better performance in practical applications.