Research On Intelligent Algorithms In Path Planning Of Mobile Robots

Mobile robotics is a multidisciplinary field that integrates knowledge from multiple fields such as intelligent systems,sensors,cognitive biology,and applied electronics,involving many cutting-edge technologies in the current world.However,there are still some problems in path planning for mobile robots,such as the search path easily falling into local extremum,and the slow convergence speed of the algorithm.To address the above issues,the main research content of this article is as follows:1.An ant colony algorithm based on potential field is proposed for the path planning problem of a single target robot with known obstacle information in the environment.However,the ant colony algorithm has the disadvantage of slow convergence speed and being prone to falling into local extremum during path optimization: incorporating the gravitational effect in the potential field algorithm into the heuristic information to improve the optimization ability of the ant colony algorithm;A non-uniform pheromone initialization distribution is designed to reduce the blindness in the early stage of the ant colony algorithm and achieve the goal of fast convergence of the algorithm;When updating information,the information of the best and worst individuals in each generation of the population is added to the update to avoid the algorithm falling into local optima.The simulation results show that the improved algorithm has improved convergence speed and optimization performance.2.An ant colony algorithm based on Laplace distribution and dynamic window fusion is proposed for path planning of single target mobile robots with unknown obstacles in the environment.However,the path of ant colony algorithm is not smooth enough to complete local obstacle avoidance.The improvements are as follows: Firstly,improve the heuristic information of ant colony algorithm,and add dynamic adjustment factors to enhance the guidance ability of heuristic information in the early stage,while enhancing the guidance ability of pheromone in the later stage of iteration;Additionally,the Laplacian distribution is introduced into the pheromone update of the ant colony algorithm to adjust the volatilization of pheromone and accelerate the convergence speed;Then,perform path geometry optimization on the obtained path to remove redundant nodes and improve path smoothness;Ultimately,the improved ant colony algorithm is integrated with the improved dynamic window algorithm to ensure that the robot safely reaches the endpoint.Simulation shows that the improved ant colony algorithm has improved optimization performance and can complete local path planning.3.A multi-objective slime mold algorithm based on Baldwin is proposed for the global path planning problem of multi-objective mobile robots.In an environment with complex obstacles,the optimization objects are path length,path smoothness,and path safety.The Baldwin learning mechanism is added to MOSMA to guide the population evolution direction and accelerate the convergence of the algorithm.The K-nearest neighbor metric is used to prune the solution set to obtain a more uniformly distributed solution set.Verified on the testing problem platform,the proposed algorithm outperforms other classical multi-objective algorithms in terms of convergence and uniformity.Ultimately,in MATLAB simulation experiments,compared with the MOSSA algorithm,the three indicators in Environment 1 improved by 1.82%,66.63%,and 0% respectively,while in Environment 2,they improved by 2.88%,16.99%,and52.17%,proving that the multi-objective slime mold algorithm achieves higher solution quality.