Research On Robot Path Planning Based On Improved Intelligent Optimization Algorithm

In recent years,with the development of technology,artificial intelligence technology has become increasingly widely used in various industries.Intelligent mobile robots,as an important branch of artificial intelligence,have been applied in many fields,such as heavy and dangerous work such as cargo handling and terrain investigation.Therefore,the development and design of mobile robots is an extremely important research direction.In the technical spectrum of intelligent mobile robot research and development,path planning technology is one of the core technologies.Efficient path planning ability can bring faster execution efficiency and lower energy consumption to the work of mobile robots,thereby bringing economic benefits to relevant enterprises and individuals.This paper focuses on the research of relevant algorithms in three problem models: global path planning,local path planning,and uncertain environment.Based on the characteristics of algorithm applicability,pathfinding efficiency,and robustness,ant colony algorithm and dynamic window method are selected as the research algorithms for global path planning and local path planning,and improvements are proposed to address some of the problems that exist in the two problem models,and relevant experiments were conducted to validate the improvement strategy.Finally,the two algorithms were fused and used to solve path planning tasks in uncertain environments.The main work of this paper is as follows:(1)When ant colony algorithm is solving the global path planning problem,the pheromone concentration is calculated by each entire path,and this method will cause the high-quality local path segments in the poor-quality path to be ignored.In this paper,a terminal distance index strategy is proposed to replace the pheromone concentration,and the high-quality local path segments are constantly stacked,thus enhancing the efficiency of finding the optimal path.Secondly,by studying the performance of different step sizes in different complexity environments,it was found that the step size with the best overall performance is 2 or 3.By integrating the multi-step strategy with the final distance index,the defect of redundant road sections in the path obtained by the multi-step strategy was improved to some extent,in order to achieve better results.(2)In response to the fact that the dynamic window method only uses trajectory prediction information from different behavioral decisions at the current moment to determine the quality of decisions when solving local path planning problems,without considering the quality of the next stage,a feedback adjustment factor is proposed based on the idea of Bayesian decision-making to enable the algorithm to consider not only the current quality of behavior decisions,but also the current situation,At the same time,it can also consider the future benefits generated based on current decisions.At the same time,considering the low utilization of environmental information in traditional DWA,an iterative modified potential field strategy is proposed to generate a force field and use it to construct a force field function,improving the evaluation function of the DWA algorithm.(3)For path planning under dynamic uncertainty in the environment,an adaptive cubic spline curve path optimization method is first used to curve the optimal path obtained from global path planning,generate a collision free curve path,and use this curve as a guide curve to further improve the evaluation function of DWA,thereby further improving its efficiency in utilizing environmental information,Propose a two-stage evaluation function for obstacle avoidance,which allows mobile robots to move according to the guidance curve by default.When sensors such as Li DAR detect the presence of dynamic obstacles and the obstacles affect the robot’s movement trajectory,the dynamic window method modified in this article will be temporarily called for local path planning to avoid obstacles.When the robot reaches a safe area,it will return to its original movement state.Continuously repeat the above behavior until reaching the predetermined target point to complete the task.(4)To test the actual effectiveness of the improved algorithm in this article,a three-wheel omnidirectional mobile robot controlled by ESP32 was used to test the performance of the algorithm.The robot is equipped with YDLIDARX2 Li DAR and uses the ROS robot system.Modeling the experimental environment through the gmapping function package and visualizing the environment map through rviz.Finally,the proposed IACODWA algorithm was tested in static environment path planning experiments and dynamic obstacle avoidance path planning experiments,and the results were compared with the traditional dynamic window method to demonstrate the innovation and effectiveness of the proposed IACO-DWA algorithm.