Research On Improved Artificial Potential Field Path Planning Algorithm

In the rapid development of artificial intelligence today,the problem of path search planning technology has been a tricky topic.How an intelligent body can autonomously plan a path to ensure that it can smoothly arrive at the destination from the starting point without obstacles is the work it needs to do autonomously in the environment it is in,and this route should be optimal or suboptimal.For example,when performing the task of autonomous driving,if the vehicle does not have a good path planning capability,it will delay the arrival time in a light way,or in a heavy way,it will have a traffic accident,causing economic losses and even threatening lives.In the process of task management for an intelligent body,its ability to avoid obstacles and the efficiency of performing path planning determine the overall efficiency of its task execution.Therefore,whether in the industrial,civil or military fields,a reasonable path planning for an executing intelligent body plays a role in determining the success or failure of a mission.Traditional path planning design algorithms do plan and output the resultant paths well when the pre-verified information in the environment is known,but in those unknown environments,especially those complex environments with various irregular obstructions,many planning algorithms will soon have outlived their usefulness.The optimal paths derived by traditional methods based on incomplete information are not guaranteed to be optimal or even feasible with complete information,do not have the ability to respond to environmental changes,lack robustness,and do not meet the requirements of "intelligence";on the other hand,traditional methods can optimize a static environment without considering time consumption and other factors.However,when the environment changes dynamically,the computational efficiency of the traditional algorithm is not enough to catch up with the environmental changes,and the planned paths are out of order and of no value.The optimal path planned by the traditional method based on static environment may collide with obstacles on the way to actual operation;even if the traditional path planning method is used for emergency obstacle avoidance after obtaining the environment change,it may cause untimely response due to deadlock or inefficiency and other problems,resulting in mission failure.The ever-evolving science and technology has increased the demand for dynamicity and accuracy of path planning technology,because traditional path planning methods have been difficult to cope with the increasing complexity of the current intelligent planning problems and their increasing application areas.In particular,in military confrontation,path planning in an environment of dynamic and incomplete information is an ever-present problem,and traditional methods are unable to meet the challenges of such demanding dynamicity and accuracy requirements,and combat units are likely to be destroyed by the time they react.For example,in military applications,path planning for static environments is difficult to ensure that combat units do not encounter danger on the way,because the environmental information obtained before path planning is usually incomplete;at the same time,in the event of a surprise attack on the way,the pre-planned route will not help combat units to evacuate quickly,timely,and safely,which may result in " The classical artificial potential field method(APF)is a method that is used in the field.The classical artificial potential field(APF)is a method that uses potential energy functions to construct a virtual potential field in which the intelligent body is instinctively displaced by potential energy differences.The traditional artificial potential field method,which has the advantages of small computation,high real-time performance and smooth planning route trajectory,is widely used in the local real-time route obstacle avoidance of intelligent body,but it has two inherent technical defects:one is the local minimum problem and the other is the target unreachability problem.In this paper,we propose a method that introduces a potential energy adjustment control factor related to the relative moving distance between the intelligent body and the moving target,which ensures that the target point is always the full potential energy minimum in the potential field and solves the problem of local minima.Meanwhile,this paper also proposes a detection method based on the movement steps and displacement of the intelligent body and an escape strategy for temporary sub-targets,which solves the target unreachability problem caused by the balance of attractive and repulsive forces.In addition,this paper introduces the Rapidly Exploring Random Tree(RRT)method,which is a random sampling-based method with both probabilistic completeness and good convergence performance,in the process of global path planning using the artificial potential field method.By analyzing the principle of algorithm expansion,we propose APF-RRT,a method that combines goal-oriented and stochastic properties to solve the above-mentioned local minima problem,based on the improvement of the existing artificial potential field method.On the other hand,the introduction of the artificial potential field method reduces the search space of the RRT method.Finally,to address the feature that the raw data of the posture provided by the war game projection platform is not easy to use directly,this paper proposes a path planning problem solving framework for the war game projection platform,which consists of three independent modules responsible for encoding the posture,solving the path planning problem and decoding the posture,so as to realize the fast solution of the path planning problem for the war game projection platform.On the basis of this framework,this paper compares the proposed method with the traditional path method using the war game projection platform,and again illustrates the advantages of the proposed method.In short,the APF-RRT fusion method is proposed to address the potential deadlock problem and the inefficiency of fast extended random tree search in the artificial potential field method,and the proposed APF-RRT path planning method is verified on the war game projection platform.