Application Example of Heuristic Algorithm Based on Machine Learning

Path planning in three-dimensional space is an important field of machine learning algorithm research. At present, there are many path planning algorithms, such as heuristic search algorithm, Dijkstra algorithm, reverse incremental search algorithm and so on. Compared with the non-heuristic algorithm, the heuristic search algorithm can more effectively guide the search direction to the target point, avoid the blind traversal of the non-heuristic algorithm, and greatly improve the search efficiency. However, when the heuristic path is blocked, the search effect may be counterproductive. Therefore, how to use the geometric method to fix the search path and make the search area information effective to improve the search effect is the focus of this paper. We need to discretize continuous terrain information into graph information, use computational geometry means, and apply genetic algorithm in machine learning to imitate the random global search and optimization of natural biological evolution mechanism, which provides a new idea for improving the search effect of heuristic algorithm. The improved genetic algorithm has good global search ability, which is helpful to find the optimal solution in three-dimensional space. In this paper, the basic principle, representative research, advantages and disadvantages of heuristic algorithm are introduced in depth, and the problem of heuristic algorithm path blocking is solved by optimizing genetic algorithm, and the future research is prospected.