Research and application of inspection path planning of heat exchange station based on improved A* algorithm and K-means

In order to solve the optimal path planning problem of multi-team heat exchange station inspection, the path planning model of heat exchange station inspection was established to minimize the driving distance of patrol vehicles, and A multi-team optimal path planning strategy based on A* algorithm combined with K-means algorithm was proposed. The strategy consists of two stages. In the first stage, K-means clustering algorithm is used to divide heat exchange stations with similar distances into inspection tasks for a group of personnel, and the Haversine formula is used to replace the Euclidean-style distance calculation, which greatly reduces the distance deviation. In the second stage, for each group of inspection tasks, the A* algorithm is improved to calculate the visiting order of heat exchange stations and the path between stations. This algorithm adopts the dynamic neighborhood search strategy, dynamically assigns the optimal search direction to the current nodes in real time according to the current adjacency node relationship, and adds variation factors into the extended node selection process. The algorithm can jump out of the trap of local optimal and continue to search for a better path. Simultaneously, a novel heuristic function is devised based on the Haversine formula to enhance its suitability for searching in road network environments, achieve global shortest path determination, and reduce path finding time. Ultimately, the algorithm was practically implemented on a geographic information visualization platform utilizing Amap API.