Path planning method for inspection robot in data center based on improved ant colony algorithm

Abstract

The invention discloses a path planning method for an inspection robot in a data center based on an improved ant colony algorithm, comprising the following steps: S1, initializing information, and simultaneously reading the shortest distance matrix and the nearest neighbor matrix between all points of entire station stored in advance; S2, obtaining a corresponding shortest distance matrix according to the current task point number. When the number of task points is large, the time for the path planning is controlled to be within one second, so that the repeated path is significantly reduced, and the total path is significantly shortened. At the same time, a new inspection path can be independently determined and re-planned according to the completion of an inspection task after encountering the obstacle. The algorithm has good robustness. In short, according to the path planning method for the inspection robot in the data center based on the improved ant colony algorithm, the inspection efficiency of a laser navigation inspection robot in the data center is significantly improved, so that the inspection robot can complete the task to be inspected in a shorter time, and at the sametime the intelligence and autonomy of the inspection robot is improved.

Description

technical field [0001] The invention relates to the technical field of inspection robots, in particular to a path planning method for a data center inspection robot based on an improved ant colony algorithm. Background technique [0002] At present, data centers use inspection robots to replace manual inspections of instrumentation equipment in data centers, in order to overcome the problems of heavy workload, high risk factor, low efficiency and poor reliability faced in manual inspections. Among them, path planning is the basis for whether the inspection robot can efficiently complete the inspection task, and at the same time affects the intelligence and performance of the inspection robot. At present, the local path planning and global path planning of the laser navigation inspection robot in the data center both use the Dijkstra algorithm. This method only considers the optimal path locally, but does not consider the optimal path globally. There are many repeated paths. ...

AI Technical Summary

Problems solved by technology

At present, the local path planning and global path planning of the laser navigation inspection robot in the data center both use the Dijkstra algorithm. This method only considers the optimal path locally, but does not consider the optimal path globally. There are many repeated paths. The shortcoming of the inspection path is long. At the same time, after encountering obstacles, the inspection will stop regardless of whether the inspection task is over. It lacks autonomy and cannot fully reflect the intelligence of the inspection robot.

At the same time, the data center inspection robot also uses the Dijkstra algorithm to plan the local path, and the simulated annealing algorithm to plan the global path. When the number of task points is relatively small (less than 100), the time-consuming planning and path length of this method are reasonable, but when the task When the number of points increases, this method is time-consuming and not suitable for large-scale inspection tasks

[0003] In this field, China's No. 201510290471.5 invention introduces a global path planning method for inspection robots based on topological point classification. The topological points of the stops are classified and merged according to the path they belong to, and the directed data structure corresponding to the topological points is established. When planning, only the topological points of the robot’s current position and target position are inserted into the directed graph data structure, which can significantly reduce the number of topological points involved in the calculation method, improve the computational efficiency of path planning and reduce the consumption of storage space. However, currently based on laser navigation The path planning algorithm of the inspection robot is planned on the basis that the relationship between topological points is a directed graph. There are many repeated paths and the total path is long. After encountering an obstacle, it stops immediately regardless of whether the inspection task is over. In the inspection, it is impossible to independently judge and plan a new inspection path based on the execution of the inspection task. Moreover, in the path planning algorithm schemes of some data center inspection robots, when the number of task points is large, the path planning is relatively difficult. Time-consuming. In addition, the current path planning method for the inspection robot in the data center has not realized that after the robot encounters an obstacle, it can determine whether there are still task points that have not been inspected according to the inspection situation of the task, and re-plan its path to ensure the inspection. Therefore, we propose a data center inspection robot path planning method based on the improved ant colony algorithm to solve the above problems

Images