Research on Power Grid Inspection Path Based on Edge Computing

Abstract

With the development of intelligent power systems, establishment of intelligent inspection paths is studied in this paper so that disadvantages such as inaccessible network, cumbersome operation, and low safety performance in existing power grid inspections can be solved. Besides, power network topology is constructed based on study of characteristics in power grid environment and requirements of inspection tasks. Moreover, two-layer heuristic algorithm is proposed in this paper based on edge computing. On the basis of ant colony algorithm, deviation degree is used to guide pheromone update so as to solve dual-objective problem in path planning. Meanwhile, annealing mechanism is introduced to effectively avoid “premature” phenomenon in ant colony algorithm, which speed up algorithm convergence. Every time simulated annealing runs, entire ant colony algorithm needs to be run. After all ants complete a round of search, optimal ants will be selected for node swapping, inversion, and translation to generate new solutions and accept new solutions according to annealing rules, which completes combination of simulated annealing and ant colony algorithm. After simulation experiment analysis, it is proved that research method proposed in this paper effectively improves global search ability and avoids falling into local optimum. Therefore, if deviation degree is used as the criterion to evaluate pros and cons of solution as well as guide pheromone update, convergence speed of algorithm can be accelerated.