Research On Swarm Intelligence Algorithms For Optimizing Power Grid Paths

The optimization of the power supply network path was studied based on the unified development requirements of State Grid Corporation and the latest national power demand level.Path planning of a power supply network is a complex and large-scale nonlinear combinatorial optimization problem.With the economy’s development,the power supply system’s decision variables increase rapidly,and the planning problem becomes more and more complicated.The classic shortest path algorithm Dijkstra algorithm is widely used and mature in path planning.However,due to the uncertainty and universality of the determination of path points and parameter values in the path optimization problem of the power supply network,the global search ability of the entire power grid is poor,and the convergence is also low,which often makes it difficult to determine the final scheme.Moreover,the solution precision and convergence speed may be low when the Dijkstra algorithm solves too many path points.The swarm intelligence algorithm is an efficient meta-heuristic search technology used to solve combinatorial optimization problems in recent years.Compared with traditional algorithms,the correlation algorithm has strong robustness,fast search speed,and high mixed-use efficiency.It has been applied to many typical combinatorial optimization problems with good results.This paper introduces the research status of power network planning and summarizes the existing planning methods.In this study,when solving the routing optimization problem of the power supply network,firstly,the application of Dijkstra algorithm in the routing problem is simulated.Because swarm intelligent algorithm has good optimization ability,this study uses the typical swarm intelligent algorithm ant colony algorithm and its improvement to solve the path optimization problem.This paper first realizes the traditional ant colony-solving path optimization problem.After analyzing why,the basic ant colony algorithm has many parameters and the operation result depends on the choice of parameters,an improved and optimized static matrix ant colony algorithm is proposed to solve the path optimization problem.Firstly,the weight factor is added during pheromone updating to enhance the solving speed.Secondly,the parameter β in the static matrix is analyzed,and the algorithm’s robustness is improved by setting the parameter value range.Finally,2-opt was used to optimize the results and improve the quality of the solutions.Due to the large number of parameters of ant colony algorithm and the significant influence of parameters on the solution,this study introduced the jumping spider algorithm,which has good application in continuous optimization problems,to automatically optimize the parameters of ant colony algorithm.The improved algorithm was finally compared with the simulation experiment data,and in terms of time and accuracy,it was better than the previous single algorithm.Finally,this paper calculates the power flow of the IEEE-6 node using the swarm intelligence algorithm and obtains the lines that need to be planned.According to the results,the system line diagram is drawn.By comparison,the improved ant colony algorithm has the least planned lines and the shortest calculation time,and a satisfactory result is obtained.