Research On Calculation Framework Design And Resource Optimization Scheduling Of Intelligent Distribution Network

With the rapid development of power construction,as a smart distribution network that directly interacts with power users,its operational safety and reliability will affect the power supply reliability,operating efficiency and user experience of power system services.At present,the explosive growth of smart distribution network terminal equipment has brought computing,storage and bandwidth pressure to traditional computing platforms,and the smart distribution network system is facing new opportunities and challenges.In recent years,many researchers have introduced edge computing technology into traditional computing frameworks,and built a cloud-edge-terminal three-layer computing framework to process large amounts of data generated by terminal devices in smart distribution networks.However,with the surge of data volume,the basic three-layer computing framework has been unable to meet the needs of massive task data.In addition,most current research has not considered the selection and deployment of edge nodes at the edge layer.Similarly,the reasonable scheduling of computing resources based on the computing framework also affects the reliable operation of each application system of the smart distribution network and the security of power grid operation.Therefore,it is of great significance to construct a reasonable intelligent distribution network computing framework and corresponding resource scheduling algorithms.This paper firstly proposes a two-level edge architecture on the basis of the cloud-edge-end three-layer computing framework,and points out the deployment location of edge nodes to realize the intelligent distribution network.Hierarchical control of the grid.Then,the resource scheduling under this framework faces the problems of large amount of tasks and computing resources,various computing types,and task preferences.An improved artificial bee colony resource scheduling algorithm based on bilateral clustering is proposed.The algorithm first performs bilateral clustering preprocessing according to the characteristic attributes of tasks and computing resources,improves the fuzzy C-means clustering algorithm,and conducts simulation experiments on the improved clustering algorithm.The shortcomings further improve the execution efficiency and accuracy.Then the same type of task class and computing resource class are matched,and the resource scheduling algorithm is further executed in the same type of task class and computing resource class set.For this purpose,an improved artificial bee colony algorithm is designed,which is introduced into the basic artificial bee colony algorithm.Simulated annealing algorithm and tournament selection method overcome the shortcomings of easy precocity and easy to fall into local optimum.Finally,the simulation experiment verifies that the improved artificial bee colony resource scheduling algorithm based on bilateral clustering is feasible and effective in the optimization of task completion time,load imbalance and task demand satisfaction.