Research On Smart Energy Distribution Optimization Scheme Based On Edge Computing

With the large-scale construction and gradual development of microgrids,in order to achieve efficient,low-latency data processing and meet the needs of smart grid users,emerging smart energy systems are constantly deployed at the edge,integrating edge computing modules into the microgrid system.Therefore,the corresponding energy systems can effectively extract terminal equipment data nearby,carry out real-time and reliable data analysis,realize the control,prediction and analysis of power generation behavior of the microgrid power generation capacity,while dispatching the power distribution interaction between the microgrid generators and usersand completing security microgrid control of the microgrid at the optimal cost under load balancing conditions.This paper intends to study the optimization design of micro-grid distribution based on data query analysis and graph calculation in the edge module.The main work of the paper is as follows:Each power generation equipment and user of the microgrid can be modeled as countless nodes in the graph.The edge computing module will use the users,electricity consumption behavior and power consumption in the microgrid as design parameters to establish a node data information database.This paper innovatively proposes a two-layer optimization control model for microgrid.In the upper model,the data feature information is obtained through the mean-shift clustering method,and optimized according to the objective function of the constructed microgrid user node with the least comprehensive cost.Using particle swarms,the algorithm obtains the optimal solution of the design model,and calculates the weight ratio of each cost attribute of the graph node through the normalization function and the weight ratio formula.On the basis of determining the weight of the graph nodes,the Dijkstra algorithm is adopted for the generated weighted graph to solve the optimal solution of the shortest path between the two nodes,and an optimal matching scheme for the user’s power transaction is given.According to the scenarios with limited computing and storage resources in the edge environment and the diverse and complex data characteristics of the access terminal,sparse matrix and memorized search are used to optimize and improve the Dijkstra path search algorithm,without destroying the overall structure of the dot network system diagram.By subtracting unnecessary nodes,the storage space occupied by them is reduced by more than 96%.Under the condition that the optimal solutions of all paths of the original algorithm and the improved algorithm are consistent,the operating efficiency of the proposed algorithm is increased by 45%,and the budget speed is increased by about 40%,which verifies that the improved algorithm cannot sacrifice computing performance.Meanwhile,The algorithm implementation complexity is extremely lower,and the operating efficiency is significantly improved.Finally,this article applies the improved algorithm to the smart energy edge computing system,and builds the micro-grid power distribution optimization module of the smart energy system,which can run on the actual grid system.To sum up,this article takes edge computing as the application scenario,mainly studies and analyzes the micro-grid power distribution optimization plan.Besides it conducts model establishment and simulation analysis of the micro-grid system.Furthermore,this thesis conducts algorithm derivation of optimization algorithms and graph search algorithms,while giving some classical algorithm examples and example verification et al.Finally,this thesis improves and optimizes the core algorithm,and applies it to edge boards to achieve smart energy distribution optimization,micro-grid user transactions and other business functions.