On this page

Optimization study of feeder consumption carrying capacity of low and medium voltage distribution networks with deep learning support

By: Shiyuan Ni 1, Sudan Lai 1, Lu Tang 1
1State Grid Fujian Economic Research Institute, Fuzhou, Fujian, 350012, China

Abstract

This paper addresses the lack of simple and effective informatization analysis tools for the selection of new energy and user access points by the planners of medium and low voltage distribution networks, considers the different types of new energy accessed by different distribution networks at different levels, proposes the objective function of assessing the new energy maximum carrying capacity of the distribution network by using the minimum of the maximum access capacity of the distribution network under each scenario and sets the constraints of assessing the safe operation of the grid at all levels, the power of the contact line and the power of each node to form a cooperative assessment model of the new energy carrying capacity of multi-level distribution networks. A hybrid optimization algorithm based on Gray Wolf algorithm and cone planning is applied to solve the carrying capacity optimization model. Compare the solution speed and accuracy of this algorithm with those of BONMIN solver, GA, PSO, SCOP and GWO. Analyze the distributed PV carrying capacity under different access methods and the effect of PV carrying capacity improvement. In the typical scenario of distributed PV, when the average value of PV output increases, the single power system starts to show light abandonment near the peak of PV output. Compared with the single power system, the Jieyang regional integrated energy system has a shorter abandonment period and a smaller amount of abandoned light, which indicates that the Jieyang regional integrated energy system based on the new energy carrying capacity assessment platform of the multilevel distribution grid has a better access scheme, produces a smaller amount of abandoned light, and has a higher level of distributed PV consumption.