Coding and genetic algorithm suitable for power distribution network and application in distribution network reconfiguration

Abstract

Disclosed are a coding and genetic algorithm suitable for a power distribution network and application in distribution network reconfiguration. Coding is conducted through names of nodes, depth of the nodes and the degrees of the nodes, male parents are selected, a later generation is obtained by cross and mutation breeding through a cross operator and a mutation operator, when the male parents are selected to be crossed and mutated, one male parent is selected at random while the other male parent is selected optimally. The cross operator is used for exchanging a connection relation between a certain node and a father node in each tree mapped by two codes. The mutation operator is used for changing a father node of a certain node in each tree mapped by the codes, wherein the certain node is provided with spare branches. Radial constraint does not need to be verified, the situation that an invalid network is operated and repaired repeatedly to obtain a valid network is avoided, and computing time is greatly shortened. The valid network can be obtained certainly in networks of any scales, and the problem that an existing method cannot be applied to reconfiguration of large-scale actual power distribution networks.

Description

technical field [0001] The invention belongs to the technical field of power distribution network automation. Background technique [0002] Distribution network reconfiguration has the function of optimizing network loss, voltage quality, power supply reliability and other operating indicators by changing the network structure through the switching operation of the existing network without adding additional economic investment, which is a part of distribution network automation. The distribution network reconfiguration problem is an NP-hard problem, and there is no mathematical analysis method to solve this problem. At present, except for a few simple heuristic algorithms based on physical models, most of the power distribution network reconfiguration uses intelligent algorithms to solve this problem. Theoretically, the intelligent algorithm can obtain the optimal network reconfiguration scheme globally, but the simple heuristic algorithm of the physical model cannot guaran...

AI Technical Summary

Problems solved by technology

The first one uses each switch of the power distribution network as an independent variable, and the on-off state of the switch is coded with 0-1. Most of the networks generated by this method are not effective networks, and the coding efficiency is extremely low.

The second method uses the basic loop of the power distribution network as an independent variable to encode the number of switches opened in the loop. The coding efficiency is greatly improved in small-scale power distribution networks compared with the first one, and is currently commonly used. The best coding method, but the coding efficiency of this method will drop sharply in large-scale power distribution networks, and it will also produce a large number of invalid networks

The power distribution network has a mandatory requirement for radial open...

Images