Font Size: a A A

Study On Time Division Service Restoration Optimization Of Active Distribution Network Considering Load Characteristics

Posted on:2020-04-26Degree:MasterType:Thesis
Country:ChinaCandidate:L H LiFull Text:PDF
GTID:2392330575498362Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
With the continuous access of distributed generation in the distribution network,the distribution network system in the traditional form is gradually changing and gradually developing towards the active distribution network.After the distributed generation is added,when the power supply is restored,the existence of the island is allowed,and the power supply area can no longer have connectivity,the fault recovery process of the active distribution network becomes a distributed generation supply in the fault area as well as the reconstruction of the main network outside the island,and their mutual coordination and cooperation after the fault occurs.In addition to the cooperation between the formation of the island and the reconstruction of the main network,due to the randomness and uncertainty of the distributed generation and the load of each node in the active distribution network will fluctuate with the change of the time of day,when the active distribution network fails,the energy demand supply relationship between the main network power supply,the distributed power supply and the user load may cannot be effectively arranged,and the original failure recovery effect may not be expected.In view of the above problems in the fault recovery optimization of active distribution network,it is of great significance to carry out the research on time division fault recovery optimization of active distribution network with load characteristics.In this thesis,when calculating the power flow of the time division fault recovery optimization of the active distribution network with load characteristics,the nodes of the distributed generation are processed into four types:PQ,PV,PQ(?)and PI to participate in the Newton-Raphson method.When the distributed generation is connected to the IEEE standard 33 nodes distribution network system,the impact on the node voltage and the impact on the network loss shows that:when the position of the DG access node is closer to the power supply point,the effect of voltage boosting is less obvious;the farther the location of the access node is from the power point,the more obvious the effect of voltage boosting;as the capacity of the distributed generation increases,the network loss appears to decrease first,and when it decreases to a certain extent,it begins to increase again;when a distributed generation of the same capacity is connected to the distribution network,the nodes that are accessed are different,and the network loss is also different,when the distributed generation is connected near to the power point,the network loss is large,and when the distributed generation is accessed away from the power point,the loss of the network is low.In the problem of active distribution network fault recovery and time division,this thesis uses K-Means clustering algorithm to divide the time period in the active distribution network fault recovery process.Then,taking into account the load characteristics and the uncertainty conditions of distributed generation output,using the improved genetic algorithm,through the addition of virtual branches,the global optimization of the main network reconstruction and the formation of distributed generation islands is realized.Through the analysis of the failure recovery examples of the IEEE standard 33 nodes distribution network system and time-varying IEEE standard 33 nodes distribution network system,it is shown that by adding a virtual branch between the power point and the BDG node,the global optimization between the main network reconstruction and the island formation after the fault occurs can be well controlled by the objective function of the genetic algorithm;through the variable population method,adaptive method and a certain degree of retention for individuals who do not meet the constraints of the distribution network,the local convergence of the genetic algorithm can be effectively reduced;by using K-Means clustering algorithm to divide the time period in the active distribution network fault recovery process,the fault recovery scheme of the active distribution network is different with time,which makes the fault recovery strategy more timeliness.
Keywords/Search Tags:active distribution network, service restoration, time division, improved genetic algorithm
PDF Full Text Request
Related items