Dynamic Service Restoration Of Distribution Systms Involving Microgrids Based On Load Curves And MAS

Uninterrupted and reliable power supply is one of the basic requirements for powersystem operation. But the faulted events are hard to be avoided in large scaledistribution systems during the practical operation. After the occurrence of the fault indistribution systems, the utility shall find the accurate fault location, isolate the fault assoon as possible, and restore the service to the outage unaffected loads in time. Amongthem, the main objective of service restoration procedures is to restore as much load aspossible by transferring outage loads caused by the fault via network reconfiguration toother supporting feeders without violating operation or security constraints. Theeffective service restoration can reduce the unaffected outage area caused by the faultand shorten the outage duration of some customers, thus reducing the outage cost ofsome customers and increasing the energy sales of the utility. Therefore, investigatingan effective method for service restoration of distribution systems is of greatsignificance to improve the security and reliability of the service operation indistribution systems. Based on the currently study on models and methods of servicerestoration of distribution systems, the main works in this paper are as following:①In the aspect of the load models，aiming at the problems of the load modelsused in existing service restoration methods, a dynamic service restoration ofdistribution systems method, which is based on load curves, is proposed in this paper.The proposed method optimally identifies the network configurations to form the set ofcandidate network configurations for each hour over the fault restoration period, takingthe minimum loss of load as the objective and taking the branch currents, node voltages,and radial network configuration as constraints while considering the hourly loadingbased on load curves, and then the service restoration plan which changes the networkconfiguration with the load variation and guarantee the minimum number of switchingoperations. The service restoration plan contains the temporal and spatial dynamiccharacteristics, and can take full advantage of load curves, further reduce the outagearea, shorten the outage duration of customers, thus reduce the cost of customers andincrease the energy sales of the utility.②Combining with the development of distribution systems, investigate theapplication of service restoration to the distribution systems involving microgrids.Aiming at the situation that the microgrids located at the backup feeders after the occurrence of the fault, a dynamic service restoration method, which takes the numberof switching operations into consideration, is proposed for the distribution systemsinvolving the microgrids. This method divides the outage duration hourly. Combiningwith consideration of the characteristic of load variation, establish the multiobjectivemulticonstraint math model for each hour over the fault restoration period consideringthe objective of minimum loss of load and minimum number of switching operations atthe same time. Through solving the math model of corresponding hour, networkconfigurations of each hour over the restoration period are obtained, further to form thecandidate set of network configurations. Then several sorted dynamic servicerestoration plans are obtained according to the pairwise comparison between theobjective of minimum total loss of load and minimum total number of switchingoperations. The multiple plans are convienent for operators to select according to theactual situation. Meanwhile, the proposed method can enlarge the available ampacity ofthe backup feeders by setting the distributed generators of microgrids fully generating,thus enlarging the restoration area and reduction of outage duration.③The future Smart Grid integrates the electricity network and communicationnetwork, containing the ability of information interaction when power delivery. Aimingat the shortcoming of the current centrialized control of service restoration ofdistribution systems, a decentrialized hierarchy MAS is established. The MASincorporates the feeder agents, switch agents, microgrid agents and distributed generatoragents. The data used in service restoration is timely refreshed by the collaboration ofagents, confirming the service restoration plan secure and optimal. On the other hand,the restoration plan is obtained and put into practise by the collaboration of agents.Minimize the workload of the operators by taking the full advantage of thecommunication network and the proposed MAS. The optimal dynamic servicerestoration is put into practice intelligently enhancing the intelligent level of thedistribution automation.④A70-node distribution system is used as the case studies for validating thefeasibility and effectiveness of the proposed method.①and②are realized in Matlabsoftware, while③is realized in JADE platform using the Java-Matlab programmingmethod.