Study On Power Distribution System Fault Location And Topology Reconfiguration

As one kind of public utilities,power distribution systems play an extreme important role in social production,daily life and so on.It is of great support meaning in the realization of the grand goal of building a moderately prosperous society.But due to its large scale and complicated operating environment,power distribution systems are much more liable to suffer from faults than transmission systems.Faults will greatly decrease the reliability of power supply.Fast fault location and isolation is an important way to improve the reliability of power supply.Fortunately,the great improvement of distribution automation provides useful information and data for the location and isolation of faults.Another measure to improve the reliability of power supply is to restore the non-fault area as fast as possible.Power distribution systems are designed as weekly meshed networks,however,for the purpose to simplify the setting of relay protection,they usually operate under open loop conditions.When a fault occurs,the whole feeder will be cut off from the system by opening the circuit breaker which is located at the beginning of the feeder.This,to some extent,extends the blackout area.Therefore,after fault is located and isolated,reconfigure the system can restore the non-fault areaAlthough the level of distribution automation in our country has been improved significantly,there is still a gap with other countries.In order to further narrow the gap and to further improve the reliability of power supply,two documents are made by the National Development and Reform Commission and the National Energy Administration.Based on the two documents,this dissertation studies the power distribution system fault location method and topology reconfiguration method to improve the power supply reliability and fault self-recovering ability.The main research contents and contributions are as follows:(1)Study on power distribution system fault location method based on multisource information.At present,the research area of distribution system fault location is mainly in high voltage and medium voltage distribution networks.Fault location methods for low voltage distribution systems are seldomly reported.Also,most of the existing methods utilize only one kind of fault information.When there is fault information loss problem caused by interruption of communication,these methods may not be available.With the development of distribution automation,different kinds of automation systems(e.g.distribution automation system-DAS,customer electric information acquisition system-CEIAS,grid production management system-GPMS,trouble call system-TCS)and monitoring terminals(e.g.feeder terminal units-FTUs and smart meters)are installed.These systems and devices provide a large amount of information and data to monitor the operating condition of distribution systems.And based on these multisource information,this dissertation presents an automatic faulted line section location method for distribution systems.It can be divided into two parts:faulted line section location method for medium-voltage distribution feeders and faulted area identification technique for low-voltage distribution systems.Then,a cooperation strategy of the two parts is proposed.The proposed method extends the research field of fault location to low voltage distribution systems.(2)Study on faulted feeder section location and isolation method for power distribution systems considering the change of topology.For the purpose of fault isolation,non-fault area restoration and network reconfiguration,the topology of distribution systems change frequently.However,the frequent change topology brings many difficulties to fault location.To overcome this,in this paper,a novel automatic matrix-based algorithm for the identification and isolation of faulted feeder sections on distribution systems is proposed.The algorithm works in two stages:the first stage automatically identifies the radial feeders that make up the whole system and represents the feeders' topology in matrix form;the second stage automatically identifies the faulted section of the identified feeder and opens the relevant switches to isolate it.The algorithm can be applied to single and multiple faults.And as it operates by using measuring device information and detecting the status of switch devices,the method does not require any electrical parameters and it is not affected by the fault type or fault resistance.At last,the cooperation strategy of the two parts is also proposed.(3)Study on split-phase power flow calculation method for unbalanced three-phase distribution systems.Firstly,different from usually balanced transmission systems,the three phases of distribution systems are generally asymmetrical for the existing of both single-phase and three-phase loads.Secondly,for the purpose of topology reconfiguration and the optimal configuration of reactive power compensation,vast amount of load flow calculation work needs to be done.Thirdly,in the next research content,a split-phase topology reconfiguration method for unbalanced distribution systems is proposed.When the method is applied,the direction of the currents flowing in the three phases may be totally opposite.Thus,the traditional power flow calculation method may be no longer applicable.In view of these three issues,a novel power flow algorithm considering the associative direction of the three phase currents is proposed.It is a forward-backward sweep method.For the existence of mutual inductance,split-phase not means to calculate the power flow for each phase individually.Test cases that are studied using the IEEE 33 node system demonstrate the effectiveness of the proposed method.(4)Study on power distribution system reconfiguration method based on improved Genetic AlgorithmFast service restoration reconfiguration is another way to improve the reliability of power supply.Genetic Algorithm is an important tool to solve the network reconfiguration problem.However,the following two difficulties prevent the application of the method in reality.Firstly,each time a fault occurs,the initial population for the new system after the fault is located and isolated should be reformed.However,there is not an automatic method to do this work so far.Secondly,due to the topological constraint,infeasible individuals are very likely to be generated during the evolution.Many new coding strategies have been developed to solve this problem,however,these methods still have different problems.To solve the first difficulty,an automatic initial population formation method is proposed.It first simplifies the system and forms the branch relationship matrix between the original system and the simplified network.Then,an initial population formation method based on the matrix and the simplified network is developed.The method can guarantee that all the generated individuals are feasible.For the second difficulty,an automatic mesh-check algorithm is proposed.It can identify whether there are loops or isolated nodes in the system which is represented by the codes.For the individuals that violate the topology constraints,they will be replaced by the new feasible individuals that are generated using the initial population formation method.In addition,to improve the convergence performance of the Genetic Algorithm,a guided mutation operator is devised.The operator firstly determines all the feasible allelic gens,then to select the best one according to the fitness.Most of the existing topology reconfiguration methods assume the distribution system is symmetric.Thus,only one phase is studied.Although there are network reconfiguration methods developed for unbalanced distribution systems,they give the same reconfiguration strategy.The three phases will be opened in the same positions.In this paper,a novel split-phase topology reconfiguration method for unbalanced three phase power distribution system is proposed.Split-phase means the three phases adopt different reconfiguration strategies.Under this conception,the new mathematic model of topology reconfiguration is reconstructed.And the proposed Genetic Algorithm is modified to solve the model.Test results of different cases demonstrate that split-phase reconfiguration method can obtain better result than non-split-phase methods.