Research Of Coordinated Planning In Medium-voltage Distribution System For Improving Power Supply Capability

As the link of end-users and generation and transmission system, distributionsystem is an important part of the reliable and safe power supply. With the rapideconomic development of cities, especially in large cities, the growing tension in landmakes it difficult to select the substation location and electricity access corridor. If thehigh-voltage (HV) substation and medium-voltage (MV) distribution network areplanned in accordance with the traditional technical principles, it will be very difficult tosupply the power demand of all-level users, and to reduce the scale of construction andthe consumption of land resources concurrently.Therefore, it is necessary to carry out in-depth research of coordinated planning ofpower distribution system at the planning level to enhance the utilization level of powersystem equipment and to reduce the occupation of the land resources, which is an urgentproblem in the development of distribution system at the present stage in China. Focusedon the HV substations and MV distribution network, this paper carries out the research ofcoordinated planning in MV distribution system to improve power supply capability, theresearch contents and results are as follows:(1) Based on the thorough analysis of the process of main transformer “N-1” checkand the mechanism of load transfer, this paper proposes the determination method ofpower supply capability considering constraints of transformer overloading and tie-linecapacity, in which some new conceptions are defined, such as the constraint matrix ofmain transformer contact capacity, the matrix of virtual contact relation. Through furtherinduction and summary, the optimization analysis model of the power supply capabilityis established, which converts the problem of the calculation of power supply capabilityto the problem of a complicated and large-scale linear optimization. Moreover, byanalyzing typical examples, this paper contrasts the characteristics of the analytic methodwith that of the model optimization method, providing effective tools for the subsequentresearch of the index evaluation of the power supply capability.(2) On the basis of the clustering analysis theory, this paper puts forward the powersupply block partition method based on the improved K-means clustering algorithm, inwhich the traditional K-means clustering algorithm are improved from the two aspects ofselecting of the initial clustering centers and determining the optimal clustering partition number, as well as defining the conception of the substation “equivalent radius” and“scope distance”, which makes the partition of power supply block change from the pointdemarcation into the area dividing. This method can cut a wide range of power supplyarea into small pieces in a reasonable way, where the substations at the same block havevery close ties with each other, and substations at different blocks may have relativelyweek contact with each other. In view of this characteristic, paying more attention to thestructure optimal planning of each power supply block will be the most effective andconvenient way to meet the distribution system’s lean plan requirements of the wholearea.(3) In order to improve the optimization efficiency by avoiding the useless searchcalculation of the substations which have no contact relationship between each other,based on the improved K-means clustering algorithm of power supply block partitionmethod, this paper proposes the method of two-stage programming based on recedingoptimization and the method of multi-objective optimization model of main transformercontact structure based on genetic algorithm. Among them, the two-stage optimizationand planning method consists of two parts, one of which is the establishment ofsubstation overall contact layout, and the other is the optimal setting method of contactcorridor between transformers, which completes the optimization settings of theconnecting passages between main transformer stations from coarse to fine and step bystep. Taking power supply capability, main transformer connecting passage number andtotal length of the connecting passage as objective functions, the paper comes up with anidea of combining genetic algorithm with sequencing strategy to solve themulti-objective optimization model. The two methods mentioned above are bothreasonable and economic way to set up connecting passage between main transformers,which can improve the maximum potential of the whole area’s power supply capability,and meet the requirements of the refined planning in regional power grid.(4) To break the traditional ideas of relying on wiring mode to guide theconstruction of the network, this paper builds the typical power supply model of the MVdistribution system. Firstly, the contact structure model of power supply block (powersupply architecture) is defined from the viewpoint of graph theory, and the matchinganalysis is carried out between relevant characteristic parameters of the model and allkinds of power supply area. Then, on the basis of optimization method of the maintransformer contact structure, the main transformer contact relations of the minimumcontact structural units whose connected degree is1~3, and related power supply capability is optimized and calculated. Finally, by introducing the typical connectionmode of MV network and combining HV substation with MV network, the paper buildsthe typical power supply model of MV distribution system based on the minimumcontact structure unit, and also calculates some related indicators such as the maximumpower supply capability, the feeder number of main transformer and the contact pointsbetween substations, in order to find some basic characteristics of the model. Relevantconclusions will provide important suggestions on the aspects of coordinated planning ofdistribution system, construction and retrofit of network, revision of the relevanttechnical guidelines of distribution network.