Short-Term Load Forecasting-Based Distribution State Estimation Including Ditributed Generation

With the rapid development of smart grids and distributed generation (Distributed Generations,DG),the real-time monitoring for distribution network becomes more and more important.As the basic core of the Distribution Management System(DMS),Distribution State Estimation (DSE) can provide a large amount of reliable and accurate data support for many of its important applications.However,the lack of real-time measurements in the distribution network, resulting in the unobservability of distribution system,which requires a lot of pseudo-measurements to ensure the observability of the distribution system and further to promote the DSE to execute smoothly.Furthermore,since the complex structure-. three-phase unbalanced and the increasing proportion of DG in distribution network,the traditionally power transmission network state estimation methods (such as Weighted Least Squares Method) could not be directly used for DSE.So,this paper proposes a short-term load forecasting-based DSE method including DG.The main work and results of this thesis are summarized as follows:1、Firstly,This thesis used the Particle Swarm Optimization-Least Squares Support Vector Machine(PSO-LSSVM) to realize the short-term load forecasting. That is,using PSO to optimize the key parameters of LSSVM,so as to establish a more accurate short-term load forecasting model which provide more accurate pseudo-measurements and further to solve the observability issue of distribution system and ensure the DSE to execute smoothly.2、Considering the increasing uncertainty of the system state and load variation caused by the randomness and the variability of the increasing proportion of DG in power system as well as the nonlinear characteristics of a few equipments in the distribution network such as voltage regulator and distributed generators, the DSE including DG is considered as a nonlinear optimization problem with multi constraint conditions, which is solved by applying a self-adaptive differential evolution algorithm with excellent global search capability and good convergence property.3、For the proposed method,this paper simulates on the three-phase balanced IEEE33-bus test system with DG and the three-phase unbalanced IEEE13-bus test system with DG respectively. The following tests were done:(1) the influence of short-term load forecasting accuracy on DSE; (2) the influence of different location and capacity of various types of DG accessing to system on DSE. (3) Comparison of different algorithms for optimization of DSE with DG. The simulation results show the feasibility of the presented method and the comparative results with other optimization algorithms verify the effectiveness of the introduced method.