Based On Wide Area Measurement System Load Modeling Study

In recent research on modern power systems, digital simulation has become the primary scientific method of decision making, program, dispatch and system operation. Load model is one of the crucial factors that influence the accuracy and reliability of the simulation results. Presently, load modeling is mostly restricted to load nodes where the power measurement equipment is available. When the number of load nodes in concern becomes large, problems come forth, such as the exorbitant cost and the difficulties in implement. This paper presents a load modeling method based on wide-area measurement system (WAMS), considering the features of WAMS. This method can be adopted to obtain load models of all the load nodes, including the ones on which no measurement equipment is available. Generally, research work in this paper covers the following aspects.Firstly, the actualities and the significance of load modeling in power systems is elaborated, as well as the concept, methods, common models and the parameter identification methods of load modeling. An idea is also brought up, that we can obtain load models based on measured data of the phasor measurement unit (PMU).Secondly, this paper introduces the features and implement of the load modeling method based on WAMS, and analyze the advantages of this method and its essential differences from the traditional ones. A composite load model is established and the clonal selection algorithm (CSA) is adopted to identify the parameters. This method is applied to the IEEE 39-bus system and the simulation results show that the response from the load model well fits the measured response, which is a proof that the proposed load modeling method and the CSA based parameter identification are feasible and valid.In addition, this paper discusses the classification and synthesis of load characteristics under WAMS based load modeling method, so as to solve the load's time variation problem. This method uses the model response gained by a standard voltage excitation as the characteristic vector and uses the CSA based clustering algorithm to research load classification. The direct synthesis identification is then applied to various kinds of load. Finally, the correctness and validity of the proposed method is verified through the load modeling research on two of the load nodes in a 3-generator-9-bus system, as well as the simulation results of its classification and synthesis.