| Distributed generation (DG) has won a rapid development for its flexible installation, less investment, cleanliness and efficiency as well as its ability to improve the function of traditional centralized grid. Accompanied with the ever increasing generation capacity, the original radial distribution networks have gradually evolved into an interconnected network of users and generations in which load flow is no longer unidirectional from substation bus lines to the load. Unavoidably, the change of grid structure and direction of load flow would have an impact on the operational characteristics of the distribution network and that of the composite load. As a result, the former composite model may fail to offer a suitable depiction of regional load characteristics. Under Such Background, relying on the National Natural Science Foundation project “Critical Theory and Technology of Power System Generalized Composite Load Modeling Considering Distributed Generation”, this paper studies the composite load characteristics of distribution network with DG and its model structure in an effort to provide model to support power system simulation.This paper has carried out in-depth research on the operational characteristics of typical DG, proposed equivalent description models for DGs such as PV, fuel cell, micro-gas turbine generator, etc. With analysis of the similarities among structure and equivalent characteristics of different inverter-based grid-connecting DGs, a unified model for this kind of DG is put forward. This paper also studies the impact of inverter-based DG’s penetration levels on descriptive ability of traditional mechanism model, and proposes a generalized composite load model structure that can represent the distribution network’s composite load characteristics with high penetration of inverter-based DGs. After thoroughly studying the output characteristics of doubly-fed wind generation system, a modeling-oriented equivalent description model of wind generation system is constructed. Then, based on numerical and trend similarity, an assessment method for load model accuracy centered on numerical processing and weight distribution is put forward. Detailed works are as follow:In the first place, equivalent models for typical DGs including PV, fuel cell, micro-gas turbine generator, direct-drive wind generator are constructed. Based on systematic research on DG’s generating principles and operational characteristics, using Matlab/Simulink, specific digital simulation models for typical DGs are established. By studying their dynamic behaviors in transient process, controlled current source model of PV array, voltage source model of fuel cell, controlled current source model of micro-gas turbine generator and its rectifier as well as that of direct-drive wind generator and its rectifier are set up. These models are then combined with general control strategy to construct their external characteristic equivalent models. With simulation and analysis of those models under different fault degrees, their effectiveness is testified.Secondly, a unified model for inverter-based DGs is proposed. From the connection mode between DGs and the grid, it defines the DG which feeds its power through inverter into the grid as inverter-based DG This category includes generation systems of PV, fuel cell, micro-gas turbine generator and direct-drive wind generator. Through analyzing the similarities of inverter-based DG on topological structure and control characteristics, and considering their operational characteristics, a unified model for inverter-based DG has come into being with complete mathematical analysis and identification strategy of model parameters. The unified model’s adaptability to depict the external characteristics of different inverter-based DGs is verified by running numerous simulations under different fault degrees.Then, the generalized composite load model of distribution network considering the impact of inverter-based DG is constructed. After systematic analysis of the impact of inverter-based DG’s penetration on the traditional mechanism model’s descriptive ability over regional composite load characteristics of the distribution network, it is pointed out that the traditional one is incapable to offer a well depiction of composite load characteristics of the distribution network under high penetration of inverter-based DG. Therefore, in a pursuit to conquer this problem, a generalized composite load model of distribution network considering the impact of inverter-based DG is proposed on the basis of the unified model of those DGs. The validity of the proposed model is testified under different fault degrees as well as varied capacity configuration ratios of different inverter-based DGs.Also, an equivalent description model of load modeling-oriented doubly-fed wind generator is built. Through systematic analysis of steady-state operational characteristics of doubly-fed wind generator (DFIG) as well as its dynamic behavior in transient state, its corresponding equivalent description model is constructed based on three-order induction motor model—an induction motor in serial connection with controlled power source. Moreover, an equivalent description model of DFIG wind generation system is proposed, that is, an internal feedback expressed by a first-order difference equation is introduced into the output equation of the model, to solve the problem of inadequate approximation capability resulted from order reduction of the model and strengthen its depiction ability of transient state.Finally, a model accuracy assessment method is proposed based on trend and numerical similarity. The data sequence is divided at the beginning according to its development. Their corresponding weight is distributed according to their importance. By doing so, the negative impact of strong relation between assessment results and sequence length occurred in regular numerical assessment is effectively curbed. Numerical similarity assessment method bases its relative deviation sequence of data sequence on steady-state value, thus avoiding the mistakes of negative numerical similarity due to large gap between corresponding elements of the sequence. Weight of relative deviation sequence’s elements is set in line with analytic hierarchy process. Trend similarity has adopted gauge transformation method to reduce the negative impact of sequence amplitude, deviation as well as noise on trend similarity assessment. By test carried out on examples based on measured and simulation data, the new assessment method is proved to be effective. |
PDF Full Text Request