Probabilistic Load Flow Calculation And Characteristic Analysis For Microgrid

Microgrid contains different types of distributed generation, which has uncertainty and fluctuation. Compared with power system, it has different structural characteristics that the branch power flow is no longer single flow direction, leading steady state characteristics of microgrid more complex. Thus it is important to study microgrid load flow in order to provide reference for preliminary planning, operation and maintenance of system. This paper starts from deterministic power flow, analyzing the operating characteristic of microgrid system around probabilistic load flow.Firstly, the deterministic mathematical models are given based on wind power system, photovoltaic power system, fuel cell generation system and micro gas-turbine system in this paper. An advanced back and forward sweep algorithm with the branch current is proposed in order to improve the traditional back and forward sweep algorithm which can only handle PQ node. The variable of back process is transformed from the branch power into the branch current. By correcting reactive power equations, multi-type nodes are handled. Simulation results show that the improved algorithm can assure linear convergence and high convergent accuracy.According to the intermittence and fluctuation of distributed generation in the real application, different types of probabilistic models are established on microgrid. The probabilistic load flow algorithm on microgrid by combination of three-point estimate method and Newton-Raphson method is adopted in this paper. The proposed algorithm has no limitation on types of probability distribution. Arbitrary type of random variables can be treated by moment estimation calculation. Microgrid system simulation results show that the proposed algorithm can well effectively and accurately reflect steady state characteristics of system and posse high practical value.The above algorithm requires mutually independent random variables. Actually, the wind speed in wind farms and the light intensity on near geographical location have correlation. Therefore, the probabilistic load flow algorithm considering the correlation of input random variables is proposed. This algorithm uses third-order polynomial normal transformation to deal with dependent random variables. And then it combines three-point estimate method to format three sampling points. Finally the sampling points of output random variables can be solved by using load flow equations. The probabilistic density curve and the probabilistic distribution curve can also be got based on Gram-Charlier expansion series. Simulation results indicate that the proposed algorithm is corrective and effective, reflecting overall the steady state characteristics of microgrid taking account into correlation. The transformation method remedies the shortage of three-point estimate method and enlarges application range.For further research on operation characteristics of the microgrid system, the example of microgrid system concluding different types of distributed generation is given to analyze dynamic characteristics. The operating condition is of two sorts: islanding mode and grid-connected mode. The active power change of distributed generation and the variation of interruptible load are studied on microgrid system. Simulation results show that the microgrid can sustain the stability of voltage and frequency. In conclude, the system runs with high reliability and safety.