Research On Power Balance And Reactive Power Allocation In PV/Wind/Battey Islanded Microgrid

Renewable energy power generation technology has developed rapidly in recent years,of which wind/PV power generation is the most widely used renewable energy because of its clean,efficient and sustainable characteristics.Distributed microgrid is the ideal organization form and large-scale application mode of wind/PV energy generation.The power balance and system stability of islanded microgrid is more difficult than grid-connected microgrid,and it is one of the key difficulties to restrict its application.This thesis studies the maximum power control strategy of distributed generation(DG),the active power balance and system configuration of islanded microgrid,and the reactive power balance and optimal allocation of off grid wind/PV microgrid.The main research work and innovative achievements include:First,the output power maximization strategy of wind/PV DG is improved and optimized.A PV power maximization control method based on fuzzy sliding mode control and voltage compensation power prediction disturbance observation method is proposed.Rapid and stable maximum power tracking is realized by using sliding mode control and robust tracking,combined with fuzzy control to reduce chattering.The estimated voltage deviation is used to control the voltage compensation,and the disturbance observation method based on power prediction is proposed to reduce the oscillation and misjudgment.The wind power generation model has strong nonlinear characteristics,and the input wind speed is unstable.To solve this problem,a sliding mode self-optimizing control method(SMOSC)and an optimal control method based on feedback linearization are proposed based on the nonlinear model of wind power generation.SMOSC has the advantages of simple implementation and high capture efficiency,without measuring wind speed.Feedback linearization can make complex nonlinear model equivalent to linear system,and improve the tracking characteristics.Second,a progressive Droop power control method for preventing overrun,and a system capacity allocation method based on artificial immune algorithm are proposed.Considering the typical application environment of island,DG source,energy storage and daily load,desalination,EV models,and integrated model of PV/wind/Battey islanded microgrid are established.The active power dispatching stategy is designed according to the environmental protection and the peak valley electricity incentive policy.Aiming at the problem that the droop control is easy to overrun and the control speed is not adjustable,a progressive Droop control method is proposed.Small signal analysis shows that the method has good stability.For the problem of capacity allocation,a multi-objective optimization allocation method based on artificial immune is proposed,which takes environmental protection and unit energy cost as objective function.The configuration optimization in zero emission,and the power balance between renewable energy and load power and discussed.Simulation results verifies the effectiveness of the control strategy.The islanded microgrid has a good capacity to EV in accordance with zero emission configuration.Third,based on power flow calculation,droop compensation control of weak nodes and reactive power allocation optimization method of equal network loss rate and Semi Invariant Method are proposed.The reactive voltage characteristics of droop control micro source are theoretically analyzed.The calculation of power flow is extended by homotopy analysis method,and the weak node is searched through the calculation of reactive margin,and the droop compensation control of reactive power supply is carried out.Two cases are discussed aiming at the reactive power allocation problem.Optimal distribution calculation of reactive power sources based on the equal network loss rate is presented for fixed parameter network.Optimization allocation on the minimum network loss probability density function using semi invariant method is proposed for random parameter network.The design steps and numerical examples of the two cases show the effectiveness of the method.