Probabilistic Assessment And Optimization Strategies For Voltage Security Of Droop-regulated Islanded Microgrids

With the constantly growing application of intermittent renewable distributed generations(DGs),microgrid has been proposed and intensively studied in recent years.For microgrids operating in islanded mode,dispatchable DGs are mainly responsible for the system frequency and voltage amplitude regulation,and two control schemes have been proposed:centralized control scheme and decentralized droop control scheme.Decentralized droop control scheme prevails at its lower-cost,reliability,modularity and expansibility.However,due to its special structure and operating mechanism,droop-regulated islanded microgrids(IMGs)is vulnerable to voltage quality and voltage instability issues especially in conditions of high penetration of intermittent DGs.In this thesis,the probabilistic voltage profile and stability evaluation and optimal strategies for voltage security improvement of droop-regulated IMGs are studied on the basis of steady state modeling.Details are as follows:1.Traditional load flow algrithom is not suitable for droop-regulated IMGs because there is no slack bus and the frequency is not const.Therefore in this thesis we propose a load flow method for both balanced and unbalanced droop-regulated IMGs.Taking voltage of each node and frequency as variables,the load flow model is built considering the philosophy of droop-controlled DGs and static voltage/frequency characteristics of loads.Newton-Raphson method is proved to be difficult to converge due to the characteristic of droop control principle,especially for unbalanced IMGs.A Newton-Raphson load flow method with optimal multipliers for droop-regulated IMGs is proposed to solve the flow equations with better convergence.By optimal multiplier deduced from higher order term of Taylor expansions of the load flow equations in each iteration,the iterative correction direction is guaranteed to be closest to the solution,and the convergence performance is improved.Numerical examples demonstrate the correctness and effectiveness of the proposed algorithm.2.Owing to the small scale,IMGs are vulnerable to voltage quality issues caused by the integration of intermittent DGs.A probabilistic method to evaluate the voltage profile of droop-regulated IMGs considering the correlation of random variables is proposed.To describe the uncertainties in output power of intermittent DG and loads,quasi-Monte Carlo method with Sobol sequence is adopted.Usually there is significant correlation between the output power of intermittent DGs in an IMG.To preserve the correlation in samples,Nataf transformation is performed.Through repetitive load flow calculation of droop-regulated IMGs by Newton-Raphson method with optimal multipliers for the samples,voltage quality metrics of mean value,standard deviation,limit violation probability,probability density function and three-phase unbalance factor are obtained.The effects of attributes that may influence voltage profiles such as droop parameters,penetration level and power factor of intermittent DG on voltage quality are discussed.The proposed method can help to evaluate and improve the maximum hosting capacity of intermittent DGs without violation in an IMG.Test results demonstrate the performance of the proposed method.3.IMG is vulnerable to instability owing to its small scale and low inertia.To analysis the static voltage stability of droop-regulated IMGs,two methods based on singular value decomposition and nonlinear programming respectively are adopted.On the basis of load flow calculation by Newton method with optimal multiplier,the minimum singular value can be obtained as a static voltage stability index by the singular value decomposition of Jacobian matrix.Meanwhile the load margin index is obtained by nonlinear programming method.Taking the randomness of output power of intermittent DGs and load fluctuation into account,a probabilistic evaluation method of voltage stability based on unscented transformation technique is proposed.To obtain the probabilistic characteristics of load margin,unscented transformation technique is utilized to transform stochastic problem to deterministic problem,which is solved by interior point method of the IPOPT toolbox.The results of the proposed method are verified by those of Monte Carlo method with Nataf transformation on an IMG system.Moreover,the adaptability of the proposed method to the correlation,the probability distribution types and parameters of random variables is testified.4.Based on the above analysis of voltage quality and static voltage stability of droop-regulated IMGs and the influence factors,the optimal settings of droop parameters and the optimal reactive power planning are studied as means to improve the voltage quality and static voltage stability in island state.First,a droop parameter settings optimization model for droop-regulated IMGs is established,which takes voltage deviation as the objective function and static voltage stability margin as one of the constraints.Then an optimal reactive power planning model of distribution system,considering the voltage quality in island state is set up,with droop parameter settings optimization considered.The locations of compensation capacitor banks are selected based on the singular value decomposition of Jacobian matrix in island state;and then the voltage deviation in island state and the economic benefits in the grid-connected state are taken as objective functions.The reactive power regulation capability of intermittent DGs is considered in both models,and Quasi Monte Carlo method and scenario reduction technology are applied to get typical scenarios to describe the randomness of intermittent DGs.Test results verify the effectiveness of the optimization strategies for voltage quality and stability improvement.