Security Constrained Economic Dispatch Of Power System With Wind And Photovoltaic Power

Nowadays,with the deterioration of environment and depletion of fossil energy,wind power and photovoltaic power have received extensive attention and utilization world-widely,as the most well developed renewable energy.However,as wind and photovoltaic power penetration in power system increase significantly,the stochastic nature of wind and photovoltaic power have caused great effects on the security of power system operations,and bring new challenges to the power system security constrained economic dispatch(SCED).In order to increase the forecast accuracy of wind and photovoltaic power,this paper proposes a short-term wind speed and solar irradiation forecast model based on nonlinear autoregressive dynamic neural network.In network training procession,in order to improve the accuracy of inter parameters fitting,this paper proposes a kind of training sample in parallel structure to guarantee the time correlation within training sample.Secondly,focusing on traditional solar forecast model's sensitivity to weather conditions,this paper adopts multi-dimensional input combination to improve forecast model's adaptability to weather condition.The comparisons with traditional solar irradiation forecast model based on focus time delay network in four typical weather conditions prove the improvements of this paper's forecast model in solar irradiation forecast accuracy and weather adaptability.Focus on the influence of wind power and photovoltaic power on system operation security and economy,This paper proposes a security constrained economic dispatch strategy that considers optimizing system state selection and spinning reserve allocation.The system state selection is judged by the criterion of occurrence probability.Considering the impacts of wind and photovoltaic power integration on the outage probability of system component,the over-current protection outage model and voltage protection outage model are employed to calculate the operational outage probabilities of transmission lines and generators respectively.The fast sorting technique is adopted to select all system states with higher occurrence probability than the system security threshold.Besides the traditional constraints on system demand and regulation range,the spinning reserve allocation additionally obeys the transmission constraints in the selected states to avoid transmission congestion in contingency situations.Benders decomposition is utilized to partition the SCED problem into the active power dispatch sub-problem and the optimal reserve allocation sub-problem for tractability.A numerical comparison with the traditional N-1 security assessment SCED is given to validate security and economic improvement of the proposed method.