Research On Storage-operation Strategy In Renewable Energy Grid

Renewable energy grid is the grid with high proportion of renewable energy.Up to now,renewable energy grid has been born and the number is increasing world-widely.Renewable energy grid represents the inevitable trend of power grid development.Power generations of some kinds of renewable energy power,such as wind power and solar power,etc.have the inherent characteristics of uncertainty,which can be characterized by randomness and intermittency.As a result,it is difficult to accurately predict these kinds of renewable energy power.The fundamental cause of power generation uncertainty is the non-storability of primary energy.Energy storage facilities are effective means to eliminate or reduce the negative influence of uncertain power generation on power balance.Pumped storage is currently the only practical high-capacity energy storage technology.Aimed at the large-scale wind power generation,this paper takes the maximization of renewable energy utilization as the basic goal and pumped storage as the basic means,and adopts the progressive connection of day-ahead dispatching,real-time scheduling and automatic power generation control as the basic route to study the storage-operation mode and related technologies for solving power balance problems in renewable energy grids.In order to effectively increase the permeability of the generation of renewable energy power,a renewable energy grid storage-operation mode is built.Power balance in renewable energy grids requires not only the instantaneous power dissipation technology,but also the continuous energy storage technology.A threestep power balance model combining day-ahead power generation plan,real-time power generation plan and automatic power generation control is established.Based on the uncertain power generation prediction and load prediction,the power generation in day-ahead dispatch scheduling is decomposed into the base load,the predictable volatility reserve,the unpredictable uncertain reserve and the atomatic power generation reserve.In view of the fact that the capacity of the large-scale energy storage is limited,predictable fluctuating power is mainly balanced by conventional power,while the uncertain power is balanced by energy storage facilities,and the residual imbalance power is borne by the automatic power generation device;when the capacity of energy storage facilities is limited,conventional power will serve to make up the shortage.Storage-operation mode is the main academic thread of the whole paper.A novel wind speed and wind power forecasting method based on the wavelet packet decomposition theory and improved Elman neural network is presented.Aiming at the problem of poor regularity of wind speed variation,wavelet packet decomposition theory is adopted to decompose wind speed data into several wavelet spaces,the wind power forecasting method based on the Elman neural network with different frequency network is hence formatted to abtain the forecasting results.A PSO algorithm with disturbance is used to train Elman neural network to avoid the optimal particle falling into local optimal solution.The day-ahead and real-time wind speed and wind power forecasting results show that the prediction accuracy is higher than that of traditional methods,such as the traditional Elman neural network,etc.Considering the power error and energy deviation in wind power forecasting,a method of capacity sizing of energy storage system based on wind power reliable output power is proposed.Based on the plan of the uncertain renewable energy power,non-parametric estimation method is adopted to fit the probability distribution of the certain wind power,moreover,the probability distribution of the uncertain wind energy concerning to energy storage dispatching cycle is abtained.Considering the wind curtailment loss and traditional energy power uncertain reserve,a model is formulated to determine the relationship between the power capacity and wind energy loss.Non-parametric kernel density estimation is adopted to estimate the confidence intervals of wind power prediction error and fluctuation ranges of the actual output of a wind farm under different confidence degrees.The relationship between energy storage system capacity and the reliable output power of the wind farm is studied,and according to the optimal cost-benefit principle,costs and benefits of energy storage system are analyzed from the perspective of reliable power output,and the corresponding energy storage system capacity sizing method is formed.A typical day example demonstrates the rationality of the proposed method.Based on the large-capacity energy storage,the reliable output power of the wind farm ensures the feasibility of day-ahead generation plans.Finally,taking the pumped storage as the large-capacity energy storage system,the power balance models of day-ahead dispatching and real-time scheduling are put forward in view of renewable energy grid intergrated with the high proportion of wind power,and the storage-dispatching strategy of renewable energy grid is proposed.An evaluation index of power stability of the power grid is also presented.The studies show that without increasing operation cost of thermal power unit,the abandoned wind power decreases with the increasing pumped storage capacity,until the wind curtailment disapears.Arming at the fact that the capacity of the pumped storage is limited,three dispatching strategies are prepared,the pumped storage respectively setting as the simple uncertain reserve,the priority volatility reserve and the priority uncertain reserve.From the perspective of maximizing the utilization of the renewable energy,the strategy with pumped storage as the priority uncertain reserve has the best dispatching effect,and the effect of the strategy with the pumped storage as simple uncertain reserve takes the second place.The studies of a theoretical simple system case and a practical Zhangjiakou power grid case demonstrate the effectiveness of storage-dispatching strategy.