Research On Economic Dispatch Of Large-Scale Wind Power Integrated System With Consideration Of Adequacy Risk

Nowadays, China’s wind power industry enjoys a rapid development under the policy incentives. As the strong uncertainty of the output of wind power, the integration of large-scale wind power increases the difficulty and risk of operation of power system. In this context, the research of economic dispatch of large-scale wind power integrated system with consideration of adequacy risk has theoretical and practical significance.In order to solve the problem of adequacy risk caused by uncertainty of wind power,, the paper does the research in four aspects from adequacy risk management perspective:(1)To reduce the uncertainty of risk source, further research of wind speed and wind power prediction method is carried out to improve forecast precision.(2) The research of the probability model of wind power prediction error is important for risk assessment and enconomic dispatch.(3) Adequacy risk assessment method is proposed to evaluate the short-term operation extreme risk of power system.(4) The economic dispatch model of power system with consideration of adequacy risk and risk decision method is proposed to make reasonable risk decision and determine the economic dispatch scheme.The paper studies the wind speed and power forecast method to decrease the predition error. The new improved BP neural network based on Adaboost algorithm (Ada_BP) is proposed to forecast short-term wind speed. Using Adaboost algorithm can solve defects of the BP neural network algorithm such as over-fitting and trapping in local optimum. The case study shows that the prediction accuracy of the proposed algorithm is higher than time series algorithm and BP neural network algorithm. The results demonstrate the accuracy and effectiveness of the proposed algorithm. For wind power prediction problem, wind power curve modeling method based on BP neural network is proposed. The direct prediction method of analysis of wind power and forecasting method based on wind power curve are studied. Through the case study, the prediction accuracy and application conditions are given.The paper studies the probability distribution model of wind power prediction error. The Accuracy and applicability of normal distribution model in describing the total wind power prediction error of power system is verified. Moreover, the time-varying mixed distribution model based on normal distribution and Laplace distribution is proposed to describe the wind power prediction error. Through the analysis of the characteristics of actual prediction error probability distribution and wind power curve, time-varying characteristics of actual prediction error probability distribution can be found and explained. The accuracy and applicability of the proposed model in describing the prediction error are demonstrated by optimization problem of the energy storage capacity for wind farm.Adequacy risk assessment index based on CVaR and method of power system in short-term operation are proposed. Probability distribution model of uncertainty factors power system are established. Conditional Value at Risk (CVaR) is applied to measure the adequacy risk, and two adequacy risk index are proposed such as risk indexes of load loss and wind curtailment based on CVaR. The case study shows that the proposed risk indexes can reflect extreme risk of power system better than loss of loss of load probability risk index and expected loss of lost load risk index and demonstrates the comprehensiveness and accuracy of proposed risk index. Moreover, the proposed risk assessment method can evaluate the risk of power system under different adopt capacities for wind power and spinning reserve levels, and help the solution of the problem of power system economic dispatch with consideration of adequacy risk.The paper studies problem of economic dispatch of large-scale wind power integrated system with consideration of adequacy risk and risk decision method. The cost-CVaR multi-objective optimization model based on portfolio theory and adaptive chaos multi objective particle swarm optimization algorithm (ACMOPSO) are proposed. In order to select the optimal solution from Pareto solution set, the risk decision method based on marginal risk is presented, which can help dispatchers determine dispatch scheme to balance the reliability and economy according to the risk preference. The case study of IEEE-118power system demonstrates the accuracy and effectiveness of the proposed ACMOPSO algorithm, cost-CVaR model and risk decision mehtod...