Research On Short-term Interval Optimal Dispatching Model And Method For Hydro-thermal-wind-solar Power System

With the gradual depletion of fossil energy and the gradual deterioration of the environment,the development and utilization of clean energy such as water,wind,and solar energy for power generation have been widely used.To better deal with the uncertainty of wind power and photoelectric power,give full play to the regulation capacity of cascade hydropower and thermal power,and carry out the joint optimal dispatching of multiple energy sources,has become an urgent problem to be solved in the current power grid dispatching.To alleviate this problem,this thesis starts from reducing the uncertainty of wind and solar power output and applying it to dispatch,researches and explores the prediction interval of the wind power and the photovoltaic power and the multi-energy dispatching of the hydro-thermalwind-solar power.The main contributions of the research can be summarized as the following:(1)To solve the problem that there are always errors in the prediction of wind power and solar power,this thesis makes an interval prediction of wind power.First,the long-term and short-term memory cycle neural network model(LSTM)is used to predict the wind power and solar power,and the gray wolf algorithm(GWO)is used to optimize the parameters of the LSTM model,to improve the prediction accuracy of wind power and solar power.Secondly,based on the prediction of wind power and solar power,the thesis classifies the prediction value and relative error of wind power and solar power by clustering technology and fits the prediction error of the classified prediction value of wind Power Point,to get the prediction range of wind power.Finally,the simulation results show that the LSTM-GWO interval prediction model can obtain high-quality short-term wind power and solar power prediction intervals.(2)Aiming at the generation dispatching problem with wind photovoltaic output uncertainty,a short-term complementary interval dispatching model(HTWSID)for hydrothermal wind solar multi-energy system is constructed.In this model,the uncertain wind and solar power output are expressed in the form of the interval,and the optimal adjustment of hydrothermal power output and wind and solar power output interval is carried out with the goal of power system operation cost,wind-solar energy abandoned minimum and compensation cost of insufficient wind power generation.In this thesis,the interval constraints in the model are transformed by the interval possibility theory,so that the interval optimization model can be transformed into a random variable model.A high-precision special ordered set linear approximation method based on the second-order discretization is proposed to deal with the non-linear functions such as the cost function of thermal power in the transformed model so that the non-linear model can be transformed into a mixed-integer linear model and solved directly by the solver.The simulation results show that the method proposed in this thesis can improve the utilization ratio of the uncertain output range of wind and power,at the same time,improve the economy of power system operation by using the characteristics of hydrothermal wind multi-energy complementary;compared with the non-linear programming method,it has an advantage in calculation accuracy and time.(3)Furthermore,considering that the prediction range of wind power can only reduce the error but can not eliminate the deviation from the actual output,a calculation model of the conditional risk value range of wind power uncertainty output range is proposed.On this basis,the short-term interval dispatching model(CVa R-HTWSID)of the hydrothermal wind power multi-energy system considering the operation risk of wind power is established.Considering the distribution characteristics of wind power output,the probability interval probability is used to transform the interval optimization model into a random variable model.For the transformed model,an improved gray wolf algorithm based on a heuristic constraint processing strategy is proposed.In this algorithm,the decision variables are heuristically adjusted by the degree of individual violation of constraints,and then the feasible solution individuals are obtained.Finally,the simulation results show that the proposed algorithm can get a group of dispatching schemes with lower comprehensive cost compared with other algorithms,which provides strong technical support for decision-makers.(4)The multi-objective generation dispatching problem with an uncertain output range of wind and solar power is discussed.Because of this problem,considering the correlation of the output of multi wind(solar)electric field,a short-term multi-objective interval dispatching model(CVa R-MOHTWSID)of the hydrothermal multi-energy system is established,which comprehensively considers the operation risk of wind power,the generation cost of the power system and the emission of pollutants.In the process of solving the model,to contain more uncertain information of wind power,a method of scene construction and reduction of wind power output range is proposed.Based on the probability sampling method,this method uses copula function and K-means clustering technology to construct the scene of joint wind power generation and uses the backward reduction method based on the area spacing to reduce the scene,to improve the efficiency of the model solution.Besides,based on the multi-objective Gray Wolf algorithm and the interval dominance relationship,and improved multi-objective interval gray wolf algorithm(MOIGWO)is proposed to solve the CVa R-MOHTWSID model.Finally,the simulation results show that the MOIGWO algorithm can obtain high-quality uniform and diverse interval solution sets when solving the CVa R-MOHTWSID model,which provides strong theoretical support for dispatching decision-makers.