Elaborate Forecasting Theory And Methods For Power Systems

All decisions are based on prediction.The prediction provides a key support for power system planning,scheduling,controlling and so on,which is the basis for the balance of the power system.Faced with the increasing global fossil energy depletion,greenhouse effect,environmental pollution and other issues,the structure and operation of future power system will be greatly changed.The change in structure of future power system is reflected in substantial increasement of new energy and power electronics.Integration of new energy sources such as wind,photovoltaic,biomass and other forms of power into the power grid will greatly increase the uncertainty of the power system and put forward higher requirements of the breadth and precision of related prediction.The power electronics of power system will make the traditional power grid lose the foundation that keeps the instantaneous balance by the mechanism inertia,which also puts forward higher demands on the prediction breadth and the accuracy of prediction.It is necessary to improve the accuracy and content of prediction to fully support the operation decision-making of future power system.In view of the above problems,this paper carried out the following research:The influence of temperature on the load has a cumulative effect,and it leads to a big difference on the load even the weather conditions and the day type of the two days are the same.The current similarity date selection methods which take weather conditions and the day type as the characteristic parameters cannot guarantee the accuracy of prediction.A dynamic similar method is proposed in this paper,and a new short-term load forecasting method is proposed by combining it with the existing methods.The load forecast of holiday is transformed the average load and the load curve prediction respectively by using decoupling model.The average load is predicted by the dynamic similar method and the static similar method is used for the load curve prediction.And the load forecasting result is the integration of the load curve forecasting result and the load reference value.Finally,the case study shows the advantages of this method in improving the load forecasting accuracy of holidays.Aiming at the time variation characteristic of load mode,a load forecasting model based on artificial neural networks has been proposed in the third section of the paper.Load model is the basis of power system operation and control simulation,accuracy of which has great impact on power grid security and stability.At present,the historical load of the same period is used as the future load model in the power system operation mode verification,which will result in simulation bias.In this paper,the history load model parameter is achieved by using the least square method at first.Second,based on the static ZIP load model and the artificial neural model,the predict day of load model parameters have been reached by referencing the load forecasting method.And then analyze the sensitivity of forecasting results.The load model prediction can be used to check the operation mode of power system.Besides,an orderly power utility based on the load model is proposed in this section.It can improve the stability and economic of power system operation.Aiming at the characteristics of various kinds of devices and complex influencing factors in fault probability forecasting,a random fault probability forecasting method based on ARMA(Auto-Regressive and Moving Average Model)model and a fuzzy fault probability forecasting method based on non-interval gray prediction have been proposed in this paper.In addition to the main factors of equipment fault rate that have been researched,the equipment historical data and micro-meteorology are taken into the fault forecasting model.For the typical power equipment,transmission lines and transformers,the fault probability forecasting models are established respectively.The data of transmission line is analyzed with SPSS(Statistical Product and Service Solutions).The ARMA method is used to identify the model and forecast the fault probability of the transmission line.In the other hand,considering the fault characteristics and data monitoring situation of power transformer,an unequal interval gray GM(1,1)power prediction model of gas solubility of different gases are used to predict the IEC three-ratio.Results of case studies show that the forecasting results can improve the fitting and prediction accuracy,and the model has no limit to the equidistant coherence of the basic data,with better practicality and adaptability.The change of structure and operation of power system in the future will greatly increase the uncertainty of the system.On the one hand,this paper is trying to improve the accuracy of the load forecasting;on the other hand,two forecasting items are proposed to provide decision-support for the future power system operation.