Research On Load Forecasting And Profit Strategy Optimization Of Retail Electricity Companies Under The Biased Electricity Assessment Mechanism

Since 2015,the power market has been undergoing a new round of reforms,The idea of which is to make the government tightens regulation on the power grid,transmission and distribution links,increases the capacity of power generation and electricity sales,as well as improves the degree of marketization of electricity suppliers,vendors and users.This reform has played a vital role in accelerating the development of the power market from the perspective of electricity sales,thus leading to the emergence of various electricity vendors.The rise of electricity vendors has contributed to improving the structure of the entire electricity market while diversifying electricity transactions.As an intermediate between electricity suppliers and electricity users,electricity vendors make the purchase of electricity from the wholesale electricity market for resale to the retail market,with profits generated out of the price difference or the provision of value-added services.However,electricity vendors are also faced with various operational risks while gaining profits.For example,there would be a deviation between electricity sales and purchases at the time of transactions,which will make electricity vendors face deviation power assessments.As a significant influencing factor for the revenues generated for electricity vendors,deviation power assessment plays a vital role in ensuring fair transactions in the electricity market.Therefore,electricity vendors were taken as the research object in this study to optimize the load forecasting and sales profit strategy intended for electricity vendors under the current mechanism of deviation power assessment.The research results obtained from this study are detailed as follows:First of all,a summary was made of the business risks faced by electricity retailers.On this basis,an analysis was conducted regarding the research significance for electricity vendors to avert biased electricity assessment,with the current status of research on related issues both at home and abroad sorted out.Through discussion about the load forecasting and deviation power assessment response strategies intended for electricity vendors from two perspectives,this study lays a theoretical foundation for further research and indicates its direction.Then,an improved adaptive genetic algorithm model was constructed in this study by analyzing the problem with the traveling salesman(TSP)optimal path planning,and power load forecasting research was carried out by optimizing the BP neural network.Herein,such factors as season,climate,temperature,precipitation and others that have a significant impact on load changes were treated as variables.Through a comparison in load forecasting results between the improved model and the traditional model,it can be discovered that the improved algorithm can achieve a faster convergence rate than the traditional algorithm,thus providing a more accurate reference method for electricity vendors to carry out load forecasting.Secondly,based on a research of the electricity sales model adopted by electricity vendors,a weekly warning model of deviation power assessment was proposed in this paper for the frequent operation or slow response caused by the daily or monthly warning of electricity vendors.Meanwhile,a strategy model was constructed for the assessment and avoidance of positive and negative deviations under a fixed price difference contract,with the maximum revenues generated for electricity vendors as the objective function,based on which the impact of various parameters on the revenues generated by the electricity vendor was compared.According to the results,the strategy of applying interruptible load,increasing controllable load,and carrying out mutual protection of electricity is effective in improving revenues for electricity vendors to a certain extent.Finally,based on the theory that the electricity vendor and the user sign up to a spread option contract,a Nash equilibrium model for the revenue of both the electricity vendor and the user was established in this study,with an improved particle swarm algorithm adopted to solve the model.The results show that the overall economic benefits received by both parties can be improved by reducing the positive deviation power assessment fee under certain conditions,so as to achieve a win-win situation for both the electricity vendor and the user.In general,the focus of this study is placed on the relevant research on the profit strategy adopted by the electricity vendor.Through the establishment of an improved load forecasting model,a deviation power weekly warning model,a deviation power optimization model,and a Nash equilibrium model for the revenues of two parties under a spread option contract,a comparison was performed before and after the transaction to support the electricity vendor in reducing the deviation and increasing revenues,which is of practical significance to improving market competitiveness for electricity vendors.