Multi-time Scale Operation Optimization Of Integrated Energy System Based On Energy Prediction

The Integrated energy system with multi-energy coupling plays an important role in realizing energy interconnection,promoting the reform of energy structure,and realizing the whole life cycle management of energy consumption.It has become a research hotspot in the energy industry.The comprehensive energy system intersects the traditional energy system with diverse equipment and complex energy forms,and there are fluctuations on both sides of source-load.In addition.in order to reduce the consumption of primary fuel,renewable energy is introduced into the energy supply side of the comprehensive energy system,which on the one hand improves the economy and environmental protection of the whole system.On the other hand,because the output of renewable energy is affected by the wind,similar to the energy demand side,it fluctuates greatly,resulting in complicated system scheduling and bringing risks to the stable operation of the system.In view of the above problems,this paper takes an integrated energy system as an example,comprehensively introduces the internal equipment of the system,uses the deep learning algorithm to effectively predict both sides of the source and load,and on this basis,further studies the effective optimization scheduling method of the integrated energy system.First of all,The accurate mathematical modeling of the energy equipment and energy network in the integrated energy system is the premise of ensuring the optimal scheduling of the integrated energy system.In this paper,the operation mode of each equipment in the integrated energy system is introduced in detail,and the corresponding mathematical modeling is given,which provides a mathematical basis for the subsequent system operation optimization.The characteristics of multiple load on the demand side are analyzed in detail,the influencing factors are analyzed from multiple time scales,and the internal rules of multiple load are summarized.Secondly,the mean influence value and Informer algorithm was used to predict the renewable energy and multiple load.After data processing,the mean influence value method based on BP neural network is applied to quantify the influence of various factors on target variables.The ProbSparse self-attention mechanism in Informer model was used to reduce the complexity of the model,simplify the network structure,reduce the usage of memory,and increase the calculation accuracy.The effectiveness of the the mean influence value and Informer algorithm is verified by comparing the prediction results of different algorithms for wind power and multiple load.Finally,a multi-time scale optimization scheduling method for integrated energy system considering energy characteristics and controllable load is proposed.In the comprehensive energy system,considering the different energy transmission rates and the different response times of the same energy production equipment,the multi-time scale optimization scheduling is proposed.The optimization scheduling is carried out from the three levels of day-ahead,day-day and real-time.The day is divided into two scheduling cycles,and the corresponding slow equipment and the corresponding fast equipment are optimized respectively.By comparing different scenarios,the results show that the proposed optimal scheduling method has good economic performance.