Research On Multi-energy Coordinated Optimal Dispatch Of User-level Integrated Energy System With CCHP

The integrated energy system(IES)contains multiple energies such as electrical,heating,cooling and natural gas energy,which involves energy conversion,distribution,coordination,and it is an important physical carrier to realize the multi-energy complementary of the energy internet.The user-level IES has the characteristics of large load fluctuations,low prediction accurac y and complex energy coupling because of small-scale.It requires more accurate and reliable multi-energy load forecasting method and real-time effective scheduling strategies.Therefore,an ultra short term load forecasting method considering multi-energy spatio-temporal coupling,a day-ahead scheduling strategy considering electricity and gas reserve,a hybrid time-scale strategy in intraday scheduling considering energy characteristics are proposed to improve the flexibility and reliability of system operation.Firstly,the spatial correlation of basic load units is analyzed and the multiple load pixels with strong correlation between adjacent load units are constructed through K-means clustering and multi-dimensional scale analysis.Various load pixels are input into their convolution-pooling channels in time order to form a multi-channel convolutional neural network(MCNN)with spatial feature extraction.Long and short-term memory network is then used to further capture the time characteristics of high-dimensional load sequence to form a deep spatiotemporal load forecasting framework of user-level IES to improve the accuracy of multiple load forecasting.Secondly,the basic structure of user-level IES is introduced and the detailed mathematical model of each device is given.In view of the uncertainty caused by the forecasting errors of multiple load and renewable energy output,several power and gas reserve scenarios are given based on the opportunity-constrained programming model.By comparing the day-ahead scheduling cost under different scenarios and confidence levels,the optimal reserve schdule is determined.At the same time,the optimal reserve confidence level suitable for user-level IES is trade-off between economy and reliability,so as to prepare for intra day adjustment.Finally,a hybrid time-scale scheduling strategy considering the differences in energy characteristics is proposed in intraday scheduling.Determining the time resolution of slow layer involing heating,cooling enegy and fast layer involing electrical enegy based on the error distribution between the measured data and the predicted data,combined with the tolerance of the imbalance of energy supply and demand in the intra-day scheduling.The low time resolution of cooling and h eating energy is regarded as the upper level and the high time resolution of electrical energy is regarded as the lower level in intraday correction.The adjustment of the intraday device output plan is completed in the order of upper and lower level corre ctions based on the day-ahead scheduling plan.The results show that the proposed strategy can take into account the imbalance between supply and demand and the number of equipment adjustments.