Research On Integrated Energy System Scheduling Based On Robust Rolling Optimizatio

With the increasing proportion of renewable energy generation,the impact of its output power uncertainty on the grid cannot be ignored.As an important form of integrating renewable energy with the main grid,the integrated energy system can smooth out the random fluctuations of renewable energy with its multi-energy complementary characteristics and high operational flexibility.However,the large-scale integration of renewable energy into the integrated energy system will lead to the problem of system supply-demand imbalance.To address this problem,this paper studies the uncertainty modeling and optimal dispatching methods of integrated energy system.The research work is mainly as follows:Firstly,the source,network and load models of the integrated energy system are constructed according to the operational characteristics of different forms of energy in the integrated energy system and the coupling mechanism.In the energy transmission and conversion link,a pipeline virtual decomposition of heat network modeling method is proposed,which solves the problem of mismatch between the calculated delay time and the actual delay of multi-pipeline coupling by decomposing the main heating pipeline into multiple virtual fine pipelines,so that each heat load is supplied by one virtual pipeline individually.The simulation results show that this method can improve the calculation accuracy of delay time compared with the weighted delay method with the same calculation step.Secondly,the output power of renewable energy is characterized by randomness and unsteadiness,and its direct grid connection will adversely affect the safe operation of the power system.To address this problem,a two-stage rolling optimization scheduling method is proposed to give full play to the flexibility of time-domain rolling optimization and the thermal inertia of the heating system to eliminate the effects of renewable energy output power uncertainty and day-ahead load prediction error to the maximum extent.The simulation verifies that the thermal inertia energy storage model of the heating system can reduce the operation cost to a certain extent,and the two-stage rolling optimization measure can maintain the contact line exchange power following the planned value and reduce the fluctuation of the battery output.Finally,a robust optimal scheduling method based on Wasserstein measure for time domain rolling distribution is proposed with full consideration of the stochastic characteristics of renewable energy power output.The method calculates the uncertainty fuzzy set by integrating the historical sample data of the energy system,while introducing contact line penalty coefficients for regulating the system contact line exchange power.The time-domain rolling distribution robust optimization incorporates the idea of rolling predictive control into the robust optimal scheduling,improves the prediction performance and reduces the conservativeness of the robust algorithm by using intra-day rolling prediction.Compared with the day-ahead robust optimization method,the domain rolling distribution robust optimization method can reduce the abandonment rate of photovoltaic power generation,and at the same time,the contact line power can deviate less from the day-ahead planned value,reducing the impact of source-load prediction errors on the grid.