Research On Energy Internet Optimal Scheduling Considering Demand Response Under Cloud-edge Environment

Energy Internet can realize the synergy,complementarity and efficient operation of electricity-dominant multi-energy systems,and its open,interconnected,distributed,and user-centric features are also in line with cloud-edge collaborative scheduling.On the one hand,the configuration of edge-side intelligent computing and information collecting equipment in the context of Energy Internet construction provides a physical basis for cloud-edge collaboration.On the other hand,cloud-edge collaboration also provides a solution for the optimal control of large amounts of demand-side controllable devices in the Energy Internet.It is very important to rely on the cloud-edge collaboration technology to realize the hierarchical optimal scheduling of various devices in Energy Internet,so as to fully exploit the potential of load-side regulation,and to achieve multi-energy complementarity and load-energy system flexible interaction.This paper conducts research on Energy Internet cloudedge collaborative scheduling considering demand response,and the main work is as follows:(1)Based on the energy system scheduling logic,physical structure and participants,an Energy Internet cloud-edge collaborative scheduling framework is established to support the implementation of energy system global scheduling as well as load-energy system interaction.The multi-time-scale scheduling process of Energy Internet is introduced,and the market environment and the scheduling solution process are analyzed in detail in the two scenarios: day-ahead scheduling and intraday demand response.A characteristic analysis method for energy consumption side entities to participate in demand response is established to support load-energy system interaction.(2)For day-ahead optimal scheduling scenario,considering the cloud-edge scheduling architecture,market environment and multi-energy network interaction process,a hierarchical optimal scheduling method for Energy Internet is proposed,considering load-energy system interaction and multi-energy complementarity.The upper layer optimization is used for the global optimal scheduling of Energy Internet,and the lower layer optimization is used for scheduling the internal adjustable equipment in a single load entity.for the above two-layer optimization scheduling method,considering the characteristics of each layer’s optimal scheduling,a two-layer optimization based on improved particle swarm algorithm is proposed.Considering the characteristics of each layer optimization model,a solution method for the twolayer scheduling model based on improved particle swarm optimization is proposed.The case study shows that the proposed method can further tap the potential of load side regulation through multi-energy demand response and effectively reduce the operating cost of Energy Internet.(3)For intraday demand response scenarios,a distributed price-based demand response pricing method based on A3C(asynchronous advantage actor-critic)algorithm LSTM(long short-term memory)network is proposed.Through the distributed training and centralized decision-making structure of the A3 C algorithm,the local utilization of load information is realized.The virtual environment based on LSTM is used to replace the actual energy system to participate in the training process,reducing the training cost of the algorithm.Calculation results show that the proposed method requires less historical data,saves the cost of demand response implementation,and is capable of supporting demand response pricing decisions in cloud-edge environment.