Energy Optimization Management Of Integrated Energy System Considering Multi-energy Response

With the development of intelligent information technology and energy Internet,the coupling between different forms of energy flows such as electric,heat and gas has been deepened in integrated energy system,which brings greater diversity of energy consumption,stronger load uncertainty,and higher energy comfort requirements to the user side.Meanwhile,the amount of multi-energy flow information that can be collected and processed by the system grows sharply.The exploded energy system information not only challenges the accuracy of system regulation and operation strategies,but also enhances the possibility of strengthening the interaction between energy supply side and demand side and of improving the comprehensive system energy efficiency.Therefore,economical,fast and accurate energy optimization management is needed for the integrated energy system.This paper focuses on typical integrated energy system represented by smart buildings and smart communities.The main work can be conclued as follows:(1)Potential of data predictive control(DPC)in smart building energy management system is studied.Firstly,a day-ahead and intraday model predictive control(MPC)energy optimization model of a typical multi-layer intelligent building system based on heat storage characteristics of envelope structure is established.Secondly,DPC control strategies based on machine learning algorithms artificial neural network(ANN)and random forest(RF)are proposed to make rapid error feedback correction for the intraday energy consumption decisions made by MPC model.The DPC methods avoid the huge time cost and effort paid by traditional MPC algorithm to learning a gray/white box model of the actual physical building system,effectively reduce the impact of load forecast errors on decision-making and provide economical,accurate and fast energy management solutions for the system.(2)A user-dominated smart community energy management and transaction method focused on mechanism design is proposed.Firstly,a user dominated demand response(UDDSR)mechanism is established.UDDSR allows community users to submit flexible integrated demand response(IDR)bids to the community energy management system(CEMS)with flexible start time,stop time and response durations with regard to users’ comfort zones for centralized heating system and home appliances,which gives maximum freedom to the IDR participants.Secondly,a reward mechanism for UDDSR participants is set up and a IDR bids aggregation optimization model based on penalty coefficient is constructed.Moreover,a community energy pool Peer-to-Peer(P2P)transaction mechanism based on user energy surplus/demand ratio is built to facilitate efficient energy usage among neighbourhoods.(3)The energy consumption optimization method of community integrated energy system based on UDDSR is studied.Firstly,the mathematical models of UDDSR participants’ responsive resource such as interruptible power load,shiftable power load and adjustable heat load are established.Secondly,the integrated building thermal model is introduced to measure the temperature requirements of the overall community users for system heat supply.Thirdly,a day-ahead cheduling model was established to optimize the energy consumption of the UDDSR-based community integrated energy system.Then,taking forecast errors of photovoltaic output,user load,outdoor temperature,and user actual UDDSR response capacity into account,a penalty mechanism is introduced to punish the particants who make imbalance response against the day-ahead IDR bids.Finally,a CVa R-based community energy optimization method with the penalty mechanism is presented to enhance the robustness of the day-ahead scheduling model of the community system under different prediction accuracy.