Identification And Operational Optimization Of Integrated Heat And Electricity Systems

In the era of “dual carbon” goals,reducing carbon emissions to cope with global warm-ing has become a global consensus,while the transformation of the energy structure is the de-terminants for the short- and medium-term low-carbon process.With the increasing amount of low-carbon electricity represented by wind power and photovoltaic power and the growing demand in urban heating,a scientific and technical problem that needs to be solved urgently in the energy system is how to coordinate thermoelectric coupling in a large-scale new energy grid-connected environment,and the whole process of “source-grid-load-storage” operation op-timization of electric energy and thermal energy can be realized in a safer,more economical and flexible way.The integrated heat and electricity system provides a feasible path to realize the coor-dinated operation of heating and power system,and improve the absorption capacity of new energy.In the framework of the integrated heat and electricity system,thermal energy and electric energy are coupled with each other at all levels,which expands the control boundary of the original isolated energy system,and can effectively improve the efficiency and flexi-bility of energy conversion.However,the theoretical research and engineering practice on the integrated heat and electricity system are still in the preliminary stage.On the one hand,the cou-pling mechanism between the production,transmission,storage,and consumption of thermal energy and electric energy is very complicated,and the characteristics of different time scales re-flected in the whole system bring severe challenges to modeling and controlling efforts.On the other hand,the inherent randomness of renewable energy significantly increases the difficulty of adjusting the integrated heat and electricity system,and also puts forward higher require-ments for system flexibility.Moreover,the integrated heat and electricity system contains a variety of energy conversion units,whose operation mechanism is different,and the operation optimization is mainly completed online due to the operation requirements,which also chal-lenges the existing real-time optimization methods.Therefore,this paper conducts research on the integrated heat and electricity system at four levels of “modeling-controlling-optimization-dispatching”.Firstly,the system identification method for integrated heat and electricity system with two-time-scale characteristic is studied,and then,the two-time-scale model predictive con-trol method for rapidly power response of integrated heat and electricity system is studied.After that,the real-time efficiency optimization method for production and conversion units in the in-tegrated heat and electricity system is studied.Finally,the flexible peak shaving method of integrated heat and electricity system is studied.The main work and innovations of this paper include:(1)Aiming at the two-time-scale modeling problem of integrated heat and electricity system,a two-time-scale system identification method is proposed based on system identification principle.To solve the problem of mutual interference of wide-frequency domain infor-mation caused by two-time-scale coupling,the design of high-pass filter and low-pass filter and information elimination method are used to separate different timescale infor-mation to enhance the SNR of different frequency bands,and then improve the overall modeling accuracy of the system.The consistency of the method is proved,and the ef-fectiveness of the method is verified by Monte-Carlo simulation.The proposed method solves the problems of biased estimation and convergence of the traditional output error and prediction error identification methods in the low frequency band of the two-time-scale system,and is robust to multivariable problems,which can provide model support for two-time-scale control optimization.(2)Aiming at the two-time-scale control problem of integrated heat and electricity system,a two-time-scale predictive control method for fast power response is proposed.According to the fast and slow dynamic characteristics,the model predictive controller is set respec-tively.Through the information interaction between the controllers,the heating steam flow is dynamically adjusted to obtain fast electric power response by using the heat stor-age capacity of the heating system on the premise of satisfying the user’s comfort,and the performance of the coordinated control system(CCS)is significantly improved.The pro-posed method makes collaborative optimization decisions for power and heating systems,and realizes flexible energy allocation from plant level to district heating level.(3)Aiming at the problem of online efficiency optimization of energy production and conver-sion units in the integrated heat and electricity system,a real-time optimization method based on system identification is proposed.The method applies dynamic data and uses a multivariate closed-loop identification algorithm to obtain gradient.The gradient uncer-tainty is estimated by the asymptotic theory,and the signal-noise ratio(SNR)is dynam-ically modified by the online experimental design method to compensate for the uncer-tainty.Finally,the online iterative optimization is realized by using the feasible direction optimization algorithm.The proposed method solves the problem of gradient and its uncertainty estimation,and forms a complete optimization process of “dynamic exper-imental design-gradient estimation-gradient error estimation-online iterative optimiza-tion”,which improves the operability of real-time optimization and obviously reduces the cost of the optimization process.(4)Aiming at the peak shaving problem of the integrated heat and electricity systemin the un-certain scenario,a day-ahead and intra-day combined heat and power scheduling strategy is proposed.Considering the peak shaving characteristics of a variety of generators and the influence of wind and photovoltaic power uncertainty,a scheduling model including peak shaving cost and thermoelectric dynamic coupling constraint is established.The pro-posed method uses a two-stage distributionally robust optimization method to deal with the day-ahead scheduling problem under uncertainty of wind and photovoltaic power,and adopts a rolling optimization method based on model predictive control to solve the intra-day scheduling problem.The proposed method balances operational economy and robustness,which significantly increases the peak shaving potential and reduces the peak shaving cost of the system.