Modelling And Control Of Thermostatically Controlled Loads For Participating In Dynamic Response Of Power Systems

The power generation share of renewable energies is increasing rapidly in power systems to deal with the environment pollution and resource shotage.However,compared with traditional generating units,such as thermal and gas power generation,renewable energy generation can bring more fluctuations and intermittence to power systems,which impacts the secure,stable and economic operation seriously.Traditional power systems achieve the system balance between generation-side and consumption-side by adjusting the output of generating units,which is called the reserve capacity.With the increasing proportion of renewable energies,the necessary reserve capacity is becoming larger rapidly,which increases the operation cost of power systems.Therefore,it is urgent to explore more regulation resources to solve the fluctuation problem caused by renewable energies.In the power consumption-side,the load structure has changed with the improvement of the living standard.More thermostatically controlled loads(TCLs),such as air conditioners,water heaters and heat pumps,are installed and have become one kind of the most important loads.The load peak valley difference of the power system is further enlarged by large-scale TCLs,especially in extremely cold and hot weather days.This also brings severe challenges to the secure and economic operation of power systems.Faced with above challenges from generation-side and consumption-side,the rapid development of information,communication,and intelligent control technologies bring new opportunities to improve the flexibility of power systems.The load power consumption can be reduced or transferred to provide regulation capacities for power systems,which is called demand response(DR).TCLs have natural advantages of participating in DR: i)the proportion of TCLs is high in the total load power consumption,and has large regulation potential;ii)TCLs have energy storage capacity,and can be regulated flexibly in the premise of ensuring users' comfort;iii)the high power consumption period of TCLs is precisely the period of power system's peak load,when the regulation capacity from generating units is insufficient and TCLs can make up for the regulation capacity shortage.Therefore,it is urgent to study the methods of using TCLs to improve the flexibility of power systems,and solve the problem of high penetration of renewable energies and insufficient system regulation capacity.However,the operation of TCLs is affected by many factors,such as building structure,thermal insulation materials,and user habits,which make it difficult to develop the TCLs' model.Besides,it is impossible to quantify the regulation capacity of TCLs accurately as generating units due to the heterogeneity.The individual capacity of TCL is small,geographical location is scattered,and the number for providing regulation capacities is large,which make it difficult to aggregate and control.In order to gurantee stability of power systems,the feasible region of TCLs' regulation is difficult to solve when considering the dynamic response.In addition,the communication delay for implementing DR is inevitable and will introduce new oscillations to power systems.The terminal controller probably have detection errors of power system operation parameters and will cause the difficulty of accurate response capacity control.It is also necessary to allocate the limited number of adjustable TCLs and limited adjustment times to different areas of power systems,to carry out DR at different time,which resultes in multi-dimensional massive parameters control optimization problem.To address above issues,this thesis has carried out in-depth research work from five aspects.(1)A typical TCL model is constructed,and a quantitative calculation method of response regulation capacity is proposed.In this manner,the aggregated TCLs can be incorporated into the existing power dispatching system,and the system operator can dispatch TCLs and traditional generating units uniformly.(2)The frequency domain model of TCLs is developed and equivalent to the model of traditional generators.On this basis,a hybrid stochastic control method of TCLs is proposed by considering the deadband control and hysteresis control,to achieve large-scale TCLs aggregation and flexible regulation.(3)The steady-state error calculation method and the frequency stability analysis method of power systems considering the dynamic response of TCLs are proposed.The regulation feasible region of TCLs is framed considering the influence of communication delay,which ensures the secure and stable operation of power systems.(4)Based on cloud-terminal measurement and detection error modification methods,a centralized-decentralized control architecture of aggregated TCLs is proposed to solve the load oscillations caused by communication delay during DR process The generalized error distribution theory and the least deviation unbiased estimation are used to improve the control accuracy of large-scale TCLs' regulation capacities.(5)A multi-area power system model considering the dynamic response of TCLs is developed,in which the large-scale TCLs' control strategy considering the area control error is proposed.A multi-area and multi-time scale regulation capacity distribution and optimal schedule method is also proposed considering the TCLs' regulation capacity and regulation times constraints,so as to achieve the global optimum of TCLs' regulation in multi-area power systems.