Research On Collaborative Control Strategy Of Demand-side Aggregated Loads

The energy transformation promotes the development of low-carbon and green renewable energy and accelerates the construction of smart grid.In the future,renewable energy will be the main energy that social development depends on.And it will be connected to the smart grid on a large scale to provide energy power for social development.However,the renewable energy has the characteristics of uncertainty and fluctuation,which make it difficult for the power grid to accurately predict its actual load output.With the integration of large-scale renewable energy,it is difficult for the grid to maintain the real-time balance of supply and demand,which seriously threatens the safe and stable operation of the grid.The traditional generation regulation method based on matching load demand has the shortcomings of high cost,low efficiency and slow response.And it is difficult to meet the demand of supply-demand balance regulation with the penetration of large-scale renewable energy.Faced with this situation,it is urgently needed for the power grid to develop other new modes to supplement the shortcomings of traditional the regulation mode and enhance the ability of maintaining supply-demand balance.Electricity market reform provides the condition for demand-side load resources to participate in the supply-demand regulation.Demand response(DR)provides an implementation form.For power grid,demand response can not only achieve peak-shaving and valley-filling,suppress load fluctuation,enhance the ability of maintaining the supply-demand balance,but also promote more renewable energy to be connected and to be accommodated.For energy consumers,it improves the interaction level and provides value-added services.Therefore,demand response will be an indispensable supporting technology for the smart grid.Residential loads are good resources for demand response.And the electric water heater(EWH)is one of the main energy loads at the residential side.So,this dissertation takes electric water heaters as an example to study cooperative control strategy of aggregated loads for demand response.The main research contents of this dissertation are as follows.Aiming at demand response for load reduction,this dissertation studies the collaborative control strategy of aggregated loads under the mode of load aggregators with fixed incentive mode.And the THT strategy considering the transferable heating time is proposed.Based on the analysis of the energy flow of an electric water heater,the calculation model of the heating time is established.And On the basis of the heating time model,the transferable heating time is defined according to the user's requirement.According to the transferable heating time,the THT strategy and its implementation architecture are designed,including grouping operation,evaluation operation,selection operation and update operation.The grouping operation defines a grouping rule of aggregated loads.The selection operation establishes a selection rule based on the transferable heating time.And the update operation formulates a real-time update rule for individuals which are chosen to participate in demand response.Aiming at demand response for load increase,this dissertation studies the collaborative control strategy of aggregated loads under the mode of load aggregators with fixed incentive mode.And the RTS strategy based on priority ranking is proposed.According to the theory of cyber-physical system,a data-driven TRS strategy and its implementation architecture are designed,including state awareness,real-time analysis,scientific decision-making and precise execution.State awareness formulates a grouping rule taking into account the constraints of real-time state and thermostat setpoint range.Real-time analysis establishes a method correcting the response load amount and a method determining the type of execution action.Scientific decision-making defines the selection priority of each load and establishes a selection rule based on the selection priority.Precise execution specifies the method of generating control signals for selected individuals.In view of the situation that data processing pressure may be high in the cloud center of load aggregator,this dissertation studies a strategy to alleviate the data processing pressure and proposes a TCIS strategy taking control interval into account.Based on the collaborative thinking of local computing and central processing,the TCIS strategy and its implementation architecture are designed.The TCIS strategy transfers some data computing work to load terminals.Based on the above idea,a characterization function of response characteristic considering control interval is defined for each load.And a response state model which reflects the response type and reponse priority,is established to support the TCIS strategy to some data computing work to load terminals.According to the defined response state model,the TCIS strategy is designed and a decision rule based on the response state is formulated.Compared with the TRS strategy,the TCIS strategy reduces the frequency of data interaction and the data processing pressure of load aggregator center,by increasing the control interval and transfer partial computing work to load terminals.Aiming at demand response for load increase,this dissertation studies the collaborative control strategy of aggregated loads under the mode of load aggregators with flexible incentive mode.And a UCOS strategy based on personalized customization is proposed to minimize reward payments to the users.Aiming at the multi-level incentive under flexible incentive mode,the optimization problem in the decision-making process is analyzed.And according to the analysis result,a data-driven UCOS strategy is designed,which includes state awareness,real-time analysis,scientific decision-making and precise execution.The generation principle of response characteristic identification for each load is defined in the logic of state awareness.In the logic of scientific decision-making,an optimal decision-making rule for load selection is established with the objective of minimizing reward payments to the users and the constraints of real-time response characteristics and users' personalized requirement.