Research On Key Technologies For The Enhancement Of Source-load Regulation Capability In Urban Power Systems

As the main form of human society development,cities have a huge demand for energy consumption,and power systems have become the lifeline of city energy supply.It is worth noticing that the traditional industries with steady load curves are moving out of the urban area because of their heavy contamination.Meanwhile,the tertiary industry and residential sector gradually become the major part of urban electricity consumption,which intensify the volatility of the power load.Besides,a large amount of distributed new energy generations are being pluged into the urban power systems.These features demand for higher regulation capability of the urban power sources.However,with the rapid increase of electricity imported by High Voltage Direct Current(HVDC)transmission lines,some traditional generators have been retired,so that the regulation capability is impaired.Under such circumstances,the conflict between the limited regulation capability and the increasing regulation demand has become one of the major problems in urban power system operation.Therefore,it is necessary to have a better control of the load volatility,to guarantee the regulation capability of urban power sources and to explore the regulation potential in the power systems.Aiming at enhancing the regulation capability of the urban power system,this dessertation studies the corresponding key technologies from both the aspects of power source and power load.The main works are summarized as follows:Short-term load forecasting technique for urban power systems is studied.A large amount of substation load curves with hourly interval are analyzed and forecasted to reflect the inherent law of urban power load,as well as to restrict the forecasting random error caused by the load volatility.Based on this,a complementary clustering method is proposed to preprocess the substation load curves,so that the number of load curves can be reduced while their smoothness can be improved.In this way,both the forecasting random error and the modelling error are reduced,and the accuracy and stability of the forecasting results are improved.An accuracy load forecasting is an important precondition of urban power system operation.The smart control strategy for urban air conditioning load is studied.Based on the smart thermostat installed by users,an operation strategy is designed to control and optimize the air conditioning load.The user preference model is firstly constructed to reclect the temperature preference of users and their trade-off between thermal comfort and electricity purchasing cost.Then Bayesian inference is introduced to quickly learn user preferences based on limited user observations.Next,a multi-stage Stochastic Expected Value Model(SEVM)is proposed to generate the optimal temperature setting schedules in the next few hours.Finally,an overall operation strategy is designd to enable the thermostat to track the changes of user preference and update the temperature schedules continuously.In addition to providing thermal comfort and energy saving for users,the smart thermostat together with the proposed operation strategy also works as a fundamental execution unit during demand response.The load flattering technique based on demand response is studied.The incentivebased demand response mechanism is proposed from the aspect of load aggregator(LA)to control and shave the air conditioning loads in the area.The interactions between users and LA are modelled by Stackelberg Game to determine the optimal pricing strategy for the LA.The existence of equilibrium is theoretically analyzed,and an iterative algorithm between LA and users is designed to quickly reach a satisfying pricing strategy while protecting the privacy of users.Finally,the optimal compensation is discussed and the elasticity of air conditioning load toward price and outdoor temperature is analyzed.The proposed load flattering technique can effectively shave the air conditioning load during peak hours,which reduces the regulation burden for the urban power grid.The method to determine the reliable online generation capacity is proposed.A main threat to the urban power grid is the deep frequency drop after the bipolar block of HVDC transmission line.Based on this concern,the relation between system requency drop under large amount of power shortage and local online generation capacity is analyzed,and a corresponding correlation model is constructed.Besides,the correlation model between ancillary service capabilities and the local online generating capacity is also provided.Based on the above two correlation models,the reliable online generation capacity can be determined according to the requirements of system frequency drop control and ancillary service adequacy.When operated under the reliable online generation capacity,the urban power system can provide enough regulation capability and guarantee the frequency stability under the impact of HVDC failure.