Optimal Scheduling Methods For Demand Response Resources At The User Side

With the rapid development and wide application of novel technologies including distribution generation,energy storage and smart power utilization,the user-side paradigm of modern power systems have changed significantly,and many new opportunities and challenges appear.Demand response is at the cutting edge of power system research,and is considered to be a crucial solution to the power and energy problem faced by our country.Developing demand response fits well with the policies and strategies of our country,and is benefitial to the long-term sustainable development of our economy and society.This thesis focuses on optimal scheduling methods for demandresponse resources at two levels,household level and aggregation level.The main work of this paper is fourfold as follows:1)A traversal-and-proving algorithm is proposed to scheduling household thermostatically controlled loads optimally based on the operational characteristics of the appliances.Both qualitative and quantitative analysis are carried on to evaluate its calculation complexity and optimality.Compared to the tratidtional algorithm,the proposed traversal-and-proving algorithm obtains more economical schedules at the premise of thermal comfort.The proposed algorithm is of great robustness and flexibility as well.2)A robust-index method is proposed to tackle the uncertainties of user behaviors at the household level.To be more specific,robust index are defined to quantify the robustness of the schedules of three different types of household loads,and they are normalized to be integrated into the household load scheduling optimization problem to improve the robustness of the schedules.Through the proposed robust-index method,the schedules with expected robustness can be obtained,and thus the risk brought by user behavior uncertainties can be limited.Besides,compared with stochastic programming,the proposed robust-index method is superior in simple modeling,being indepenet of historic data and lower computation cost.3)A robust scheduling method is proposed to tackle the uncertainty of household photovoltaic systems.Specifically,the robust counterpart of the household load scheduling optimization problem with photovoltaic systems is established and transformed to a second-order programming problem that can be easily solved by existing optimization tools.Schedules with different robustness levels can be obtained by the proposed method to resist the uncertainty of household photovoltaic systems.Compared to the traditional worst-case method,the proposed robust scheduling method is more flexible and capable of obtaining better schedules.4)An optimal scheduling method is proposed to schedule aggregated industrial thermostatically controlled loads in combined energy and frequency response markets.Practical energy and frequency response markets in the UK are considered,and bitumen tanks are taken as an example.The optimization problem is established and a heuristic three-stage algorithm is proposed to solve it.Under different tariff structures and frequency response fee levels,the proposed method brings more economic benefits for load owners compared to the traditional hysteresis control and the optimization method that considers the energy market only.