Research On Electricity Price Mechanism Design And Electricity Devices Random Scheduling Algorithms For Demand Response

In order to deal with the energy crisis and some shortcomings of the traditional electricity grid, Smart grid is advocated, which is a hot topic in the development of new technology and industry all over the world. Demand-side response is a key imponent in the future smart grid and it has some fantastic advantages, such as fast response, fewer emissions, and lower cost. It can be used to reduce system peak tariff and the risk of electricity price fluctuations, optimize the allocation of energy resources, and ensure the stability of electricity market, playing an important role in the electricity industry and economic development. The development of smart grid technologies lays a foun-dation for the implementation of real-time pricing, due to its integration of advanced information, control and communication technology. What’s more, the smart grid en-ables the function of plug-and-play, which can achieve a seamless connection between the home energy storage devices and the grid. The popularity of PHEV and some other energy storage devices also brings new challenges to the schedule of electricity devices under the real-time energy price. This paper studies the design of real-time electrici-ty pricing mechanism, and random scheduling of electricity appliances and household energy storage device, under the condition of uncertain electricity price.First, we consider the design of real-time energy price under the condition of un-certain users’electricity load, using the method of mechanism design in game theo-ry. We consider a residential electricity network, composed of an electricity service provider and many residents. The service provider purchases the energy from the elec-tricity grid and assigns it to users with a reasonable charge. The amount of energy needed by residents is uncertain. The consumers are considered to be selfish, and s- trategic, who are reluctant to report their true power demand. Therefore, we investigate a pricing mechanism, inducing users to report their true information about the energy demand. We build a system model to describe the consumer’s utility and electricity costs. We develop a pricing algorithm, which maximizes the expectations of average social welfare, and meets incentive compatibility and individual rationality constraints as well. We give the simulation results, showing that our proposed strategy can help users to reduce financial costs and improve society welfare.Second, we discuss the scheduling problem of household appliances and storage devices, under the condition of uncertain energy price, and advocate a distributed on-line learning scheduling algorithm. In our system model, the consumer has some de-ferable energy request. While the delay will cause a negative effect on the satisfaction level of users. Also, the user is equipped with a storage device, which has the ability to charge and discharge. Based on this model, we explore the problem about how to minimize the user’s expected average cost. Because the energy request is influenced by users’habits, it is hard to characterize the randomness of energy request quantitative-ly. Therefore, we propose a stochastic scheduling algorithm based on online learning method, to jointly schedule the operation of household electrical equipments and stor-age devices. This algorithm has a low complexity, and can be adaptive to the changes of external environment factors. As we can be seen from the simulation results, the proposed algorithm can help the user to reduce his total cost.Finally, the research work in this thesis is concluded and the future research di-rections is discussed.