Demand Response Mechanism Models And Its Application Considering Uncertainty

The adoption of Demand Response (DR) concept has changed the traditional way of satisfying the growing power demand merely by investing in generations. Rather, demand-side resources are now supplements of the supply side generations. DR alleviates the tight power balance between supply and demand, protects the environment and conserves energy. In fact, DR exhibits uncertainty in its mechanism, where its characteristics are affected by various factors that are uncertain. Therefore, an in-depth study of modeling and applications of DR mechanism considering uncertainty is significant for DR scheduling in the electricity market.This thesis studies the mechanism of DR and proposed mathematical models considering its uncertainty. Major works in this thesis are summarized as following:1、Analyzed DR uncertainty characterizes based on uncertainty theoryThe uncertainty factors in DR programs are analyzed comprehensively, including the different viewpoints of consumer response types and uncertainty theory branches. The current research focuses are summarized as three aspects:randomness, fuzziness and roughness, providing the theoretical basis of the following studies in DR modeling. The analysis also provides future research topics regarding DR uncertainty analysis.2、Proposed a linear DR model using customer baseline load considering uncertaintyLoad variability index, weather sensitivity index, bias index and accuracy index are adopted to assess the calculation method of baseline load. The clearing method for interruptible load considering uncertainty is proposed. On this basis, The DR linear model is formulated and applied to the initial stage DR program implement. The cost function including the payment cost and the risk cost of electricity companies is proposed as well as the detailed operation constraints.3、Proposed a quadratic DR model for virtual power plant considering uncertaintyClarifying the structure of electricity market, the participating mode of DR is studied, considering the characteristic of different market models. With the progress of Smart Grid and electric industry reform, the structure of electricity market develops toward a competitive mode. Based on the concept of virtual power plant, the deterministic and uncertain models of incentive-based and price-sensitive DR. models are proposed, which can be optimized to bid with generators together. The derived cost functions are quadratic and are similar to the traditional generator.4、Adopted the proposed linear DR in day-ahead dispatching optimization.In the background of initial-stage DR application, the DR linear model based on consumer baseline load considering uncertainty is adopted to formulate the chance constrained model to meet the uncertain characteristic of load curtailment requirement. The optimization decision method of uncertain interruptible load is proposed based on risk assessment and chance constraint.5、Adopted the proposed quadratic DR model in an integrated generation-consumption dispatch scheme.In the background of popularized-stage DR application, the DR quadratic model is adopted for virtual power plant (VPP) considering uncertainty. Incentive-based and price-sensitive DR are included in an integrated generation-consumption dispatching scheme together with traditional generators. The inclusion of DR based VPP improves the system regulation capacity and capability for renewable integrations. The consideration of DR uncertainty also improves system reliability, reduces potential risk costs, and promotes the customers for better response reliability.