Reserve Optimization And Distributed Dispatch For Power Systems With A High Share Of Wind Power

Power systems with a high share of wind power are not only an important feature of the future development of smart grids but also an important way to promote energy low-carbon transition.However,due to the inherently strong uncertainty of wind power,it also brings new challenges to the system operation.To improve the system's wind power consumption capability and ensure the safety of the entire network,it is essential to optimize the operating reserve schedule.Most related researches only optimize the reserve provided by generators within a single system,which fails to make a reasonable wind consumption capability assessment of the current system and cannot comprehensively consider the coordination of multiple agents and multiple generation resources.Under this background,this thesis conducts theoretical research on the reserve scheduling problem of power systems with a high share of wind power from the four perspectives of ‘multi-scenarios',‘multi-resources',‘multi-levels' as well as ‘multi-areas',and proposes a distributed framework to solve the multi-agent collaborative scheduling problem,which is mainly summarized as follows:1)In terms of ‘multi-scenarios',a concept of optimal wind power consumption point is proposed to evaluate the system's wind consumption capability and to ensure the wind consumption level in multiple possible scenarios.Based on that,a two-stage chance-constrained model is established to coordinately optimize the unit commitment and reserve schedule.This model also takes into account the effect of battery energy storage on improving system flexibility.To avoid solving a large-scale optimization problem incorporating multiple scenarios and improve the convergence of the solution process,a novel optimality-check-only bilinear Benders decomposition method is proposed.Case studies show that the computational efficiency of this method can be improved by more than one order of magnitude compared with the original one.2)In terms of ‘multi-resources',by modeling wind farm reserve,incentive-based demand response,and price-based demand response,the potential of multiple reserve resources to improve system flexibility is fully utilized.Based on the two-stage adaptive robust optimization,a multiple reserve scheduling model is established.This model uses an uncertainty set to describe the wind uncertainty,which guarantees the system operation safety under the worst operating scenarios.By adjusting the size of the uncertainty set,the cost-effectiveness and safety of system operation can be balanced.The most advanced column-and-constraint generation algorithm is used to solve the proposed model.Case studies further analyze the application scenarios of these reserve resources and their impact on the system's wind consumption capability.3)In terms of ‘multi-level',a coordinated robust reserve scheduling model for the coupled transmission and distribution systems is proposed.This model not only determines the base-state boundary power but also optimizes its adjustment capacity in the re-dispatch state to promote multi-level coordination for handling the entire system wind uncertainty.The uncertainty of the active distribution networks' equivalent load is also characterized to ensure the boundary power balance.Based on the alternating direction multiplier method,the decomposition mechanism of the coupled systems is proposed,with its feasibility proved.A distributed framework is established to ensure regional information privacy and reduce the communication burden between levels.A two-layer iterative process is further proposed to improve the convergence property.4)In terms of ‘multi-area',a risk-based robust reserve scheduling model is established for the multi-area interconnected power systems.By optimizing the tie-line power and its adjustable range,the multi-area generation resources are coordinated,which enables reserve sharing among connected areas,improves the cost-effectiveness of system operation,and promotes wind power consumption.By introducing the probability density function of wind output into the two-stage robust model,the system operational risk is also evaluated,with an admissible region of wind power obtained correspondingly.Based on the alternating direction multiplier method,a distributed framework is established.A dynamic penalty factor adjustment strategy is further proposed to avoid parameter choosing and improve the convergence of this method.