Smart Grid Dispatch Optimization For System Resilience Enhancement

The frequency and intensity of natural disasters have greatly increased over the past decades,and they are expected to further increase in the coming future.As a consequence,it becomes a criti-cal issue worldwide to deal with the disasters and their impacts.Under this background,the concept of resilience was introduced,and resilience-oriented dispatch soon takes the place of traditional dis-aster reduction methods for power grids.However,we note that current practice and research are still in the exploratory stage,and neither basic concepts nor fundamental principles are well estab-lished.In particular,there is yet no sufficient research in the field of resilience-oriented power grid dispatch,and there is a general lack of decision-making methods for system resilience enhance-ment.To close this gap,this dissertation clarifies the definition and quantification of resilience and presents three fundamental principles of resilience enhancement.Moreover,this dissertation puts forward an integrated preventive-emergency response for power systems,an integrated gas-electricity dispatch for coupled energy systems,and a cyber-constrained optimal power flow model for cyber-physical systems.The contributions of this dissertation are as follows:Firstly,the basic concepts of resilience,including definitions and quantifications,are carefully investigated here.We analyze the basic model of power grid disaster defense,and put forward three fundamental principles of how to enhance the system resilience for smart grid.These three principles correspond to traditional power systems,coupled energy systems,and cyber-physical systems,respectively.They constitute a three-dimensional framework,which can be used to guide the system resilience enhancement of smart grid.Secondly,an integrated method of preventive and emergency responses for power grid re-silience enhancement is put forward.Focusing on the most basic components of grids,we here propose an integrated resilience response framework,which not only links the situational aware-ness with resilience enhancement,but also provides effective and efficient responses in both pre-ventive and emergency states.Preventive response here considers generator re-dispatch and topol-ogy switching,while emergency response includes generator re-dispatch,topology switching and load shedding.The core of the proposed framework is a two-stage robust mixed-integer optimiza-tion model,which can be easily extended for other dispatch strategies and solved by the proposed nested C&CG algorithm.This dissertation verifies the potential of topology switching and more importantly,the superiority of the proposed integrated resilience response.Then,a joint dispatch method of coupled gas-electric energy systems is put forward.As mod-ern energy systems have evolved into a complex network of coupled systems and the role of gas-electric energy systems is so outstanding,we here model the operation of gas-fired generators in the electricity network and electricity-driven compressors in the gas network,and propose a joint dis-patch method of coupled gas-electric energy systems.We compare the proposed dispatch method to the traditional independent dispatch model,corresponding to perfect global visibility and no global visibility,respectively.To deal with the non-convex Weymouth equations,a sequential second-order cone programming algorithm is designed to solve the model.Through this work,we report a phenomenon we term as "cascading imbalance",which is an emerging type of cascading effect that has never been reported before.As the proposed joint dispatch strategy can avoid the above cascading effect,it will lead to lower loss not only in total cost but also in individual energy supply.Lastly,a cyber-constrained optimal power flow model for cyber-physical power systems is put forward.Here we model the power network and the cyber network explicitly,and take into account not only the dependence of power networks on cyber networks but also the dependence of cyber networks on power networks.With the introduction of cyber networks,however,it is very difficult to directly solve the cyber-constrained optimal power flow model because of the highly non-linear constraints.What is worse,the curse of dimensionality makes it hard to model even intermediate cyber-physical systems explicitly.As a countermeasure,we in this dissertation propose an extended maximum flow method to obtain the optimal solution.This work highlights the importance of the proposed model for smart grid resilience enhancement and will also provide insights for other interdependent systems that have bidirectional interactions.