Font Size: a A A

Risk Analysis And Corrective Control Of Power System Cascading Failures

Posted on:2023-04-30Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y M LiuFull Text:PDF
GTID:1522306902972099Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
Modern power systems are indispensable energy industries to support social life and its reliable operation is essential to interdependent critical infrastructure,such as transportation or communication systems.While enjoying convenience and comfort brought by electricity,protection of the power system against catastrophic accidents has been an unavoidable challenge for researchers.In recent years,several blackouts caused by cascading failures have severely affected production and daily life.Meantime,with tighter electrical coupling within internal power networks,the influence scope and propagation speed of cascading failures are further expanded.Therefore,it is necessary to assess the risk of power system blackout caused by cascading failure,and identify vulnerable components that trigger or accelerate the propagation of cascading failures.Then,based on this,control research on cascading failures has the important value for the risk management of cascading failures and security operation of the power system.Focusing on the core problem of power system cascading failure defense,this thesis systematically studies several key issues,including the risk assessment of system blackouts,identification of vulnerable components and cascading failure control.The main work of this thesis is summarized as follows.(1)A blackout risk analysis method based on the balance structure of the power system operating state is proposed.For assessing the risk of power system blackout caused by cascading failure,considering a marked difference in importance of branches and coupling effect between branches on the system dynamic behavior,the transmission branches are represented as virtual nodes and the power flow interactive relations between the transmission branches as virtual links,and then a power system associated network with weight and direction is constructed,where importance degree of branches is used as node weight and link weight is the link sign characterizing the equilibrium of system power flow distribution.Based on the structural equilibrium theory of signed networks,the balance structure pattern and classification of power system operating states is established and analyzed.On this basis,a comprehensive of power network imbalance evaluation index is proposed to represent the power flow distribution characteristic and quantify power system disturbance dissolving ability,which can assess system blackout risk.(2)Vulnerable branches identification method considering the propagation characteristics of cascading failures is proposed.To identify vulnerable components that trigger or expand the propagation of cascading failures,fault chain model of power system was constructed by selecting branch with maximum outage probability among all the overloaded branches as the outage branch at the next cascading step during cascading failure forecasting.The cascading failure process sets generated by fault chain model were used to construct a weighted cascading faults graph by merging the repeated components in different cascading fault processes and cumulating the corresponding weight values between components.The cascading failure process sets is compressed into a directed and weighted cascading failure graph.And cascading failure graph was transformed into the cascading fault space-time graph(CFSTG)according to the community characteristics of the grid structure.In terms of vulnerability origin and whether branch can promote the cascading failure process to spread across regions,the vulnerable branches were divided into four types and the indices of node degree in CFSTG were employed to recognize vulnerable branches.Various vulnerable branches are verified by timing attack and the index of cascading fault propagation was proposed to quantify the impact of vulnerable branches on the spread of cascading failures.The proposed method can accurately identify the vulnerable branches that trigger or expand the propagation of cascading failures and distinguish inducing factors of different cascading steps.(3)Multi-step blocking control method based on path-driven is proposed for mitigation of cascading failures.Because cascading failure caused by overloading is a multi-step propagation process,traditional control method(i.e.,a specified cascading step)severely limits the available variables in the control model and may provide an inflexible control scheme with which to correct cascading failure at high cost.For the whole cascading failure process,a multi-step blocking control model considering flexibility of control timing is proposed.The propagation mechanism of cascading failure caused by overload is briefly described,and the dynamic interaction between cascading failure propagation paths and blocking control strategies is analyzed.For the controllable cascade steps,under the branch outage screening criteria,the outage probability constraints of the failures/non-failures path are extracted.The proposed model with flexibility of control timing is designed by optimizing probability weighting expected cost and overload risk of cascading failure.In this control method,generation re-scheduling and load shedding are selected as the main remedial actions.Control schemes generated by the proposed control method could mitigate the cascading failures and result in good performance in reducing the risk of power system blackout caused by cascading failure.(4)A multi-step cascading failure control based on dynamic evolutionary is proposed.To further improve the efficiency of cascading failure control,under extreme operating conditions,such as the lower dispatchable capacity of the generator outputs,multi-step blocking control method based on path-driven cannot provide an effective control scheme to eliminate cascading outages.For this problem,the dynamic evolutionary model is optimized to reduce the risk of cascading failures caused by tripped branch.A multi-layer coding genetic algorithm is employed to obtain the optimal remedial control schemes.The first half of a chromosome is an integer string to code the cascade failure path and the second half is a real number string used to record the cascading outage probabilities.Compared with the multi-step blocking control based on path-driven,this method has better economy and stronger applicability.
Keywords/Search Tags:Cascading failure, risk of power system blackout, vulnerable branches, multi-step blocking control
PDF Full Text Request
Related items