Maintenance Outage Scheduling Considering Covariates In Power Systems

The factors influencing the operation of components in power system,such as equipment health condition,weather condition,etc.,are defined as covariates.The current models are not able to quantify the effect of covariates on equipment reliability,criticality and maintenance scheduling.The dissertation focuses on the evaluation of components reliability and criticality,and the maintenance scheduling in power system.The main contributions are presented below.The equipment is subject to service age and health conditions.The forced outage rate(FOR)is modeled by proportional hazard model(PHM).Two reliability indices are calculated:survival function and mean residual life.The health condition is under periodical monitoring,however,the health condition transition can happen at any time.Analytical method and Monte Carlo method are respectively presented to evaluate the two reliability indices.In numerical studies,the reliability results of different models are illustrated.Moreover,the advantages and disadvantages between analytical method and Monte Carlo method are discussed.An integrated framework is proposed to identify components criticality in reconfigurable distribution systems.The model considers covariates such as prevailing severe weather conditions and component aging conditions,common mode outages,and repair rates in distribution systems.The effects of covariates on FOR of components are quantified by the PHM.A recursive sampling method is proposed to generate Monte Carlo scenarios with the given forced outage rate.In each scenario,the distribution system reconfiguration is optimized efficiently using a standard mixed-integer second-order cone program.Then absolute and relative criticality indices are calculated to describe the criticality of components.A new scenario is generated subsequently and the process is continued in the following iteration until the convergence condition is satisfied.The final absolute and relative criticality indices are the average over multiple scenarios.The influence of component aging condition and outage characteristics,weather condition,and the system reconfiguration on component criticality is distinguished and exhibited in case studies.The case studies illustrate the effectiveness of proposed model and the solution method.A three layers nested framework is proposed which coordinates short-term generation and transmission maintenance scheduling with midterm maintenance decisions by considering the effects of short-term security-constrained unit commitment(SCUC).The FOR is modeled by PHM,and a recursive sampling method is introduced in the proposed Monte Carlo-based framework for generating scenarios.The outer loop is Monte Carlo iteration.For each Monte Carlo iteration,an iterative dynamic scenario updating approach is introduced to consider interactions among covariate conditions,random component outages,and maintenance outage scheduling.The co-optimization problem is decoupled into three separate optimization subproblems by Lagrangian relaxation(LR).Case studies on the 6-bus system and the IEEE 118-bus system are used to exhibit the effectiveness of proposed framework.