Abstract
Planned maintenance scheduling is a major focus area for power utilities as they strive to ensure continuous and reliable power supply to their customers. As businesses, they also aim to achieve this at the lowest cost. The complexity of power systems is increasing as new generating units are added to power systems in order to supply power to the growing economies. This has culminated into the generator maintenance scheduling (GMS) problem which seeks to ensure that generators are taken offline for maintenance optimally before they breakdown in a way that enables the power system to run cost effectively and reliably. This research is focused on developing a generator maintenance schedule using a tri-objective model. Three objective functions are combined in order to give a robust model. The first objective is reliability which seeks to create an even reserve margin over the planning period by minimizing the sum of squared reserves. The second objective is minimizing the cost of power production associated with a maintenance schedule. In this research the production cost is taken as just the fuel cost since it makes up the greatest component of the cost. The third objective is minimizing the risk of a generator failing before goes for maintenance. In essence, the third objective seeks to minimize breakdowns of generating units leading to unplanned outages. Two solution methods are utilized in modelling and solving the generator maintenance scheduling problem. The first is an exact solution method using mathematical modelling software, Advanced Interactive Multi-dimensional Modelling System (AIMMS). The second solution method is a recently developed metaheuristic algorithm called Exchange Market Algorithm (EMA). An investigation of how effective the two solution methods are in solving a case study that is common in literature is done. They are then used to build the tri-objective model and solve for a maintenance schedule. It is found that compared to each other, the performance of the solution methods is determined by the size of the problem. Advanced Interactive Multi-dimensional Modelling System (AIMMS) gives the best solution for the smaller optimization problem while Exchange Market Algorithm (EMA) gives a better solution for the larger optimization problem. The tri-objective generator maintenance scheduling model finds a trade-off schedule for the individual objectives.
M.Ing. (Electrical and Electronic Engineering)