A linear mixed-integer programming approach for the unit commitment problem

Abstract EN

In this paper, we consider the unit commitment problem that is concerned with the shortterm (hourly) scheduling of thermal-electric generators in order to satisfy intermediateload demand.

The objective is to minimize total fixed and start-up costs, subject todemand, spinning reserve, capacity, and various operational constraints such as minimum uptime and downtime requirements, and maximal changes in output from one time period to the next (ramping constraints).

In essence, the problem is a nonlinear discrete optimization program.

In contrast with previous implicit formulations accompanied by specialized decomposition approaches, we provide an explicit linear mixed-integer programming formulation for this problem that is capable of easily accommodating several types of complicating side-constraints, and is amenable to standard commercial software application.

In particular, we present some encouraging computational results to demonstrate the efficacy of using CPLEX-MIP (version 6.5) in handling this new formulation of the problem.