HYBRID GENETIC SIMULATED ANNEALING APPROACH TO SHORT-TERM MULTIPLE-FUEL-CONSTRAINED GENERATION SCHEDULING

This paper develops a new formulation far short-term multiple-fuel-con strained generation scheduling. in the formulation, the power balance constraint, generator operation limits, fuel availability factors of g enerators, efficiency factors of fuels and the supply limits of fuels are taken fully into account. A fuzzy set approach is included in the formulation to find the fuel schedules, which meet the take-or-pay fue l consumption as closely as possible or maximise the utilisation of th e cheap fuels, within a generation schedule. The new formulation is co mbined with genetic algorithms, simulated-annealing and hybrid genetic /simulated-annealing optimisation methods to establish new algorithms for solving the problem. A method for forming the initial candidate so lutions in the genetic-based and hybrid-based algorithms is also devel oped. This method has also been incorporated into the simulated-anneal ing-based algorithm. The new algorithms are demonstrated by applying t hem to determine the most economical generation schedule for 25 genera tors in a local power system and its fuel schedule for 4 different typ es of fuels.