Adaptive Balancing of a Bank of Sugar - Beet Presses Using a Genetic Algorithm with Variable Local Search Range

The paper describes an application of a genetic algorithm with a variable local search operator for load balancing in energy-intensive processing of sugar beet. The variable local search operator enables the genetic algorithm to track optima of the time-varying dynamic systems. It constitutes an adaptive tracking mechanism as it is triggered only when the running average of the best performing members of the population deteriorates. Results obtained on simulation suggest that the genetic algorithm-based optimiser can significantly improve energy efficiency of the real press station. The technique is believed to be a feasible engineering solution to a general load balancing problem between a number of parallel processing units with a measure of performance which reflects only an overall effectiveness of the system controlled and not that of individual processing elements. 1 Introduction In this paper we report on an application of a genetic algorithm (GA) with a variable local search o..