Optimal Design and Process of Threshing Units Based on a Genetic Algorithm. II. Application

This paper illustrates the application of an optimization technique based on a multiobjective genetic algorithm (Miu, 2001), for the design and operation parameters of an axial threshing unit. The genetic algorithm includes float-point representation of the design and functional parameters as well as crop properties that are subject to optimization. The population has 100 individuals (potential solutions) and its size is constant throughout the search process. Classic genetic operators (mutation and crossover) are applied. Improvement of the algorithm performance was obtained using elitist selection strategy. Effects of equal and selective weights in the evaluation function are analysed. Performance of the optimized model is compared with industry design obtained with classical approaches and the improvement in performance is emphasized.