A New Manner of Crossing in the Genetic Algorithm for Resolving Job Shop Problem (JSP)

The minimization of the makespan of the job shop problem with J jobs and M machines is NP-hard problem. To resolve it we apply a genetic algorithm [1,2]. We use a real coding for the representation of chromosomes. The originality of our variant lies by the choice of two effective crossover operators and one for mutation, As well as their respective probability, determined by numeric simulations on several examples known in the literature. We compare our results on ''OR library'' benchmarks [3] with Adamas and al. [4], Ombuki and al [5] and Yamada and al [6] results.

A Probabilistic Approach to Performance Forecasting for FMS as Part of an Automated Factory

Specific features of a machining FMS operation in a complex production environment (a machine shop or an automated factory) are considered. A probabilistic approach to FMS performance estimation, ensuring a steady production rhythm of an automated assembly factory, is suggested. A probabilistic description of FMS performance estimation based on the concept of production potential and its respective probability of achieving the goal value with due account of the reliability of FMS production facilities and various organization factors is given. A complex optimality criterion of FMS production potential in an FA environment is described. A method of machining FMS production tasks planning in terms of time and volume of output is suggested. Application of the suggested method is illustrated by the example of designing a machine-tool automated factory.