An immune system approach to scheduling in changing environments

摘要

This paper describes the application of an artificial immune system, (AIS), model to a scheduling application, in which sudden changes in the scheduling environment require the rapid production of new schedules. The model operates in two phases: In the first phase of the system, the immune system analogy, in conjunction with a genetic algorithm, (GA), is used to detect common paterns amongst scheduling sequences frequently used by a factory. In phase II, some of the combinatoric features of the natural immune system are modelled in order to use the detected patterns to produce new schedules, either from scratch or starting from a partially completed schedule. The results are compared to those calculated using an exhaustive search procedure to generate patterns. The AIS/GA analogy appears to be extremely promising, in that schedules corresponding to situations previously encountered can easily be reconstructed, and also in that the patterns are shown to incorporate sufficient information to potentially construct schedules for previously unencountered situations.