Designing Polymorphic Circuits with Periodical Weight Adjustment

A polymorphic circuit can perform two or more functions under different conditions without the need of extra components. Those functions can be activated by environmental signals, such as temperature, power supply voltage, illumination, and so on. So far, existent evolutionary algorithms can only evolve polymorphic circuits of small scales. In this paper, the periodical weight adjustment method is proposed for the Evolutionary Strategy to evolve polymorphic circuits. The experimental results demonstrate that the Evolutionary Strategy with the periodical weight adjustment can perform better than existent evolutionary techniques for polymorphic circuits is evolved for the first time.