Assignment of Unexpected Tasks in Embedded System Design Process Using Genetic Programming

Embedded systems need to execute some tasks. The designer needed to predict every behaviour of the system. However the problem appears when system meets unexpected situation. In some cases the system architecture cannot be modified or such operation is too expensive. In this paper we present a novel developmental genetic programming based method for assignment of unexpected tasks in embedded system design process. Our approach does not modify the system architecture. The proposed method evolves decision trees. The new individuals are obtained during evolution process after using genetic operators: mutation, crossover, cloning and selection. After genotype to phenotype mapping the ready system is obtained.