Gradient-based immune algorithm for optimization of dynamic environments

A novel immune algorithm suitable for dynamic environments (GIDE) is proposed based on a biological immune mechanism. GIDE models the dynamic process of artificial immune response with gradient-based diversity operators. Unlike traditional artificial immune algorithms, which require that randomly generated cells be added to the current population to explore its fitness landscape, GIDE uses a gradient-based diversity operator to speed up optimization in dynamic environments. Other immune algorithms are compared to GIDE by using Moving Peaks Benchmarks. Preliminary experiments showed that GIDE can maintain high population diversity during the search process, while simultaneously speeding up optimization. Thus, GIDE is useful for optimization of dynamic environments.