Synergistic Hybridization of GP and DE: Innovations in Evolutionary Computation

Innovations in evolutionary computation often emerge from combining distinct optimization paradigms. This paper presents a novel hybrid approach: the synergistic integration of Genetic Programming (GP) and Differential Evolution (DE), referred to as GP*. In this framework, GP serves a dual role. It computes objective function values for DE while simultaneously generating candidate solutions in the form of symbolic expressions. We demonstrate this concept in the task of chaos stabilization using the Hénon map. The GP* algorithm enables efficient exploration of the solution space by taking advantage of the complementary strengths of GP and DE. Experimental results show that in some cases, GP* outperforms conventional two-phase strategies in which GP and DE operate independently, achieving better solution quality and faster convergence. Beyond chaotic system control, the GP* framework is applicable to broader classes of problems involving tightly coupled structural and parametric optimization. This work introduces a new perspective on algorithmic hybridization, where structural evolution directly defines the fitness landscape for parameter search. The results presented here contribute to the field of genetic programming and evolutionary optimization and open avenues for further exploration in interpretable modeling, controller synthesis, and dynamic system design.