Simulation-based optimization of a solar water heating system by a hybrid genetic-binary search algorithm

A proper design of a large solar heating system is important to maximize the benefit of the system. The system hydraulics, control parameters and dimensions of single components are usually tried to be optimized towards achieving better system performance at lower costs. The complexity of the target functions, a large number of optimization parameters and boundary conditions imposed on the system require application of advanced numerical optimization techniques and tools as well as additional software. In this paper, a hybrid genetic algorithm is proposed and applied to optimization of a solar combisystem. The hybrid algorithm couples the CHC genetic algorithm with the binary (n-ary) search method. The results of the optimizations show that the proposed algorithm is almost two times faster than the pure CHC genetic algorithm and, in separate cases, more reliable in finding the global optimum. The parallel version of the algorithm was implemented in GenOpt (generic optimization software) and applied to optimization of a solar combisystem modelled and simulated with the simulation software TRNSYS. Results of the optimization are presented in this paper.