Designing Solar Chimney Power Plant Using Meta-modeling, Multi-objective Optimization, and Innovization

Summary

Proposes a design approach for solar chimney power plants in which a costly simulator is replaced by a two-objective surrogate built with genetic programming, checked against neural network and regression alternatives, then optimized with a genetic multi-objective algorithm. Design rules are extracted from the resulting Pareto front, and the surrogate models stay within about 4% error relative to the simulator.

Description written by Solar Tower. The original abstract is available from the publisher via the link below.