Multi-objective Optimization of Thermal Control Strategies for Multifunctional Structures

A thermomechanical electronic multifunctional structure prototype has been modeled and optimized. The model focuses on the description of thermal and electrical phenomena, but leaves aside structural issues. It couples a three-dimensional thermal network with representations of different possible thermal control laws, namely on/ off control, proportional logic, proportional-integral-derivative strategy, and the usage of positive temperature coefficient heaters. The parametric model was first validated and correlated through a comparison with simple physical solutions, and then with the actual results of a thermal vacuum test. Multiobjective optimization (based on genetic algorithms) has been used to define the best heater layout options, to identify the best control strategy in terms both of panel isothermia and energy consumption, and to fine-tune the parameters of the selected control strategy. The research reported in this paper has led to the definition of an optimal thermal control solution. An examination of the optimization results has shown that the simultaneous adjustment of the geometrical layout as well as the control strategy and its parameters can lead to energy savings of about 52%.