Multi-objective inverse design of finned heat sink system with physics-informed neural networks

This study proposes a new inverse design method that utilizes a physics-informed neural network (PINN) to parameterize the geometric and operating inputs, enabling the identification of optimal heat sink designs by starting with the desired objectives and working backward. A specialized hybrid PINN is designed to accurately approximate the governing equations of the conjugate heat transfer processes. On this basis, a surrogate model derived from the hybrid PINN is constructed and integrated with multi-objective optimization and decisionmaking algorithms. The results of an example finned heat sink system are presented, showcasing the accelerated search for Pareto-optimal designs. The proposed method nearly halved the search time to approximately 113.9 h in comparison with the traditional methods. Moreover, three representative scenarios-high-performance design, equilibrium design, and low-cost design -were compared to visualize the real-time changes in the multiphysics field, facilitating improved physical inspection and understanding of the optimal designs.

Automated summary

This study proposes a new inverse design method using a physics-informed neural network to identify optimal heat sink designs. A hybrid PINN accurately approximates the governing equations of heat transfer processes, and a surrogate model is constructed for integration with optimization algorithms. The proposed method accelerates the search for Pareto-optimal designs and reduces search time. Comparing different scenarios facilitates real-time observation of multiphysics field changes, improving understanding of optimal designs.