Common approaches to improving hybrid perovskite solar cell/thermoelectric systems optimize variables one at a time, overlooking how strongly light, electricity, and heat interact. In a recent study, researchers from Ningbo University and Ningbo University of Technology addressed this challenge by introducing a machine learning-driven global optimization method that captures these complex couplings.
Using a rigorously validated numerical model, the team simultaneously optimized six key factors, including light-side temperature, absorption layer thickness, and thermoelectric element density. The results included maximum energy efficiency of 20.0%, performance gains of +4.3% over non-optimized systems, +6.8% over standalone perovskite solar cells and precisely quantified values for layer thickness, semiconductor leg dimensions, and more.
This work shows that co-optimization across multiple variables is critical to unlocking the true performance of hybrid energy systems. This approach provides a fast, flexible, and cost-effective design strategy, adaptable to new materials and diverse operating conditions.
