Abstract:Based on our fifteen-year experience in numerical simulations of various solar cells, we have recently published an article about ten common simulation steps intrinsic to typical electrical and optical device simulations. We have developed an open-source framework, UniSolar, to streamline numerical simulations with commercial software packages like Sentaurus by just editing ten sets of respective variables that include all the necessary information for these steps. We have demonstrated that multidimensional optical and electrical simulations for heterojunction solar cells can be easily carried out without in-depth knowledge of Sentaurus. To further enhance the capability of UniSolar, typical process simulations are also integrated into our framework and process sequences are well represented by one set of variables. In this work, we will present multidimensional process, optical and electrical simulations of a TOPCon solar cell with 11 sets of variables. To accurately model the light trapping effect of random upright pyramids, 150 pyramids of dynamic base centers and heights are created within a surface area of $5^{\prime}5\ \mu \mathrm{m}2$ . To account for optical losses due to free-carrier absorption, process sequences to form a boron emitter of 90 Ω/sq on the planar surface are applied to the pyramidal surface to diffuse the corresponding emitter. The process sequences encompass three process steps: boron-doped oxide deposition, boron diffusion process, and oxide strip. Reasonable agreements for quantum efficiency and light current density - voltage measurements are achieved by employing state-of-the-art models and tuning interface recombination parameters. This work indicates that multidimensional processes and optical and electrical simulations are easily accessible due to the comprehensive framework design in UniSolar. Additionally, fitting measurement curves is also easy by appending multiple shorthand representations to a variable. Our powerful framework embraces global collaboration within the photovoltaic community and could be used to generate an extensive dataset for low-cost artificial intelligence training.

Bayesian Optimization with Experience for Fast Development of Monolithic Tandem Solar Cells: Simulation Case Study

Autonomous Optimization of an Organic Solar Cell in a 4-dimensional Parameter Space

Bayesian Reverse Design of High-Efficiency Perovskite Solar Cells Based on Experimental Knowledge Constraints

Performance prediction and optimization of perovskite solar cells based on the Bayesian approach

A Combined Numerical Modeling and Machine Learning Approach for Optimization of Mass-Produced Industrial Solar Cells

Performance Optimization of Photovoltaic and Solar Cells Via a Hybrid and Efficient Chimp Algorithm

Improving the efficiency and stability of perovskite solar cell: Application of innovative machine learning algorithm

Physics-guided characterization and optimization of solar cells using surrogate machine learning model

Enhancing Tandem Solar Cell's Efficiency through Convolutional Neural Network-Based Optimization of Metasurfaces

Machine Learning Assisted Optimization of Perovskite Thin Film Fabrication Process and Assessment of Feature Importance

Utilizing machine learning algorithm in predicting the power conversion efficiency limit of a monolithically perovskites/silicon tandem structure

Embedding Physics Domain Knowledge into a Bayesian Network Enables Layer-by-Layer Process Innovation for Photovoltaics

Molecular design and performance improvement in organic solar cells guided by high‐throughput screening and machine learning

Performance Optimization of Monolithic All-Perovskite Tandem Solar Cells under Standard and Real-World Solar Spectra

Design of Experiments with the Support of Machine Learning for Process Parameter Optimization of All‐small‐molecule Organic Solar Cells

Photovoltaic Cells Parameter Estimation Using an Enhanced Teaching–Learning-Based Optimization Algorithm

Accelerated exploration of efficient ternary solar cells with PTB7:PC<sub>71</sub>BM:SMPV1 using machine-learning methods

Easy Access to Typical Multidimensional Process/Optical/Electrical Simulations for Solar Cells by Tuning 11 Variables

Device Simulation–Based Multiobjective Evolutionary Algorithm for Process Optimization of Semiconductor Solar Cells

The Spring of Processing Chemistry in Perovskite Solar Cells-Bayesian Optimization.

Machine learning for accelerating the discovery of high-performance donor/acceptor pairs in non-fullerene organic solar cells