Optimization of silicon solar cell fabrication based on neural network and genetic programming modeling

Abstract

This study describes techniques for the cascade modeling and the optimization that are required to conduct the simulator-based process optimization of solar cell fabrication. Two modeling approaches, neural networks and genetic programming, are employed to model the crucial relation for the consecutively connected two processes in solar cell fabrication. One model (Model 1) is used to map the five inputs (time, amount of nitrogen and DI water in surface texturing and temperature and time in emitter diffusion) to the two outputs (reflectance and sheet resistance) of the first process. The other model (Model 2) is used to connect the two inputs (reflectance and sheet resistance) to the one output (efficiency) of the second process. After modeling of the two processes, genetic algorithms and particle swarm optimization were applied to search for the optimal recipe. In the first optimization stage, we searched for the optimal reflectance and sheet resistance that can provide the best efficiency in the fabrication process. The optimized reflectance and sheet resistance found by the particle swarm optimization were better than those found by the genetic algorithm. In the second optimization stage, the five input parameters were searched by using the reflectance and sheet resistance values obtained in the first stage. The found five variables such as the texturing time, amount of nitrogen, DI water, diffusion time, and temperature are used as a recipe for the solar cell fabrication. The amount of nitrogen, DI water, and diffusion time in the optimized recipes showed considerable differences according to the modeling approaches. More importantly, repeated applications of particle swarm optimization yielded process conditions with smaller variations, implying greater consistency in recipe generation.

ANN-Model for Plasma Chemical Vapor Deposition Process

“…On the testing lathe in the laboratory, O2 was used to simulate the reaction gas, while plasma was generated. Via adjusting some control variables, the relationship of gas flow, furnace temperature, reflective microwave power, and pressure inside substrate-tube was observed and shown as The ANN-model is a feed-forward network with hidden layer [4]. Multiple layers of neurons with nonlinear transfer functions allow the network to fit the nonlinear relationships between input and output.…”

“…The ANN-model is trained with the available data shown in Fig. 2 [4,5]. According to this model and Eq.…”

Optimization of Solar Cell Production Lines Using Neural Networks and Genetic Algorithms

“…ML has the capacity to learn a wide variety of nonlinear patterns in high-dimensional datasets and can be used to build data-driven models of process intradependencies and interdependencies . ML-based techniques for PV manufacturing have been explored for solar cell material design, optimizing individual processes, and a combination of processes . It has also been explored with regard to DoE optimization, quality control, and troubleshooting with access to wafer tracking .…”