Lighting Design Optimization in a Plant Factory Using an Optical Simulation and Differential Evolution

The light environment in a plant factory with artificial light (PFAL) can be determined based on the optical properties of the plants or structural components and lighting design (LD). Although an optimal LD can create the desired light environment, determining it is difficult because it is a complex problem with multiple variables, objectives, and constraints. In this study, we aimed to search for optimal LD using the ray-tracing method for optical simulations and ε-constrained multi-objective differential evolution for optimization. First, the LD of a cultivation system was optimized, and komatsuna plants were cultivated under the optimal LD. The correlation coefficients between the objective functions and corresponding indices calculated from the komatsuna dry weights were larger than 0.75. Therefore, an LD suitable for a specific cultivation purpose can be obtained through optimization. Thereafter, the LD of a virtual cultivation system was optimized. The photosynthetic rate and its uniformity showed a broadly positive correlation, whereas electric energy use efficiency showed trade-off relationships with other objective functions. Therefore, the threshold of electric energy use efficiency could serve as an indicator for selecting a suitable LD. These results demonstrate that LD optimization can be used to suggest optimal LD with concrete values for practical application.