A photosynthetic rate prediction model using improved RBF neural network

Area of Science:

• Plant Physiology
• Computational Biology

Background:

• Existing photosynthetic rate prediction models struggle with low speed and accuracy.
• Accurate modeling is crucial for effective light environmental regulation in agriculture.

Purpose of the Study:

• To analyze the effects of light quality on photosynthesis rate.
• To develop a novel, highly accurate photosynthetic rate prediction model.
• To optimize model performance using a Quantum Genetic Algorithm (QGA).

Main Methods:

• Utilized "golden embryo formula 98-1F1" cucumber seedlings for experiments.
• Employed LI-6800 to measure photosynthetic rates under varied light quality, intensity, and temperature.
• Developed a prediction model based on Radial Basis Function (RBF) optimized by QGA.

Main Results:

• The QGA-RBF model achieved a determinant coefficient of 0.996.
• Linear fitting showed a slope of 1.000 and an intercept of 0.061.
• The proposed model demonstrated superior accuracy compared to six other artificial intelligence algorithms.

Conclusions:

• The QGA-optimized RBF model significantly improves photosynthetic rate prediction accuracy.
• This enhanced model offers a robust theoretical basis for light environmental regulation.
• The findings provide a more efficient tool for agricultural applications involving light management.