Modeling Of The Chrysanthemum Plant Tissue Culture Growth Based On Genetic Algorithm And Neural Network

China is an agricultural country with a large population. How to increase the production and productivity of agricultural is a big challenge. The simulation model with greenhouse is a powerful tool to get a optimization of f growth management and environmental regulation of greenhouse crop. Therefore, the simulation model of greenhouse crop growth is important to realize the greenhouse crop production and environmental control, and to improve the economic and ecological efficiency of greenhouse crop production of China.In this study, the experiment of sugar-free clone plantlets under different nutrient culture media types, temperature, light intensity, relative humidity, and CO2 concentration was carried on by the large space training device in industrialized agriculture research center of Shandong University of Technology from 2008 to 2010. Based on physiological and ecological processes of chrysanthemum growing in greenhouse, a simulation model of dry matter in greenhouse chrysanthemum was established through systematic analysis of experimental data.However, in practical application, neural network convergence is slow. For large search space, multi-peak and non-differentiable function, the network is vulnerable to falling into local extreme point. On the other hand, initial weights of neural network, thresholds, and network structure is with great random because of no basis for the choice. It is difficult to select the initial point with the overall situation, and therefore to find the global optimum is less likely. All these affected the neural network generalization. For the lack of neural networks, we use a new search algorithm-Genetic Algorithm (Genetic Algorithms called GA) for optimization of neural networks to select the air temperature, relative humidity, light intensity, CO2 concentration and leaf area as the main factors in the training device. Using the sugar-free chrysanthemum plantlets as the training and testing samples to model and simulate plant dry weigh. The results show that the method for tissue culture growth modeling is suitable for the graphics simulation of clone plantlets. The forecast data u can be verify the feasibility and accuracy of the model for the further in-depth study.