Relative humidity prediction of northern greenhouse environmental factors on the basis of a radial basis function neural network
Chunling Chen1, 2, Long Wang2, Tongyu Xu2, Jiawei Qi2
1Key Laboratory of Protector Horticulture, Ministry of Education, Shenyang Agriculture University, 120 Dong Ling Road, Shenyang, China
2School of Information and Electrical Engineering, Shenyang Agriculture University, 120 Dong Ling Road, Shenyang, China
With its advantages of abundant resource, popularity, and efficiency, solar greenhouse is the only type of greenhouse that is widely used in Northern China. This study proposes a simulation prediction model that is based on a radial basis function artificial neural network. This model is suitable for dealing with humidity in northern solar greenhouses. We select 600 groups of training data to establish the network model and to verify its accuracy. We then randomly select 80 groups for validation. With a 7.35% average error rate, the prediction model shows satisfactory performance. Thus, the results can be used to predict the relative humidity curve in a greenhouse, as well as provide a scientific basis for reasonable regulation and control of a greenhouse environment.