A Growth Prediction Model for Greenhouse Tomatoes Based on KNN-SVM Algorithm
In order to solve the problem of low prediction accuracy of the existing greenhouse tomato growth model, a non-linear model of greenhouse tomato growth and development was established based on the basic characteristics of tomato physiology, and the environmental pa-rameters in the greenhouse were used as model variables. This model described the influence of environmental factors, such as temperature, humidity, soil temperature and soil moisture in the greenhouse on the growth rate of tomato. The model had good explanatory ability and high precision. First of all, various sensors were used to collect various environmental data of tomato growth in the greenhouse of Jilin City, Jilin Province. Then, the actual data of the tomato greenhouse was preliminarily processed, and then the KNN algorithm was used to supplement missing and abnormal data, and correlation analysis was carried out. Finally, based on the processed tomato crop growth data, we considered the real-time feedback of tomato crops to the greenhouse environment. Combining with the correlation, we used the SVM optimization algorithm to analyze the greenhouse tomato data of Jilin Economic Development Zone from 2020 to 2021. After simulation, the accuracy rates of SVM, LDA and LR were 0.904, 0.885 and 0.865, respectively. The results showed that SVM could better predict the growth changes of tomato. The establishment of the greenhouse tomato crop-environment interaction model laid a good foundation for the greenhouse environment prediction control.
Greenhouse environmentEnvironmental monitoringKNN-SVMGrowth prediction model
万方数据
维普
