温室环境的支持向量机回归建模

SVM Regression Modeling for Greenhouse Environment

温室气候是一个复杂的动态系统 ,传统的建模方法很难建立精确有效的温室气候模型。本文引入一种支持向量机回归建模方法来建立温室气候模型 ,这种方法根据历史数据建立气候模型 ,并用当前数据检验修正模型。对实际温室数据进行了建模实验 ,取得了较好的效果。

Greenhouse weather is a complex dynamic system, whose mathematical model usually is not precious and even invalid if built by the classical modeling methods. Therefore, a modeling method by using support vector machines (SVM) regression is presented. The models of greenhouse weather were established according to the history data, and then were modified and corrected on-line using the current data. A model for a practical greenhouse was built by the method and the results from the model were tested. The method was proved to be good in modeling of a greenhouse.