Abstract: This paper analyzed the importance of temperature control in the process of air－conditioner running and the shortcomings of traditional methods,used Support Vector Machine（SVM）,which based on Statistic Learning Theory（SLT）,to forecast future temperature values in the process,adopted numerical experiments to verify the feasibility of this method,and the experiment results suggested that it has higher precision and superior real time performance than traditional methods.This method can give better decision information to the temperature control system of air－conditioner.It seemed that SVM method has some applied value and a hopeful future in this area．

摘要： 分析了空调工作过程中温度控制的重要性与传统方法的不足,将基于统计学习理论的支持向量机方法用于控温过程中,对温度进行实时预测。采用数字实验对所提方法的可行性进行验证,结果表明该方法可以获得比传统方法更高的预测精度,实时性较高,能为空调控温系统提供更好的决策支持,具有较大的发展潜力和实用价值。