基于支持向量回归的污泥发热量预测模型设计

Prediction model design of sludge calorific value based on support vector regression

针对污泥发热量测定过程中出现的设备腐蚀问题,以污泥的工业分析数据为基础,分别选择线性核函数、多项式核函数及径向基核函数建立基于支持向量回归的污泥发热量预测模型,并比较了不同模型的预测效果及评价参数。结果表明,基于径向基核函数建立的污泥发热量预测模型预测效果较好,其预测结果偏差值在允许范围内,可以满足对污泥发热量进行估算的要求。

In view of the equipment corrosion problem in the process of sludge calorific value determination, based on the industrial analysis data of sludge, polynomial kernel function and radial basis function were selected to establish the sludge calorific value prediction model based on support vector regression, and the prediction effects and evaluation parameters of different models were compared.The results showed that the prediction effect of the sludge calorific value prediction model based on radial basis function is better.The deviation of the prediction results is within the allowable range, which can meet the requirements of sludge calorific value estimation.