Carrousel oxidation ditch has been widely used in sewage disposal system, however the result of sewage disposal was affected by water quality and environmental factors, so it is difficult to build a precise prediction model. The existing method of machine learning algorithm usually gets a poor result in prediction. In order to precisely predict the result, this essay uses the multi-kernel wavelet support vector machine when build a model. The outcome of this experiment demonstrates that the new method improve the degree of definition in forecasting, and it is suitable for actually online prediction.%Carrousel氧化沟广泛应用于城市污水处理,但污水处理的效果受到水质和环境因素影响很大,难以建立精确的预测模型。现有的机器学习方法普遍预测效果较差,为了准确预报污水处理的效果,本文采用多核小波支持向量机进行建模,实验表明该方法提高了预报的精确度,适合用于氧化沟系统的实时在线预测。
展开▼