生物氧化预处理过程中不确定性数据处理

发布时间：2018-06-10 07:19

本文选题：鲁棒支持向量回归 + 二阶锥规划　；参考：《计算机工程与设计》2017年07期

【摘要】：氧化槽进气量是生物氧化预处理过程中重要的操作调控参数,由于氧化槽内成分复杂,预处理过程具有强非线性、耦合性和大时滞性的特点,采集到的数据往往包含一定的不确定性。针对这一问题,提出一种基于机会约束的鲁棒支持向量回归方法。通过将支持向量回归模型中常规约束转换成机会约束,得到相应的鲁棒回归模型,将鲁棒回归模型转化为求解二阶锥规划问题。仿真结果表明,所提算法能够有效对数据不确定性进行处理,为生物氧化预处理过程提供新的方法。
[Abstract]:The air intake of the oxidation tank is an important operating control parameter in the biological oxidation pretreatment process. Because of the complex components in the oxidation tank, the pretreatment process has the characteristics of strong nonlinearity, coupling and large time delay. The collected data often contain some uncertainty. To solve this problem, a robust support vector regression method based on opportunistic constraints is proposed. By converting the normal constraints in the support vector regression model into chance constraints, the corresponding robust regression model is obtained, and the robust regression model is transformed into a second-order cone programming problem. Simulation results show that the proposed algorithm can effectively deal with data uncertainty and provide a new method for biological oxidation pretreatment process.
【作者单位】：新疆大学电气工程学院;
【基金】：国家自然科学基金项目(61463047)
【分类号】：TP18;TQ033

【相似文献】