【作者】 段滈杉； 乔俊飞； 蒙西； 汤健；

【Author】 DUAN Hao-Shan;QIAO Jun-Fei;MENG Xi;TANG Jian;Faculty of Information Technology, Beijing University of Technology;Beijing Key Laboratory of Computational Intelligence and Intelligent System;

【机构】 北京工业大学信息学部； 计算智能与智能系统北京市重点实验室；

【摘要】 城市固体废物焚烧是当前固废处理的主要技术之一,对MSWI中产生的有毒气体——氮氧化物(NOx)的实时检测能有效控制NOx的排放。工业现场中采用高精密仪器—烟气排放连续监测系统对烟气排放中的NOx浓度进行检测,测量结果受环境影响较大,且设备维护成本高。针对上述问题,文中提出了一种基于模块化神经网络的MSWI过程NOx软测量方法。首先,利用模糊c-均值算法进行任务分解,将任务分解成不同的子任务;其次,针对不同的子任务,采用径向基函数神经网络分别设计软测量子模型,建立特征变量与NOx间的非线性关系;最后通过级联神经网络对子网络输出进行集成。采用基准实验和某MSWI厂实际数据验证了提出方法的有效性。更多还原

【Abstract】 Municipal solid waste incineration is one of the main technologies of current solid waste treatment. The real-time detection of the toxic gas-nitrogen oxides(NOx) generated in MSWI can effectively control the emission of NOx. In the industrial field, instruments are mainly used for spot-type detection of NOx concentration in flue gas emissions. The measurement results are greatly affected by the environment, and equipment maintenance costs are high. In response to the above problems, a soft measurement method for MSWI process based on modular neural network is proposed in this paper. Firstly, use the fuzzy c-means algorithm to decompose the task into different sub-tasks; secondly, for different sub-tasks, the radial basis function neural network is used to design the soft measurement sub-models respectively, establishing the nonlinear relationship between characteristic variables and NOx; Finally, the output of the sub-network is integrated through the cascade neural network. The effectiveness of the proposed method was verified by benchmark experiments and actual data from a MSWI plant.更多还原