论文部分内容阅读
实际工业过程都具有非线性等特征。传统的监控方法有将降维后的非线性数据映射到高维线性空间再进行数据处理,实现过程的监控。本文是在一种否定选择算法的基础上,首先利用最大方差展开(MVU)方法对正常高维数据进行降维,再利用否定选择算法直接对降维后的多维非线性数据建立“超球体群”模型,实现对过程的监控,保证工业过程的平稳运行。仿真实验是基于TE模型进行的,仿真结果表明该方法较传统方法及其他改进方法具有更好的监控能力,说明了该方法的有效性。
The actual industrial processes are non-linear and other characteristics. The traditional methods of monitoring the dimensionality of the non-linear data will be mapped to high-dimensional linear space and then data processing to achieve the process of monitoring. On the basis of a negative selection algorithm, this paper first reduces the dimension of normal high-dimensional data by using the Maximal Variance Expansion (MVU) method, and then uses the negative selection algorithm to establish the hyper-sphere Group "model, to achieve the process of monitoring, to ensure the smooth operation of industrial processes. The simulation experiment is based on the TE model. The simulation results show that the method has better monitoring ability than the traditional method and other improved methods, and shows the effectiveness of the method.