基于智能蚂蚁算法的脱硫静态模型优化

为了自动寻找脱硫过程的规律和知识,对脱硫过程进行决策支持,采用RBF神经网络作为建模工具,针对建模过程中出现的RBF中心和宽度难以确定的难点,在分析蚂蚁算法机理的基础上,提出了使用智能蚂蚁算法对RBF神经网络模型的中心和宽度进行自适应选择,从而达到模型训练精度和范化能力的一个最优的平衡,进而提高模型的预报精度;在分析脱硫工艺原理的基础上,通过有效的数据预处理,最后进行仿真分析,其模型的预报精度好于传统脱硫静态模型,具有一定的实用性和推广价值。

Static Model Optimization for Desulphurization Based on Intelligent Ant Algorithm

In order to search for the law and knowledge of desulphurization process to make decision support for the process,Radial Based Function（RBF） neural network is used as modeling tool.According to the difficulty in determining RBF center and width in the process of modeling,based on the analysis of ant algorithm mechanism,this paper presents that intelligent ant algorithm can be used to make self-adaptation selection for the center and width of RBF neural network model so as to reach optimal balance between the training accuracy and generalization and further to increase predication accuracy of the model.On the basis of analyzing the principle of desulphurization technology and by effective data preprocessing,simulation analysis is conducted,the forecasting accuracy of the model is better than traditional static model for desulphurization,and this model has certain practicability and popularization value.