【作者】 孙颖； 张袅娜； 王玉华； 卢秀和； 杨洪侠；

【Author】 Sun Ying Zhang Niaona Wang Yuhua Lu Xiuhe Yang Hongxia (Changchun University of Technology Changchun 130012 China)

【机构】 长春工业大学；

【摘要】 针对大孤山选矿厂的磁选工艺过程的多指标、强耦合、时变、非线性和大滞后等特点,使得基于数学模型的常规优化方法难以应用,本文忽略磁选过程具体的物理化学作用,采用用于大工业过程建模的多输入层小波神经网络建立单个磁选车间能源消耗预测模型,即输入变量不是由同一层输入,而是根据变量起作用的前后次序分别在网络的不同层输入,从而使网络的规模减小;然后利用遗传算法的全局寻优能力对神经网络的权值进行优化,从而建立关于磁选工艺能源消耗数据优化的神经网络预测模型。仿真和实验结果证实了所提方法的有效性。更多还原

【Abstract】 According to the characteristics of multivariate,strong coupling,time varying,nonlinear and long time-delay in the magnetic separation process of DAGUSHAN Mineral Separation Plant,The conventional methodologies for optimal control based on mathematical model is difficult to use.In this paper,multi-input layer wavelet neural network for modeling large scale industrial processes is used to establish the energy consumption forecasting model of single magnetic separation workshop through neglect the specific physical and chemical actions.The input variables are not input in one layer,but in different layers according to their action sequences. Then the genetic algorithm(GA) is used to optimize the weights of neural network.Thus the Optimal Neural Network prediction model based on energy consumption data of magnetic separation process is established.The validity of the method mentioned can be proved by simulation and the experiment result.更多还原