提出了基于卡尔曼滤波与灰度预测的化工过程异常工况发展趋势预测方法.在化工过程异常工况下,首先,测量系统主要变量的数值,通过卡尔曼滤波处理测量数据,估计测量数据的真实值;其次,使用校正后的结果作为观测值,建立基于灰色预测理论的系统主要变量发展趋势的预测模型,根据模型计算系统主要变量随时间的变化数值;最后,分析预测结果,采取相应的措施,对化工过程异常工况进行有效管理.通过对实际案例的分析计算,说明了该方法的有效性,在化工过程异常工况下,为预测系统关键变量的发展趋势提供理论依据.
Proposed a Kalman filter and grey system based abnormal event trend prediction method. The proposed method is comprised of three parts when an abnormal event is encountered: i) the values of principle variables in the process are measured, these values are processed by Kalman filter method to get more accurate actual values; ii) with these estimated values, the abnormal event trend prediction model is formulated based on the grey system theory, and then evaluate the values of principle variables with time; and iii) after analyzing these calculated results, the trend of the abnormal event can be predicted, and then the desired actions can be taken in time to prevent the abnormal event getting worse. An actual temperature runaway ease is studied to demonstrate the effectiveness of the proposed method, and this method is helpful to manage abnormal event situation in chemical process.