Operating optimality assessment based on optimality related variations and nonoptimal cause identification for industrial processes

In this study, a novel online operating optimality assessment based on optimality related variations and nonoptimal cause identification method is proposed for industrial processes. The optimality related variations are extracted from each steady performance grade by analyzing the common and unique variations among steady performance grades, which avoids the time-consuming data alignment. When the optimality related variations are used in assessment, both the robustness and sensitivity of the assessment method are improved compared with the PCA-based assessment for its abilities in highlighting the process variations related to operating performance and excluding those unrelated variations. Based on the similarities between the optimality related variations of the online data and that of each steady performance grade, the process operating performance can be evaluated as the steady performance grade or the conversion process between performance grades, and this provides more information for the in-depth understanding of the process operating. For nonoptimal operating performance, the nonoptimal cause identification strategy is developed for further production adjustment and performance improvement. Finally, the efficiency of the proposed method is illustrated with a case of gold hydrometallurgical process.     

基于改进的动态因果图推理算法的复杂工业过程运行状态在线评价

复杂工业生产中,为获得优质产品和更高经济效益,需保证生产过程运行于"正常"且"优"的状态.针对定量信息与定性信息同时存在的生产过程,提出一种基于改进的动态因果图(dynamic causality diagram,DCD)的过程运行状态评价方法.该方法以因果图理论为基础,扩充了定量分析和定性分析结合描述的动态因果图;针对复杂工业生产过程中存在多个相互关联的事件联合作用导致一个事件发生的特点,提出用联合因果强度来描述事件间关系的方法.最后以金湿法冶金氰化浸出过程为应用背景,分别利用传统DCD与改进DCD两种方法建立评价模型,对比分析评价结果,验证了文中提出方法的有效性和先进性.     

基于GMM和贝叶斯推理的多模态过程运行状态评价

为使综合经济效益最大化,生产过程应保持在最优运行状态等级.针对多模态过程运行状态等级优劣判断问题,提出一种运行状态等级评价方法.该方法对同一运行状态等级的多模态数据建立一个高斯混合模型(Gaussian mixture model,GMM),确保特征提取的准确性,避免模态划分问题.至于在线评价策略,本文采用贝叶斯推理,确定当前运行状态属于各等级的后验概率.并引入滑动窗口,判定当前运行状态等级,有效解决多模态过程运行状态在线评价问题.针对"非优"运行状态,本文提出一种基于变量偏导数的贡献计算方法,对导致过程运行状态等级"非优"的原因变量进行追溯.最后,通过田纳西–伊斯曼(Tennessee–Eastman,TE)过程验证所提方法的有效性.     

基于Fisher 判别分析的过程运行状态在线评价

为了实时掌握生产过程运行状态, 提出一种基于Fisher 判别分析(FDA) 的过程运行状态在线评价方法. 提出 离线数据分类与识别算法, 以识别不同稳定运行状态的建模数据及其对应的状态等级; 利用FDA提取各个稳定运行 状态的特征属性, 建立评价模型; 在线评价时, 通过“时间窗口”数据特征与各个状态等级的相似度, 实时评价过程运行状态. 将所提出的方法应用于某湿法冶金过程的仿真结果验证了该方法的有效性.

     

A two-layer fuzzy synthetic strategy for operational performance assessment of an industrial hydrocracking process

The operating performance may deteriorate from the optimal condition in industrial hydrocracking process due to changes in product requirements, equipment performance degradation, environmental noises, and random disturbances, etc. Therefore, it is necessary to carry out effective online operating performance assessment for it. In this paper, a two-layer fuzzy synthetic strategy is proposed for operating performance assessment of the hydrocracking process. First, an operating performance assessment index is constructed based on the concept of positive deviation of the quality variables. As there are many products in the process, an assessment system is further built based on the performance index. To calculate the indicators in the assessment index system in real-time, partial least squares is used for prediction. To solve the problem of limited labeled samples in the prediction, the process variables are classified into several subspace blocks according to their correlations with different key indicators. Then, a two-layer fuzzy synthetic evaluation strategy is designed to evaluate the actual operating performance of the hydrocracking process. The proposed strategy is finally illustrated with a series of cases on an industrial hydrocracking process in China.     

工业过程的子空间模型辨识

子空间模型辨识方法(SMI)是一类新兴的直接估计线性状态空间模型的黑箱建模方法,近年来获得了广泛关注.和传统的线性建模方法相比,SMI的优势不仅在于算法本身的简单可靠,也在于它的状态空间表达.本文首先简要介绍了SMI的基本思想以及3种基本算法(N4SID,MOESP,CVA).然后将这类方法应用于一个实际的工业过程建模,同时对3种SMI基本算法和一种传统辨识算法—预测误差方法(PEM)进行了研究对比.     

基于线性二次型高斯基准的多变量预测控制技术经济性能评估

由于受控过程参数的漂移及缺乏维护,令采用的控制器性能逐渐降低,需要做经济性能评估,以确保其最佳运行状态.因为目前最小方差评估算法没有考虑控制器的约束条件,对此我们采用线性二次型高斯(linearquadratic Gaussian,LQG)基准的模型预测控制(model predictive control,MPC)双层优化控制结构,将控制和输出的加权值引入上层经济性能指标,通过求解LQG问题获取控制与输出方差关系的离散点集,进一步拟合Pareto最优曲面方程,建立优化命题并求解最优经济指标及设定值.对延迟焦化加热炉的多变量MPC控制进行了性能评估及分析,证明该方法可以改进控制器设计,提高经济效益.     

基于ISDAE模型的复杂工业过程运行状态评价方法及应用

工业过程的运行状态评价对保证产品质量及提升企业综合经济效益具有重要意义. 针对工业过程中存在强非线性、信息冗余以及不确定性因素影响而难以建立稳健可靠的运行状态评价模型问题, 提出一种基于综合经济指标驱动的稀疏降噪自编码器模型(Comprehensive economic index driven sparse denoising autoencoder, ISDAE)的复杂工业过程运行状态评价方法. 首先, 在SDAE (Sparse denoising autoencoder)模型中引入综合经济指标预测误差项, 迫使SDAE学习与综合经济指标相关的数据特征, 建立ISDAE特征提取模型. 其次, 将ISDAE模型所学特征作为输入训练运行状态识别模型, 级联特征提取模型和运行状态识别模型并通过微调网络结构参数获得运行状态评价模型. 另外, 针对非优状态, 提出一种基于自编码器贡献图算法的非优因素追溯方法, 通过计算变量的贡献率识别非优因素. 最后, 将所提方法应用于重介质选煤过程, 验证所提方法的有效性和实用性.     

RFID与条码融合的离散制造过程自动标识技术

针对现代制造业对物流与信息流实时同步的需求,结合条码技术和射频识别技术的特点,提出一种基于两者融合的制造过程自动标识技术。给出的技术包括物理融合和数据融合两个层次。详细阐述了物理融合的串行及并行融合模式和技术实施要点,并将数据融合归纳为将条码信息转移到电子标签、将电子标签信息转移到条码和电子标签之间信息的相互转移三类数据融合方式。结合具体生产线,说明了该技术在生产现场应用的方法。结果表明,提出的技术可以提高制造企业物流与信息流的一致性,为保证企业信息同步提供了支撑。     

An Integrated Artificial Neural Network Fuzzy C-Means-Normalization Algorithm for performance assessment of decision-making units: The cases of auto industry and power plant

Efficiency frontier analysis has been an important approach of evaluating firms’ performance in private and public sectors. There have been many efficiency frontier analysis methods reported in the literature. However, the assumptions made for each of these methods are restrictive. Each of these methodologies has its strength as well as major limitations. This study proposes two non-parametric efficiency frontier analysis sub-algorithms based on (1) Artificial Neural Network (ANN) technique and (2) ANN and Fuzzy C-Means for measuring efficiency as a complementary tool for the common techniques of the efficiency studies in the previous studies. Normal probability plot is used to find the outliers and select from these two methods. The proposed computational algorithms are able to find a stochastic frontier based on a set of input–output observational data and do not require explicit assumptions about the functional structure of the stochastic frontier. In these algorithms, for calculating the efficiency scores, a similar approach to econometric methods has been used. Moreover, the effect of the return to scale of decision-making unit (DMU) on its efficiency is included and the unit used for the correction is selected by notice of its scale (under constant return to scale assumption). Also in the second algorithm, for increasing DMUs’ homogeneousness, Fuzzy C-Means method is used to cluster DMUs. Two examples using real data are presented for illustrative purposes. First example which deals with power generation sector shows the superiority of Algorithm 2 while the second example dealing auto industries of various developed countries shows the superiority of Algorithm 1. Overall, we find that the proposed integrated algorithm based on ANN, Fuzzy C-Means and Normalization approach provides more robust results and identifies more efficient units than the conventional methods since better performance patterns are explored.     

在线参数辨识和扩展卡尔曼算法的锂离子电池SOC估算研究

荷电状态(SOC)的准确估计对锂离子电池的在线实时监测和安全控制具有重要意义。以中航锂电池为研究对象,选择二阶阻容(RC)模型对电池工作特性进行表征,并结合多种工况情形对锂离子电池进行研究分析。考虑到参数辨识的初值对在线辨识修正效果的影响,搭建仿真模型与电池脉冲工况特性比较验证,仿真误差在0.05 V以内。在此基础上,构建含有遗忘因子的递推最小二乘法(FFRLS)的在线参数辨识系统,对电池动态应力测试工况(DST)进行仿真预测,相对误差在1.50%以内。针对离线参数辨识的不足,采用在线参数辨识结合扩展卡尔曼(EKF)算法对工况下电池SOC进行估计。试验结果表明,在线参数辨识下,EKF算法能够有效表征系统SOC估算,相对误差精度在0.3%以内。