误差校正新方法在重整装置芳烃收率在线软测量中的应用

针对软测量建模中含有误差会极大影响模型精度问题,提出一种将数据误差校正、模型参数校正和模型结构校正技术相结合的软测量误差三重校正方法.该方法可以有效地处理测量数据的过失误差和随机误差,从而提高软测量精度,并且该方法直接面向数据,实现方便.应用该方法对某芳烃厂的连续重整装置的芳烃收率软测量进行误差校正,取得了很好的效果.     

共中心点(CMP)层位校正

在地震资料处理中,由于动校正时速度难以精确的提取,导致存在剩余动校正量。为了减小剩余动校正量对地震分辨率的影响,本文采取多道相关的方法建立模型道,按最大能量准则拾取剩余动校正量,然后采用最小平方滤波器进行滤波插值来完成高精度动态拉伸校正,以便消除剩余动校正量的影响,从而保持地震记录的原有分辨率,并且可以使叠加过程中的信噪比得以提高,理论试算与实际资料处理说明该方法的的效果比良好,不仅减小剩余动校正量影响,而且也提高了信噪比。     

双线性约束过程的鲁棒自适应数据校正方法

采用污染正态分布模型进行数据校正,相对于传统的最小二乘方法具有较好的鲁棒性,然而参数估计结果的精确度依赖于误差发生概率和方差比值两个先验模型参数的选取,这在实际生产中难以获得,采用固定的方差比也不符合实际,因而其应用受到了限制。本文针对污染正态分布模型的不足,提出了一种鲁棒自适应误差分布模型,该模型具有与标准正态分布模型相似的分布密度函数,不同之处在于采用鲁棒自适应可变权重因子调节误差方差,通过放大显著误差方差,减小其对参数估计的影响。将该模型用于双线性约束数据校正问题,并采用Lagrange乘子法得到鲁棒自适应最小二乘分析解,同时还对鲁棒自适应数据校正中的测量数据相关性问题进行了研究。仿真结果证实了该方法的有效性。     

卡尔曼滤波器在储罐液位监控系统中的应用

提出了基于卡尔曼滤波器的储罐液位监控方法,通过卡尔曼滤波器估计储罐的真实液位,再对其进行监测与控制。采用ST语言实现卡尔曼滤波算法,并将其应用在Rockwell PLC平台上,现场测试结果表明:卡尔曼滤波能很好地滤除储罐运行时产生的各类噪声,在判断储罐液位趋势及提高液位控制精度等方面都收到了较为理想的效果。     

NIRS法测定液碱中氢氧化钠含量的方法研究

采用酸碱滴定法来测量液碱中氢氧化钠含量作为原始数据,同时收集扫描近红外谱图,建立校正模型,最终建立了NIRS(近红外)测定液碱中氢氧化钠含量的方法。采用MSA和T检验验证了测量系统和对比数据的可靠性。NIRS具有分析速度快、无需前处理、不使用化学试剂的优点。快速高效的出具分析数据,更有利益服务于工艺,提高工艺生产效率。     

基于Unscented卡尔曼滤波器的管道泄漏的快速检测与定位

针对扩展卡尔曼滤波器(EKF)对非线性模型进行状态估计时,需要将非线性模型进行线性化,而天然气管道模型的非线性程度十分严重的情况,提出一种基于无偏卡尔曼滤波器(UKF)估计的管道泄漏检测与定位方法.该方法是一类用采样策略逼近非线性模型的方法,避免了线性化,并且不需要计算Jacob矩阵,UKF的收敛速度明显优于EKF,能够快速地检测出泄漏.仿真结果说明UKF在泄漏检测中相对于EKF的优势.     

基于模型性能评估的递推PLS建模及应用

为了降低递推部分最小二乘(RPLS)建模方法的模型校正频率,开发了一种基于模型性能评估的RPLS(MPA-RPLS)模型.首先,根据过程的初始特性,自动生成模型的置信限,以均方根误差(RMSEP)为性能指标,评估模型性能;依据模型性能的评估结果,选择性地启动模型校正和置信限校正.然后,引入滑动平均滤波器消除过程变量中的噪声,探讨噪声对模型性能的影响程度.最后,将MPA-RPLS模型应用于一个化学反应过程--C₈芳烃临氢异构化过程,基于大量工业数据,进行仿真验证.仿真结果表明:本文开发的模型仅以微小的精度损失换取了模型计算效率的大幅提高(即模型校正频率大幅下降);滑动平均滤波器可有效地处理变量的噪声,改善模型的预测精度.     

基于参数化模型的近红外干涉图相位校正方法

干涉图是傅里叶变换近红外光谱仪测量获得的最原始信息,干涉图的相位校正直接关系到近红外光谱的质量。本文提出利用参数化模型对干涉图进行相位校正,该方法认为仪器测量所得的相位误差与波数之间存在函数关系。利用同一台仪器所测得的大量干涉图拟合出模型的参数,即可获得相位误差与波数之间的关系。对新样品的干涉图相位校正时,直接利用已拟合的关系进行校正即可获得高质量的干涉图。实验结果表明,采用参数化模型校正干涉图的效果比仪器自带Mertz算法效果好。利用参数化模型校正干涉图可提高干涉图处理的效率,为近红外光谱分析技术的研究提供了新的研究思路。     

基于神经网络的化工过程测量数据在线校正技术的研究

研究了人工神经网络在化工过程测量数据校正中的应用，提出了新的样本构造方法和神经网络的在线训练策略。对某乙烯装置裂解气分离系统测量数据，应用自行设计开发的改进算法的神经网络与数据校正系统集成运行，结果表明基于神经网络的数据校正技术能对测量数据中所含的随机误差和过失误差进行同时校正，提高过程数据的精度和校正过程的稳定性，同时满足数据校正的实时性要求。     

过程系统变负荷下的数据校正与参数估计方法

过程系统的数据校正与参数估计是进行实时操作优化与过程控制的基础。过程系统变负荷下由于模型参数变化的非线性及显著误差的影响,导致数据校正与参数估计的结果不准确,从而影响实时操作优化与过程控制的效率。针对此问题,本文提出了一种用于变负荷下的数据校正与参数估计方法。此方法主要包括过程的稳态检测与数据采样,多工况下的数据聚类和基于多组测量的数据校正与参数估计。首先选择有效和可靠的过程测量数据,根据变负荷下工况的波动性与系统的非线性特征进行数据聚类,最后基于聚类结果调整模型参数使得模型输出与过程测量数据偏差最小。此方法可有效地减小模型参数变化的非线性及显著误差对数据校正与参数估计结果的影响。基于现场的测量数据,将此方法应用于空气分离流程系统中,结果显示了基于此方法的数据校正与参数估计结果更准确。     

Estimation of time-dependent reaction coefficient with extended Kalman filter approach

This work introduces an extended Kalman filter (EKF) to the estimation of the unknown time-dependent reaction coefficient based on the concentration measurement data. An autocatalytic reaction pathway is chosen as a model problem. This inverse estimation algorithm does not assume any functional form of the reaction coefficient. The performance of the proposed algorithm is verified through the numerical experiments with the exact and the contaminated concentration measurement data.     

Numerical study of pleated fabric cartridges during pulse-jet cleaning

We established a numerical model and used computational fluid dynamics (CFD) analysis to observe transient flow behavior across pleated filter cartridges in a dust collector during pulse-jet cleaning. The numerical results were in good agreement with the filter-testing data during important periods including during pressure ramp-up and valve opening. Larger errors for predicting overpressure occurred during the pressure ramp-down period likely due to the uniformity of the filters' permeability. This confirmed that the numerical model demonstrated the cleaning efficiency and the local cleaning quality of three different filter cartridges with different filter dimensions and pleat ratios. Data calculated from the average static pressure on the filters' surfaces were more closely correlated to cleaning efficiency than overpressure. The surface static pressure distribution along all filter cartridges showed that the top area of the filter cartridge is difficult to clean because of the lower surface pressure generated by the pulse jet. Filter cartridges with higher pleat ratios were found to have greater instances of incomplete cleaning due to the large variation of static pressure distribution along the filter cartridges. Our results showed that although information such as average pressure and pressure distribution on the filter surface is difficult to obtain by physical measurement, this data is tractable using CFD analysis and is useful for filter design and system optimization.     

数字滤波技术在聚合物熔体在线流变仪中的应用

根据聚合物熔体在线流变仪测量系统中自变量参数熔体压力在整个测量过程中的变化特点,研究了熔体压力测量精度对最终因变量熔体剪切黏度的影响。在实验测量和理论分析的基础上,选择适应系统变化特点的滤波算法对系统变化过程的压力进行数字滤波,实现提高最终测量参数熔体剪切黏度测量精度的目的。结果表明,通过递推平均滤波算法,可以有效提高聚合物熔体剪切黏度的测量精度。     

Filtration and Dewatering of Fine Coal

Abstract

A fundamental study on filtration and dewatering of fine coal is described. Experiments are being conducted in three areas: (1) The microscopic analysis of filter cakes, (2) The measurement of equilibrium desaturation and (3) The determination of filtration and dewatering rates. Preliminary experimental results are presented together with some observations on the microstructure of filter cakes. A three dimensional network model has also been developed and is being used to analyze experimental data.     

PVC浆料中残留VCM含量测定方法的探讨

介绍了测定PVC浆料中残留VCM含量的方法,并对浆液测定法、滤饼测定法、仿国际法进行了理论分析,确定了最佳测定方法--滤饼测定法。通过工艺参数调整试验,用大量测定的数据证明该方法准确可靠。     

激光传感器在轮胎偏心度测量中的应用

介绍激光传感器在轮胎偏心度测量中的应用。采用激光传感器测量,对测量值进行滤波后可得到轮胎偏心度值。通过激光传感器对同一轮胎在试验机上进行不卸胎重复偏心度测量10次并通过FIR数字滤波器消除测量信号中的各种干扰,测量结果与机械式传感器测量结果对比证明,采用激光传感器测量轮胎偏心度可大大提高数据的准确性和稳定性。     

Improved Estimation of the Langmuir Adsorption Isotherm Using Multiscale Filtering

Abstract

Adsorption isotherms play an important role in the design and analysis of adsorption processes. These isotherms are estimated empirically from measurements of adsorption process variables. Unfortunately, these measurements are usually contaminated with errors that degrade the accuracy of estimated isotherms. Therefore, these errors need to be filtered for improved isotherm estimation accuracy. Multiscale wavelet-based filtering has been shown to be a powerful filtering tool. In this work, multiscale filtering is utilized to improve the estimation accuracy of the Langmuir adsorption isotherm in the presence of measurement noise in the data by developing a multiscale isotherm estimation algorithm. The idea behind the algorithm is to use multiscale filtering to filter the data at different scales, use the filtered data from all scales to construct multiple isotherms, and then select among all scales the isotherm that best represent the data based on a cross-validation mean squares error criterion. The developed multiscale isotherm estimation algorithm is shown to outperform the conventional time-domain estimation method through a simulated example.     

Development of an Automatic Beta Gauge Particulate Sampler with Filter Cassette Mechanism

A beta gauge particulate sampler for measuring the aerosol mass concentration in the ambient air is described. The instrument is automatically calibrated with two self-calibration mass standards during each sampling period, while it samples particles continuously with minimum sampling dead-time loss. Key design features of the instrument based on the attenuation of beta radiation include filter cassette mechanism, auto-calibration system, low sampling dead-time, high sensitivity, and straightforward audit procedures. The instrument consists of three main components: PM 10 inlet, mechanical filter movement system, and control and data processing system. The mechanical filter movement system includes particle collection system with filter cassette magazine, g -ray measuring module and particle sampling module, auto-calibration system, and flow control system. The control and data processing system performs filter cassette movement control, sampling pump control, and data analysis. The instrument has been tested in the field to compare the measurement results with those by gravimetric mass measurement. The developed beta gauge instrument has been proved to be an efficient measuring guage for the ambient particulate mass determination.     

Enhancing Controller Performance via Dynamic Data Reconciliation

Measured values of process variables are subject to measurement noise. The presence of measurement noise can result in detuned controllers in order to prevent excessive adjustments of manipulated variables. Digital filters, such as exponentially weighted moving average (EWMA) and moving average (MA) filters, are commonly used to attenuate measurement noise before controllers. In this article, we present another approach, a dynamic data reconciliation (DDR) filter. This filter employs discrete dynamic models that can be phenomenological or empirical, as constraints in reconciling noisy measurements. Simulation results for a storage tank and a distillation column under PI control demonstrate that the DDR filter can significantly reduce propagation of measurement noise inside control loops. It has better performance than the EWMA and MA filters, so that the overall performance of the control system is enhanced.     

An experimental evaluation of Kalman filtering

The effectiveness of the stationary form of the discrete Kalman filter for state estimation in noisy process systems was demonstrated by simulated and experimental tests on a pilot plant evaporator. The filter was incorporated into a multivariable, computer control system and resulted in good control despite process and/or measurement noise levels of 10%. The results were significantly better than those obtained when the Kalman filter was omitted or replaced by conventional exponential filters. In this application the standard Kalman filter was reasonably insensitive to incorrect estimates of initial conditions or noise statistics and to errors in model parameters. The filter estimates were sensitive to unmeasured process disturbances. However this sensitivity could be reduced by treating the noise covariance matrices R and Q as design parameters rather than noise statistics and selecting values which result in increased weighting of the process measurements relative to the calculated model states.