NONLINEAR DATA RECONCILIATION METHOD BASED ON KERNEL PRINCIPAL COMPONENT ANALYSIS

In the industrial process situation, principal component analysis (PCA) is ageneral method in data reconciliation. However, PCA sometime is unfeasible to nonlinear featureanalysis and limited in application to nonlinear industrial process. Kernel PCA (KPCA) is extensionof PCA and can be used for nonlinear feature analysis. A nonlinear data reconciliation method basedon KPCA is proposed. The basic idea of this method is that firstly original data are mapped to highdimensional feature space by nonlinear function, and PCA is implemented in the feature space. Thennonlinear feature analysis is implemented and data are reconstructed by using the kernel. The datareconciliation method based on KPCA is applied to ternary distillation column. Simulation resultsshow that this method can filter the noise in measurements of nonlinear process and reconciliateddata can represent the true information of nonlinear process.

An NT-MT Combined Method for Gross Error Detection and Data Reconciliation

An NT-MT combined method based on nodal test (NT) and measurement test (MT) is developed for gross error detection and data reconciliation for industrial application. The NT-MT combined method makes use of both NT and MT tests and this combination helps to overcome the defects in the respective methods. It also avoids any artificial manipulation and eliminates the huge combinatorial problem that is created in the combined method based on the nodal test in the case of more than one gross error for a large process system. Serial compensation strategy is also used to avoid the decrease of the coefficient matrix rank during the computation of the proposed method. Simulation results show that the proposed method is very effective and possesses good performance.

A Novel Robust Nonlinear Dynamic Data Reconciliation

Outlier in one variable will smear the estimation of other measurements in data reconciliation （DR）. In this article, a novel robust method is proposed for nonlinear dynamic data reconciliation, to reduce the influence of outliers on the result of DR. This method introduces a penalty function matrix in a conventional least-square objective function, to assign small weights for outliers and large weights for normal measurements. To avoid the loss of data information, element-wise Mahalanobis distance is proposed, as an improvement on vector-wise distance, to construct a penalty function matrix. The correlation of measurement error is also considered in this article. The method introduces the robust statistical theory into conventional least square estimator by constructing the penalty weight matrix and gets not only good robustness but also simple calculation. Simulation of a continuous stirred tank reactor, verifies the effectiveness of the proposed algorithm.

A New Method to Solve Robust Data Reconciliation in Nonlinear Process

Data reconciliation is an effective technique for providing accurate and consistent value for chemical process. However, the presence of gross errors can severely bias the reconciled results. Robust estimators can significantly reduce the effect of gross errors and yield less-biased results. In this article, a new method is proposed to solve the robust data reconciliation problem of nonlinear chemical process. By using several technologies including linearization method, penalty function, virtual observation equation, and equivalent weights method, the robust data reconciliation problem can be transformed into least squares estimator problem which leads to the convenience in computation. Simulation results in a nonlinear chemical process demonstrate the efficiency of the proposed method.

A Modified Outlier Detection Method in Dynamic Data Reconciliation

Data reconciliation technology can decrease the level of corruption of process data due to measurement noise, but the presence of outliers caused by process peaks or unmeasured disturbances will smear the reconciled results. Based on the analysis of limitation of conventional outlier detection algorithms, a modified outlier detection method in dynamic data reconciliation (DDR) is proposed in this paper. In the modified method, the outliers of each variable are distinguished individually and the weight is modified accordingly. Therefore, the modified method can use more information of normal data, and can efficiently decrease the effect of outliers. Simulation of a continuous stirred tank reactor (CSTR) process verifies the effectiveness of the proposed algorithm.

Integration of Sensor Network Synthesis and Steady-State Online Data Reconciliation

Sensor network synthesis means to select certain variables to be measured, to select suitable sensors, to observe all the important variables, and to get reliable and accurate process data. The steady state online data reconciliation can remove the influences of errors in the measurements on the acquired data, and get consistent process data that satisfy the process constraints. These two techniques have been studied often but separately. This paper proposed to integrate these two techniques into a unified system to get better information about the process data. An industrial application of the integrated system to a crude oil distillation unit was also described, where valid process data were obtained and applied for online process optimization system, which led to a profit increase of about four million RMB annually.

Gross Error Detection and Identification Based on Parameter Estimation for Dynamic Systems

The detection and identification of gross errors, especially measurement bias, plays a vital role in data reconciliation for nonlinear dynamic systems. Although parameter estimation method has been proved to be a pow-erful tool for bias identification, without a reliable and efficient bias detection strategy, the method is limited in ef-ficiency and cannot be applied widely. In this paper, a new bias detection strategy is constructed to detect the pres-ence of measurement bias and its occurrence time. With the help of this strategy, the number of parameters to be es-timated is greatly reduced, and sequential detections and iterations are also avoided. In addition, the number of de-cision variables of the optimization model is reduced, through which the influence of the parameters estimated is reduced. By incorporating the strategy into the parameter estimation model, a new methodology named IPEBD (Improved Parameter Estimation method with Bias Detection strategy) is constructed. Simulation studies on a con-tinuous stirred tank reactor (CSTR) and the Tennessee Eastman (TE) problem show that IPEBD is efficient for eliminating random errors, measurement biases and outliers contained in dynamic process data.

Detection of gross errors using mixed integer optimization approach in process industry

A novel mixed integer linear programming (NMILP) model for detection of gross errors is presented in this paper. Yamamura et al.(1988) designed a model for detection of gross errors and data reconciliation based on Akaike information cri- terion (AIC). But much computational cost is needed due to its combinational nature. A mixed integer linear programming (MILP) approach was performed to reduce the computational cost and enhance the robustness. But it loses the super performance of maximum likelihood estimation. To reduce the computational cost and have the merit of maximum likelihood estimation, the simultaneous data reconciliation method in an MILP framework is decomposed and replaced by an NMILP subproblem and a quadratic programming (QP) or a least squares estimation (LSE) subproblem. Simulation result of an industrial case shows the high efficiency of the method.

Application of Automatic Mass Balance in Metal Accounting of Iron Ore Plant

The purpose of this work is to present state of the art mass balances closing strategies in ore treatment plants to report metal accounting in a reliable manner. The current methodology is based on usage of an expert system which seizes and filters the plant data, analyses situations of equipment stops and mostly instruments failures. It is still possible to internally create a strategy to check the data of an instrument failure. The system is coupled to software with a Lagrangian model of mass balance closure that receives the pre-treated data. Thus, the data goes through filtering and reconciling processes to generate a reliable mass balance of the ore treatment plant. This made it possible to obtain a system that is able to tell the real productivity of an ore plant through reliable mass balances. Residence times on equipment are critical to the balance, thus, they require good logic to eliminate errors due to the characteristics of non-steady state. Among the next challenges is the inclusion of a soft sensor as a replacement of an instrument required for mass balance closure.

Improvement of Wired Drill Pipe Data Quality via Data Validation and Reconciliation

Wired drill pipe（WDP）technology is one of the most promising data acquisition technologies in today s oil and gas industry.For the first time it allows sensors to be positioned along the drill string which enables collecting and transmitting valuable data not only from the bottom hole assembly（BHA）,but also along the entire length of the wellbore to the drill floor.The technology has received industry acceptance as a viable alternative to the typical logging while drilling（LWD）method.Recently more and more WDP applications can be found in the challenging drilling environments around the world,leading to many innovations to the industry.Nevertheless most of the data acquired from WDP can be noisy and in some circumstances of very poor quality.Diverse factors contribute to the poor data quality.Most common sources include mis-calibrated sensors,sensor drifting,errors during data transmission,or some abnormal conditions in the well,etc.The challenge of improving the data quality has attracted more and more focus from many researchers during the past decade.This paper has proposed a promising solution to address such challenge by making corrections of the raw WDP data and estimating unmeasurable parameters to reveal downhole behaviors.An advanced data processing method,data validation and reconciliation（DVR）has been employed,which makes use of the redundant data from multiple WDP sensors to filter/remove the noise from the measurements and ensures the coherence of all sensors and models.Moreover it has the ability to distinguish the accurate measurements from the inaccurate ones.In addition,the data with improved quality can be used for estimating some crucial parameters in the drilling process which are unmeasurable in the first place,hence provide better model calibrations for integrated well planning and realtime operations.

Synergies from off-gas analysis and mass balances for wastewater treatment--Some personal reflections on our experiences

Looking back at over a decade of research by herself and her group,the author advocates the added value of gas phase measurements and the application of mass balances,as well as the synergetic benefits obtained when combining both.The increased application of off-gas measurements for greenhouse gas emission monitoring offers a great opportunity to look at other components in the gas phase,particularly oxygen.Mass balances should not be strictly reserved for modellers but also prove useful while conducting lab experiments and studying full-scale measurement data.Combining off-gas measurements with mass balances may serve not only to quantify greenhouse gas emission factors and aeration efficiency but also to follow dynamic concentration profiles of dissolved components without dedicated sensors and/or to calculate other unmeasured variables.Mass-balance-based data reconciliation allows for obtaining reliable and accurate data,and even more when combined with off-gas analysis.

Reconciliation of size-density bivariate distributions over a separating node

Data reconciliation considers the restoration of mass balance among the noise prone measured data by way of component adjustments for the various particle size or particle density classes or assays over the separating node. In this paper, the method of Lagrange multipliers has been extended to balance bivariate feed and product size-density distributions of coal particles split from a settling column. The settling suspension in the column was split into two product fractions at 40% height from the bottom after a minute settling of homogenized suspension at start. Reconciliation of data assists to estimate solid flow split of particles to the settled stream as well as helps to calculate the profiles of partition curves of the marginal particle size or particle density distributions. In general, Lagrange multiplier method with uniform weighting of its components may not guarantee a smooth partition surface and thus the reconciled data needs further refinement to establish the nature of the surface. In order to overcome this difficulty, a simple alternative method of reconciling bivariate size-density data using partition surface concept is explored in this paper.