Search

Data reconciliation is extensively applied to improve the accuracy and reliability of plant measurements. It relies on process models ranging from simple mass and energy conservation equations to complete causal models. The precision of reconciled data mainly depends on the complexity and quality of plant models used to develop data reconciliation observers. In practice, the difficulty of obtaining detailed models prevents the application of powerful observers like the Kalman filter. The objective of this study is to propose a methodology to build a model for a flotation circuit to support the implementation of a Kalman filter for dynamic data reconciliation. This modeling approach extracts essential information from the plant topology, nominal operating conditions, and historical data. Simulation results illustrate that applying a Kalman filter based on a rough empirical model that has been correctly tuned gives better estimates than those obtained with sub-model based observers. [ABSTRACT FROM AUTHOR]

Abstract: Data reconciliation is a well-known technique to improve accuracy and reliability of plant measurements. It relies on process models that could range from simple mass and energy conservation equations to complete causal dynamic models. Generally, precise estimates imply detailed plant models that could be difficult to build and update in practice. The trade-off between modeling efforts and estimation performances has thus lead to various approaches to deal with plant dynamics. The objective of the paper is to review and compare most common observers used for dynamic data reconciliation in the mineral and metallurgical processing industries. Comparisons are carried out using a separation unit and a flotation circuit as simulated benchmark plants. Observer performances are evaluated in terms of variance reduction. Strengths and weaknesses of the different methods are highlighted. Aspects such as estimation of model parameters, detection of gross errors, and handling of bilinear equations and plant non-linearities are discussed. [Copyright &y& Elsevier]

Abstract: To reduce impacts of measurement errors on plant variables, data reconciliation is widely applied in process industries. Reconciled measurements are used in applications such as performance monitoring, process control, or real-time optimization. However, precise estimation generally relies on accurate and detailed process models which could be difficult to build in practice. The trade-off between estimate precision and model complexity is a relevant challenge motivating the development of effective observers with limited modeling efforts. This paper proposes a data reconciliation method based on a simple mass and/or energy conservation sub-model that also considers the autocovariance function of plant node imbalances. The observer is applied to simulated benchmark plants and its performance is evaluated in terms of variance reduction and robustness against modeling errors. Results show a superior performance in comparison with classical sub-model based methods and reveal less performance degradation than the Kalman filter in presence of model uncertainties. [Copyright &y& Elsevier]