Just-in-Time Correntropy Soft Sensor with Noisy Data for Industrial Silicon Content Prediction

An ensemble JITL method based on multi-weighted similarity measures for cold rolling force prediction

In the cold tandem rolling process, the product quality and yield are affected by the accuracy of rolling force prediction directly. Fix prediction model is not applicable to the multi-operating conditions rolling environment. In addition, appropriate samples can be hardly selected by a single similarity measure because of the insufficient process knowledge. In order to solve these issues, an ensemble just-in-time-learning modeling method based on multi-weighted similarity measures (MWS-EJITL) is proposed. Firstly, multi-weighted similarity measures is used to select relevant samples. Then, the local model is constructed and the output value of the query data is estimated. Finally, the ensemble learning strategy is adopted to integrate the outputs of each local model. On this basis, the cumulative similarity factor is introduced to optimize the number of samples of local modeling, and the similarity threshold is set to update the local model adaptively. The rolling force prediction experiment verify the effectiveness and accuracy of MWS-EJITL method.

Autoencoder-based nonlinear Bayesian locally weighted regression for soft sensor development

The framework of locally weighted learning (LWL) has established itself as a popular tool for developing nonlinear soft sensors in process industries. For LWL-based soft sensors, the key factor for achieving high performance is to construct accurate localized models. To this end, in this paper a nonlinear local model training algorithm called nonlinear Bayesian weighted regression (NBWR) is proposed. In the NBWR, the nonlinear features of process data are first extracted by the autoencoder; then, given a query sample a local dataset is selected on the feature space and a fully Bayesian regression model with differentiated sample weights is developed. The benefits of this approach, which include better consistency of correlation, stronger abilities to deal with process nonlinearities and uncertainties, overfitting and numerical issues, lead to superior performance. The NBWR is used for developing a soft sensor under the LWL framework, and a real-world industrial process is used to evaluate the performance of the NBWR-based soft sensor. The experimental results demonstrate that the proposed method outperforms several benchmarking soft sensing approaches.

An Improved Normalized Mutual Information Variable Selection Algorithm for Neural Network-Based Soft Sensors

In this paper, normalized mutual information feature selection (NMIFS) and tabu search (TS) are integrated to develop a new variable selection algorithm for soft sensors. NMIFS is applied to select influential variables contributing to the output variable and avoids selecting redundant variables by calculating mutual information (MI). A TS based strategy is designed to prevent NMIFS from falling into a local optimal solution. The proposed algorithm performs the variable selection by combining the entropy information and MI and validating error information of artificial neural networks (ANNs); therefore, it has advantages over previous MI-based variable selection algorithms. Several simulation datasets with different scales, correlations and noise parameters are implemented to demonstrate the performance of the proposed algorithm. A set of actual production data from a power plant is also used to check the performance of these algorithms. The experiments showed that the developed variable selection algorithm presents better model accuracy with fewer selected variables, compared with other state-of-the-art methods. The application of this algorithm to soft sensors can achieve reliable results.

Soft Sensing of Silicon Content via Bagging Local Semi-Supervised Models

The silicon content in industrial blast furnaces is difficult to measure directly online. Traditional soft sensors do not efficiently utilize useful information hidden in process variables. In this work, bagging local semi-supervised models (BLSM) for online silicon content prediction are proposed. They integrate the bagging strategy, the just-in-time-learning manner, and the semi-supervised extreme learning machine into a unified soft sensing framework. With the online semi-supervised learning method, the valuable information hidden in unlabeled data can be explored and absorbed into the prediction model. The application results to an industrial blast furnace show that BLSM has better prediction performance compared with other supervised soft sensors.

Temperature Measurement Method for Blast Furnace Molten Iron Based on Infrared Thermography and Temperature Reduction Model

The temperature measurement of blast furnace (BF) molten iron is a mandatory requirement in the ironmaking process, and the molten iron temperature is significant in estimating the molten iron quality and control blast furnace condition. However, it is not easy to realize real-time measurement of molten iron temperature because of the harsh environment in the blast furnace casthouse and the high-temperature characteristics of molten iron. To achieve continuous detection of the molten iron temperature of the blast furnace, this paper proposes a temperature measurement method based on infrared thermography and a temperature reduction model. Firstly, an infrared thermal imager is applied to capture the infrared thermal image of the molten iron flow after the skimmer. Then, based on the temperature distribution of the molten iron flow region, a temperature mapping model is established to measure the molten iron temperature after the skimmer. Finally, a temperature reduction model is developed to describe the relationship between the molten iron temperature at the taphole and skimmer, and the molten iron temperature at the taphole is calculated according to the temperature reduction model and the molten iron temperature after the skimmer. Industrial experiment results illustrate that the proposed method can achieve simultaneous measurement of molten iron temperature at the skimmer and taphole and provide reliable temperature data for regulating the blast furnace.

A KPI-Based Probabilistic Soft Sensor Development Approach that Maximizes the Coefficient of Determination

Advanced technology for process monitoring and fault diagnosis is widely used in complex industrial processes. An important issue that needs to be considered is the ability to monitor key performance indicators (KPIs), which often cannot be measured sufficiently quickly or accurately. This paper proposes a data-driven approach based on maximizing the coefficient of determination for probabilistic soft sensor development when data are missing. Firstly, the problem of missing data in the training sample set is solved using the expectation maximization (EM) algorithm. Then, by maximizing the coefficient of determination, a probability model between secondary variables and the KPIs is developed. Finally, a Gaussian mixture model (GMM) is used to estimate the joint probability distribution in the probabilistic soft sensor model, whose parameters are estimated using the EM algorithm. An experimental case study on the alumina concentration in the aluminum electrolysis industry is investigated to demonstrate the advantages and the performance of the proposed approach.