Industrial Soft Sensor Optimized by Improved PSO: A Deep Representation-Learning Approach
Soft sensors based on deep learning approaches are growing in popularity due to their ability to extract high-level features from training, improving soft sensors’ performance. In the training process of such a deep model, the set of hyperparameters is critical to archive generalization and reliabil...
| المؤلفون الرئيسيون: | , , |
|---|---|
| التنسيق: | مقال |
| اللغة: | English |
| منشور في: |
MDPI AG
2022-09-01
|
| سلاسل: | Sensors |
| الموضوعات: | |
| الوصول للمادة أونلاين: | https://www.mdpi.com/1424-8220/22/18/6887 |
| _version_ | 1827656865811529728 |
|---|---|
| author | Alcemy Gabriel Vitor Severino Jean Mário Moreira de Lima Fábio Meneghetti Ugulino de Araújo |
| author_facet | Alcemy Gabriel Vitor Severino Jean Mário Moreira de Lima Fábio Meneghetti Ugulino de Araújo |
| author_sort | Alcemy Gabriel Vitor Severino |
| collection | DOAJ |
| description | Soft sensors based on deep learning approaches are growing in popularity due to their ability to extract high-level features from training, improving soft sensors’ performance. In the training process of such a deep model, the set of hyperparameters is critical to archive generalization and reliability. However, choosing the training hyperparameters is a complex task. Usually, a random approach defines the set of hyperparameters, which may not be adequate regarding the high number of sets and the soft sensing purposes. This work proposes the RB-PSOSAE, a Representation-Based Particle Swarm Optimization with a modified evaluation function to optimize the hyperparameter set of a Stacked AutoEncoder-based soft sensor. The evaluation function considers the mean square error (MSE) of validation and the representation of the features extracted through mutual information (MI) analysis in the pre-training step. By doing this, the RB-PSOSAE computes hyperparameters capable of supporting the training process to generate models with improved generalization and relevant hidden features. As a result, the proposed method can generate more than 16.4% improvement in RMSE compared to another standard PSO-based method and, in some cases, more than 50% improvement compared to traditional methods applied to the same real-world nonlinear industrial process. Thus, the results demonstrate better prediction performance than traditional and state-of-the-art methods. |
| first_indexed | 2024-03-09T22:35:19Z |
| format | Article |
| id | doaj.art-f1542bdce1694283b40341f07a71c942 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T22:35:19Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-f1542bdce1694283b40341f07a71c9422023-11-23T18:50:53ZengMDPI AGSensors1424-82202022-09-012218688710.3390/s22186887Industrial Soft Sensor Optimized by Improved PSO: A Deep Representation-Learning ApproachAlcemy Gabriel Vitor Severino0Jean Mário Moreira de Lima1Fábio Meneghetti Ugulino de Araújo2Computer Engineering and Automation Department, Federal University of Rio Grande do Norte, 3000 Senador Salgado Filho Avenue, Natal 59078-970, RN, BrazilComputer Engineering and Automation Department, Federal University of Rio Grande do Norte, 3000 Senador Salgado Filho Avenue, Natal 59078-970, RN, BrazilComputer Engineering and Automation Department, Federal University of Rio Grande do Norte, 3000 Senador Salgado Filho Avenue, Natal 59078-970, RN, BrazilSoft sensors based on deep learning approaches are growing in popularity due to their ability to extract high-level features from training, improving soft sensors’ performance. In the training process of such a deep model, the set of hyperparameters is critical to archive generalization and reliability. However, choosing the training hyperparameters is a complex task. Usually, a random approach defines the set of hyperparameters, which may not be adequate regarding the high number of sets and the soft sensing purposes. This work proposes the RB-PSOSAE, a Representation-Based Particle Swarm Optimization with a modified evaluation function to optimize the hyperparameter set of a Stacked AutoEncoder-based soft sensor. The evaluation function considers the mean square error (MSE) of validation and the representation of the features extracted through mutual information (MI) analysis in the pre-training step. By doing this, the RB-PSOSAE computes hyperparameters capable of supporting the training process to generate models with improved generalization and relevant hidden features. As a result, the proposed method can generate more than 16.4% improvement in RMSE compared to another standard PSO-based method and, in some cases, more than 50% improvement compared to traditional methods applied to the same real-world nonlinear industrial process. Thus, the results demonstrate better prediction performance than traditional and state-of-the-art methods.https://www.mdpi.com/1424-8220/22/18/6887particle swarm optimizationsoft sensorsdeep learningstacked autoencodersmutual information |
| spellingShingle | Alcemy Gabriel Vitor Severino Jean Mário Moreira de Lima Fábio Meneghetti Ugulino de Araújo Industrial Soft Sensor Optimized by Improved PSO: A Deep Representation-Learning Approach Sensors particle swarm optimization soft sensors deep learning stacked autoencoders mutual information |
| title | Industrial Soft Sensor Optimized by Improved PSO: A Deep Representation-Learning Approach |
| title_full | Industrial Soft Sensor Optimized by Improved PSO: A Deep Representation-Learning Approach |
| title_fullStr | Industrial Soft Sensor Optimized by Improved PSO: A Deep Representation-Learning Approach |
| title_full_unstemmed | Industrial Soft Sensor Optimized by Improved PSO: A Deep Representation-Learning Approach |
| title_short | Industrial Soft Sensor Optimized by Improved PSO: A Deep Representation-Learning Approach |
| title_sort | industrial soft sensor optimized by improved pso a deep representation learning approach |
| topic | particle swarm optimization soft sensors deep learning stacked autoencoders mutual information |
| url | https://www.mdpi.com/1424-8220/22/18/6887 |
| work_keys_str_mv | AT alcemygabrielvitorseverino industrialsoftsensoroptimizedbyimprovedpsoadeeprepresentationlearningapproach AT jeanmariomoreiradelima industrialsoftsensoroptimizedbyimprovedpsoadeeprepresentationlearningapproach AT fabiomeneghettiugulinodearaujo industrialsoftsensoroptimizedbyimprovedpsoadeeprepresentationlearningapproach |