Extração de características representativas para o desenvolvimento de sensores virtuais industriais: uma abordagem baseada em aprendizado profundo
Deep learning is growing in popularity in virtual sensor modeling problems - the soft sensors - applied to industrial processes of accentuated nonlinearity. Virtual sensors can generate estimates of process variables, which are associated with quality indexes in realtime. Thus, such sensors are a vi...
Na minha lista:
| Autor principal: | |
|---|---|
| Outros Autores: | |
| Formato: | doctoralThesis |
| Idioma: | pt_BR |
| Publicado em: |
Universidade Federal do Rio Grande do Norte
|
| Assuntos: | |
| Endereço do item: | https://repositorio.ufrn.br/handle/123456789/44630 |
| Tags: |
Adicionar Tag
Sem tags, seja o primeiro a adicionar uma tag!
|
| Resumo: | Deep learning is growing in popularity in virtual sensor modeling problems - the soft
sensors - applied to industrial processes of accentuated nonlinearity. Virtual sensors can
generate estimates of process variables, which are associated with quality indexes in realtime. Thus, such sensors are a viable alternative when the variables of interest are difficult
to measure due to some limiting factor: unavailability of hardware sensors or large measurement intervals. Traditional machine learning strategies show difficulties to model such
sensors. Typically, industrial processes are highly nonlinear, and the amount of available
labeled data is scarce. Due to that, the extraction of representative features present in the
abundant amount of unlabeled data has become an area of interest in the development of
virtual sensors. From the aforementioned premises, a new virtual sensor modeling technique based on deep learning and representation, which integrates stacked autoencoders
(SAE), mutual information (MI), long short-term memory (LSTM), and aggregation bootstrap, is proposed. First, in the unsupervised stage, the SAE structure is hierarchically
trained layer-by-layer. After a layer’s training, MI analysis is conducted between the target outputs of the model and the representations of the current layer to assess the learned
characteristics. The proposed method removes irrelevant information and weights the retained ones. The given weights being proportional to the relevance of the representation.
Therefore, this approach can extract deep representative information. In the supervised
step, called fine-tuning, an LSTM structure is coupled to the tail of the SAE structure
to address the intrinsic dynamic behavior of the evaluated industrial systems. Further,
a ensemble strategy, called bootstrap aggregation, combines the models obtained in the
supervised training phase to improve the performance and credibility of the virtual sensor. The proposal uses two industrial nonlinear processes, widely used as benchmarks,
to evaluate the performance of the models generated by the proposed technique in the
implementation of soft sensors. The results show that the proposed virtual sensors obtained better prediction performance than traditional methods and several state-of-the-art methods. |
|---|