Evolutionary Optimization of Spiking Neural P Systems for Remaining Useful Life Prediction
Remaining useful life (RUL) prediction is a key enabler for predictive maintenance. In fact, the possibility of accurately and reliably predicting the RUL of a system, based on a record of its monitoring data, can allow users to schedule maintenance interventions before faults occur. In the recent l...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-03-01
|
| Series: | Algorithms |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1999-4893/15/3/98 |
| _version_ | 1797473274286833664 |
|---|---|
| author | Leonardo Lucio Custode Hyunho Mo Andrea Ferigo Giovanni Iacca |
| author_facet | Leonardo Lucio Custode Hyunho Mo Andrea Ferigo Giovanni Iacca |
| author_sort | Leonardo Lucio Custode |
| collection | DOAJ |
| description | Remaining useful life (RUL) prediction is a key enabler for predictive maintenance. In fact, the possibility of accurately and reliably predicting the RUL of a system, based on a record of its monitoring data, can allow users to schedule maintenance interventions before faults occur. In the recent literature, several data-driven methods for RUL prediction have been proposed. However, most of them are based on traditional (connectivist) neural networks, such as convolutional neural networks, and alternative mechanisms have barely been explored. Here, we tackle the RUL prediction problem for the first time by using a membrane computing paradigm, namely that of Spiking Neural P (in short, SN P) systems. First, we show how SN P systems can be adapted to handle the RUL prediction problem. Then, we propose the use of a neuro-evolutionary algorithm to optimize the structure and parameters of the SN P systems. Our results on two datasets, namely the CMAPSS and new CMAPSS benchmarks from NASA, are fairly comparable with those obtained by much more complex deep networks, showing a reasonable compromise between performance and number of trainable parameters, which in turn correlates with memory consumption and computing time. |
| first_indexed | 2024-03-09T20:12:26Z |
| format | Article |
| id | doaj.art-39ecf9505ddc40cfb6aa2d829a1fdebf |
| institution | Directory Open Access Journal |
| issn | 1999-4893 |
| language | English |
| last_indexed | 2024-03-09T20:12:26Z |
| publishDate | 2022-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Algorithms |
| spelling | doaj.art-39ecf9505ddc40cfb6aa2d829a1fdebf2023-11-24T00:09:05ZengMDPI AGAlgorithms1999-48932022-03-011539810.3390/a15030098Evolutionary Optimization of Spiking Neural P Systems for Remaining Useful Life PredictionLeonardo Lucio Custode0Hyunho Mo1Andrea Ferigo2Giovanni Iacca3Department of Information Engineering and Computer Science, University of Trento, 38123 Trento, ItalyDepartment of Information Engineering and Computer Science, University of Trento, 38123 Trento, ItalyDepartment of Information Engineering and Computer Science, University of Trento, 38123 Trento, ItalyDepartment of Information Engineering and Computer Science, University of Trento, 38123 Trento, ItalyRemaining useful life (RUL) prediction is a key enabler for predictive maintenance. In fact, the possibility of accurately and reliably predicting the RUL of a system, based on a record of its monitoring data, can allow users to schedule maintenance interventions before faults occur. In the recent literature, several data-driven methods for RUL prediction have been proposed. However, most of them are based on traditional (connectivist) neural networks, such as convolutional neural networks, and alternative mechanisms have barely been explored. Here, we tackle the RUL prediction problem for the first time by using a membrane computing paradigm, namely that of Spiking Neural P (in short, SN P) systems. First, we show how SN P systems can be adapted to handle the RUL prediction problem. Then, we propose the use of a neuro-evolutionary algorithm to optimize the structure and parameters of the SN P systems. Our results on two datasets, namely the CMAPSS and new CMAPSS benchmarks from NASA, are fairly comparable with those obtained by much more complex deep networks, showing a reasonable compromise between performance and number of trainable parameters, which in turn correlates with memory consumption and computing time.https://www.mdpi.com/1999-4893/15/3/98Spiking Neural P SystemsNEATRemaining Useful Lifepredictive maintenanceCMAPSS |
| spellingShingle | Leonardo Lucio Custode Hyunho Mo Andrea Ferigo Giovanni Iacca Evolutionary Optimization of Spiking Neural P Systems for Remaining Useful Life Prediction Algorithms Spiking Neural P Systems NEAT Remaining Useful Life predictive maintenance CMAPSS |
| title | Evolutionary Optimization of Spiking Neural P Systems for Remaining Useful Life Prediction |
| title_full | Evolutionary Optimization of Spiking Neural P Systems for Remaining Useful Life Prediction |
| title_fullStr | Evolutionary Optimization of Spiking Neural P Systems for Remaining Useful Life Prediction |
| title_full_unstemmed | Evolutionary Optimization of Spiking Neural P Systems for Remaining Useful Life Prediction |
| title_short | Evolutionary Optimization of Spiking Neural P Systems for Remaining Useful Life Prediction |
| title_sort | evolutionary optimization of spiking neural p systems for remaining useful life prediction |
| topic | Spiking Neural P Systems NEAT Remaining Useful Life predictive maintenance CMAPSS |
| url | https://www.mdpi.com/1999-4893/15/3/98 |
| work_keys_str_mv | AT leonardoluciocustode evolutionaryoptimizationofspikingneuralpsystemsforremainingusefullifeprediction AT hyunhomo evolutionaryoptimizationofspikingneuralpsystemsforremainingusefullifeprediction AT andreaferigo evolutionaryoptimizationofspikingneuralpsystemsforremainingusefullifeprediction AT giovanniiacca evolutionaryoptimizationofspikingneuralpsystemsforremainingusefullifeprediction |