My SciELO
Services on Demand
Article
Indicators
- Cited by SciELO
Related links
- Similars in SciELO
Share
Ingeniería Electrónica, Automática y Comunicaciones
On-line version ISSN 1815-5928
Abstract
VEGA-GONZALEZ, Iliover; PUENTES FIGUEROA, Omar; TRUTIE-CARRERO, Eduardo and RAMIREZ-BELTRAN, Jorge. Algorithm to detect bursts in water pipes for implementation in low-power devices. EAC [online]. 2022, vol.43, n.1, pp. 47-64. Epub June 10, 2022. ISSN 1815-5928.
The current work presents a new algorithm to detect bursts in water pipes automatically, focused on its implementation in low-power embedded devices. The algorithm was implemented in an MSP430FR5994 microcontroller and it does not require computer devices with the processing capability to execute high-complexity mathematical operations, unlike most of the existing algorithms. The developed algorithm uses the approach of detecting bursts by enhancing the singularities they create and making decisions based on the application of thresholds. Instead of applying a single mathematically-complex threshold, two simpler ones are used that are compatible with the processing capability of low-power embedded devices. Two implementations for the convolution operation were tested using the microcontroller’s Low Energy Accelerator module; it was found that the convolution in the time domain is faster than the FFT-based method in an average of 65.3%. The algorithm detects bursts with a detection probability of 100% and a probability of false positives of 0.13% on the performed tests.
Keywords : low power; burst detection; MSP430 microcontroller; Wavelet Transform; Low Energy Accelerator module.