Homological Time Series Analysis of Sensor Signals from Power Plants

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Melodia, Luciano, Lenz, Richard
Format: Preprint
Published: 2021
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918402189164544
author Melodia, Luciano
Lenz, Richard
author_facet Melodia, Luciano
Lenz, Richard
contents In this paper, we use topological data analysis techniques to construct a suitable neural network classifier for the task of learning sensor signals of entire power plants according to their reference designation system. We use representations of persistence diagrams to derive necessary preprocessing steps and visualize the large amounts of data. We derive deep architectures with one-dimensional convolutional layers combined with stacked long short-term memories as residual networks suitable for processing the persistence features. We combine three separate sub-networks, obtaining as input the time series itself and a representation of the persistent homology for the zeroth and first dimension. We give a mathematical derivation for most of the used hyper-parameters. For validation, numerical experiments were performed with sensor data from four power plants of the same construction type.
format Preprint
id arxiv_https___arxiv_org_abs_2106_02493
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle Homological Time Series Analysis of Sensor Signals from Power Plants
Melodia, Luciano
Lenz, Richard
Machine Learning
Signal Processing
In this paper, we use topological data analysis techniques to construct a suitable neural network classifier for the task of learning sensor signals of entire power plants according to their reference designation system. We use representations of persistence diagrams to derive necessary preprocessing steps and visualize the large amounts of data. We derive deep architectures with one-dimensional convolutional layers combined with stacked long short-term memories as residual networks suitable for processing the persistence features. We combine three separate sub-networks, obtaining as input the time series itself and a representation of the persistent homology for the zeroth and first dimension. We give a mathematical derivation for most of the used hyper-parameters. For validation, numerical experiments were performed with sensor data from four power plants of the same construction type.
title Homological Time Series Analysis of Sensor Signals from Power Plants
topic Machine Learning
Signal Processing
url https://arxiv.org/abs/2106.02493