Homological Time Series Analysis of Sensor Signals from Power Plants

Abstract

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.

arXiv commentCode available at https://codeberg.org/Jiren/TwirlFlake

References in corpus (8)

• cuDNN: Efficient Primitives for Deep Learning
• Searching for Activation Functions
• Statistical topological data analysis using persistence landscapes
• Dualities in persistent (co)homology
• giotto-tda: A Topological Data Analysis Toolkit for Machine Learning and Data Exploration
• State Space Representations of Deep Neural Networks
• Persistent Homology as Stopping-Criterion for Voronoi Interpolation
• Estimate of the Neural Network Dimension using Algebraic Topology and Lie Theory