This work aims to propose an original method for representing signals collected by position and acceleration sensors and their processing in order to obtain a reliable reconstruction of the trajectory performed. The double benefit obtained consists in the improvement of the estimation of the position reached by controlled axes and in the reduction of the data to represent the available information. Various approximation algorithms using B-spline have been tested through numerical tests. A data fusion algorithm was then identified capable of combining several independent sources, profoundly different in their characteristics (frequency, spectral density, signal / noise ratio), paying particular attention to eliminating drift error caused by the integration. The research was completed with a first phase of experimental validation which resulted in the fusion of signals from three different sources (encoder, camera and accelerometers). Copyright (c) 2023 The Authors.
A Data Fusion Algorithm for Digital Twin of Controlled Axes: Development, Implementation and Test on Machine / Silvestri, M.; Poloni, D.; Riboli, M.; Corradini, F.. - ELETTRONICO. - 56:2(2023), pp. 11900-11905. (Intervento presentato al convegno 22nd IFAC World Congress tenutosi a Yokohama, Japan nel July 9-14, 2023) [10.1016/j.ifacol.2023.10.603].
A Data Fusion Algorithm for Digital Twin of Controlled Axes: Development, Implementation and Test on Machine
Silvestri M.
Writing – Original Draft Preparation
;Riboli M.Methodology
;
2023-01-01
Abstract
This work aims to propose an original method for representing signals collected by position and acceleration sensors and their processing in order to obtain a reliable reconstruction of the trajectory performed. The double benefit obtained consists in the improvement of the estimation of the position reached by controlled axes and in the reduction of the data to represent the available information. Various approximation algorithms using B-spline have been tested through numerical tests. A data fusion algorithm was then identified capable of combining several independent sources, profoundly different in their characteristics (frequency, spectral density, signal / noise ratio), paying particular attention to eliminating drift error caused by the integration. The research was completed with a first phase of experimental validation which resulted in the fusion of signals from three different sources (encoder, camera and accelerometers). Copyright (c) 2023 The Authors.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
