Process outputs, as reflected in the measured signals, are generally stochastic in nature. This chapter focuses on processing of sensor outputs and the identification of faults that may occur during laser processing, for example, cracking, porosity, no weld, and so on, and is limited to classification methods. Measured signals are often contaminated by noise. The Fourier transform involves the transformation of information from the time domain to the frequency domain. Data reduction is used to reduce the initial size of the data to make subsequent computations faster, with minimal loss of information. The chapter focuses on pattern recognition and neural network analysis. Feature extraction involves transformation of the signal, along with a reduction in the data size. Neural networks represent the classifier by nodes which are interconnected, with information available in the network being contained in the magnitude of the links between the processors.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Processing of Sensor Outputs


    Beteiligte:


    Erscheinungsdatum :

    07.02.2023


    Format / Umfang :

    30 pages




    Medientyp :

    Aufsatz/Kapitel (Buch)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch





    Integrated position sensor with dual synchronized outputs

    HANUS PETR / HAJEK ONDREJ / MANDEL ONDREJ et al. | Europäisches Patentamt | 2025

    Freier Zugriff

    INTEGRATED POSITION SENSOR WITH DUAL SYNCHRONIZED OUTPUTS

    HANUS PETR / HAJEK ONDREJ / MANDEL ONDREJ et al. | Europäisches Patentamt | 2024

    Freier Zugriff


    INTEGRATED POSITION SENSOR WITH DUAL SYNCHRONIZED OUTPUTS

    HANUS PETR / HAJEK ONDREJ / MANDEL ONDREJ et al. | Europäisches Patentamt | 2024

    Freier Zugriff