Pulse type recognition using a neural network
Students Name: Tkhor Sofiia Kostiantynivna
Qualification Level: magister
Speciality: Informatively-Instrumentation Technologies in Roboticotekhnic
Institute: Institute of Computer Technologies, Automation and Metrology
Mode of Study: full
Academic Year: 2021-2022 н.р.
Language of Defence: ukrainian
Abstract: In this master’s qualification work a system for recognizing the waveforms of measurement signals using a neural network was created, measurement signal models for neural network training were developed, the research and optimization by the criterion of minimum measurement error in measurement signal recognition was conducted, and the instability of measurement recognition results was investigated. An analysis of the modern system of weighing in motion - WIM (weight-in-motion) was conducted. The study is focused on: - the dependence of measurement signal recognition error using a neural network on the number of training pairs for its training; - the dependence measurement signal recognition error using a neural network on the number of inner layers of the neural network; - the dependence of measurement signal recognition error using a neural network on the level of interference that comes with the signal. The results of the study are presented in the form of tables and graphs. The measurement error study was performed in the Matlab software.