Perangkat Pemantauan Stres Non-Kontak Menggunakan Radar Continous Wave dan Mikrofon

Tania, Mavelyn Clarissa (2025) Perangkat Pemantauan Stres Non-Kontak Menggunakan Radar Continous Wave dan Mikrofon. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5023211004-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Stres adalah kondisi atau reaksi tubuh terhadap tekanan dari lingkungan sekitar. Stres yang terjadi terus-menerus dapat memengaruhi kesehatan mental dan fisik. Stres dapat diidentifikasi melalui pengisian kuesioner atau konseling, tetapi dapat dipengaruhi oleh berbagai macam faktor. Stres juga dapat diidentifikasi dengan electrocardiogram (ECG) dengan menganalisis fitur detak jantung. Fitur detak jantung dan pernapasan memiliki hubungan dengan kondisi bawah sadar terkait aktivitas saraf simpatis dan parasimpatis yang dapat digunakan untuk memantau stres. Namun, ECG hanya dapat mengukur detak jantung secara kontak, sehingga dapat menyebabkan ketidaknyamanan. Perangkat pemantauan stres secara non-kontak dapat dikembangkan menggunakan radar dan mikrofon. Teknologi radar dapat digunakan untuk menggantikan ECG dengan mendeteksi detak jantung dan pernapasan dari jarak tertentu berdasarkan pergerakan dada. Tingkat stres juga dapat mempengaruhi artikulasi saat berbicara. Sinyal radar diproses dengan algoritma Ensemble Empirical Mode Decomposition (EEMD) untuk memisahkan sinyal detak jantung dengan pernapasan, sedangkan sinyal vokal dapat dipisahkan dari suara lingkungan dengan metode Voice Activity Detectiond (VAD). Dilakukan ekstraksi fitur dan analisis fitur pada sinyal detak jantung, pernapasan, dan suara agar dapat digunakan dalam sistem klasifikasi Late Fusion dengan LightGBM dan Logistic Regression dalam membedakan kondisi stres rendah, sedang, dan tinggi. Kombinasi LightGBM dan Logistic Regression unggul dalam klasifikasi stres dengan akurasi 85,7%.
===========================================================================================================================================================
Stress is a condition or reaction of the body towards pressure from the surrounding environment. Prolonged stress may affect mental and physical health. Stress can be identified through questionnaire or counselling but can be influenced by a combination of factors. Stress may also be identified by electrocardiogram (ECG) through analysing heart rate features. Heart rate and breathing features have a relationship with subconscious states related to sympathetic and parasympathetic nerve activity which can be used to monitor stress. However, ECG can only measure heart rate by contact, which may cause discomfort. A non-contact stress monitoring device can be developed using a radar and microphone. Radar technology can be used to replace ECG by detecting heart rate and respiratory rate from a certain distance based on chest movement. Stress levels can also affect articulation when speaking. Radar signals are processed with the Ensemble Empirical Mode Decomposition (EEMD) algorithm to separate heartbeat signals from breathing, while vocal signals can be separated from environmental noises with the Voice Activity Detection (VAD) method. Feature extraction and feature analysis are implemented on heart rate, respiration, and voice signals in order to be used in the Late Fusion classification system with LightGBM and Logistic Regression in discriminating low, medium, and high stress states. The combination of LightGBM and Logistic Regression excelled in stress prediction with an 85.7% accuracy.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Late Fusion, Mikrofon, Non-Kontak, Radar Continous Wave, Stres, Continous Wave Radar, Late Fusion, Microphone, Non-Contact, Stress
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7871.674 Detectors. Sensors T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7872 Electromagnetic Devices T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.S65 Automatic speech recognition.
Divisions:	Faculty of Electrical Technology > Biomedical Engineering > 11410-(S1) Undergraduate Thesis
Depositing User:	Mavelyn Clarissa Tania
Date Deposited:	04 Aug 2025 08:11
Last Modified:	04 Aug 2025 08:11
URI:	http://repository.its.ac.id/id/eprint/125324

Actions (login required)

View Item