Skripsi

IMPLEMENTASI PENGENALAN PERINTAH SUARA BERBASIS DEEP LEARNING PADA ROBOT FORKLIFT MINI

---

Interaction between humans and machines is key to optimizing operational activities involving the movement of goods in industrial environments such as warehouses. The voice control system requires performance that is highly dependent on the reliability of the voice recognition model used. This study aims to develop and analyze the performance of deep learning architectures, Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), and Wav2Vec, for real-time voice control systems on mini forklift robots. To improve the model's ability to handle voice variations, the optimal model was compiled and selected, which had resource efficiency that impacted hardware memory. The results of voice data training for the CNN model achieved a training accuracy of 93.64% (loss: 0.1096) compared to LSTM with an accuracy of 89.4% (loss: 0.307). Compared to the pretrained Wav2Vec model, which has neural network weights created by Meta Facebook, after fine tuning the voice command data of the forklift robot, it produced a training accuracy of 97.14% (loss 0.6561). Real-time testing was conducted with several testers of different genders, resulting in success rates of 77.78% for CNN, 80% for LSTM, and 84.44% for Wav2Vec. RAM usage measurements also showed that LSTM used 292MB, which was lighter than CNN at 364MB and Wav2Vec at 623MB when applied to robot control hardware. This demonstrates that the application of deep learning in forklift robots for voice command recognition is very effective and responds well to commands.

---

Availability

Detail Information

Series Title

-

Call Number

T1866532025

Publisher

Indralaya : Prodi Teknik Elektro, Fakultas Teknik Universitas Sriwijaya., 2025

Collation

xvii, 122 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

629.807

Content Type

Text

Media Type

unmediated

Carrier Type

other (computer)

Edition

-

Subject(s)

Prodi Teknik Elektro
Teknik Kontrol

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment