Project, 2022.12–2023.11

Development of an Intelligent Motor Control Algorithm Based on Customer Environments

This LG Electronics industry-academia project (Dec 2022 to Nov 2023) started from a practical problem: washer and dryer motors run on fixed PID parameters that take time and engineering effort to tune and cannot adapt to customer conditions such as temperature, humidity, and load. The team connected a MATLAB/Simulink motor simulator to a Python reinforcement learning model and built an adaptive motor control system in which LSTM-TD3 sets the speed and sensorless control gains dynamically. After the interim review, a second theme was added: an automated substitute-load compensation system meant to cut down the repeated experiments caused by PCB variation. Joonho Seo took part as a participating researcher. At the final review the project reported both KPIs as achieved (adaptive motor control with shorter response time and lower sensorless error; substitute-load compensation with current error within 0.1 A), including a 0.03 s (62%) reduction in time to target speed at 50 RPM and a 2.5 RPM (50%) reduction in sensorless error at 100 RPM.

Where

LG Electronics industry-academia project

My role

Participating researcher

Status

completed

2 KPIs reported as achieved (adaptive motor control; substitute-load compensation with current error within 0.1 A)