7.7.19. Induction Motor Speed Prediction
Predict induction motor speed from electrical signals.
7.7.19.1. Overview
This example demonstrates speed estimation for induction motors using electrical measurements. Sensorless speed estimation eliminates the need for mechanical speed sensors, reducing cost and improving reliability.
Application: Industrial motor drives, pumps, fans, compressors
Task Type: Time Series Regression
Data Type: Multivariate (voltage and current signals)
Induction Motor Relationships:
Relationships between electrical parameters and motor speed
7.7.19.2. Configuration
common:
target_module: 'timeseries'
task_type: 'generic_timeseries_regression'
target_device: 'F28P55'
dataset:
dataset_name: 'induction_motor_speed_prediction'
training:
model_name: 'REGR_4k_NPU'
training_epochs: 50
batch_size: 32
testing: {}
compilation: {}
7.7.19.3. Running the Example
cd tinyml-modelzoo
./run_tinyml_modelzoo.sh examples/induction_motor_speed_prediction/config.yaml
cd tinyml-modelzoo
run_tinyml_modelzoo.bat examples\induction_motor_speed_prediction\config.yaml
7.7.19.4. Dataset Details
Input Variables:
Phase voltages
Phase currents
DC bus voltage (optional)
Output:
Motor speed (continuous value in RPM or rad/s)
7.7.19.5. See Also
Torque Measurement Regression - Torque estimation