MATLAB Modeling & Simulation
This architecture blueprint demonstrates how you can:
- Seamlessly integrate MATLAB/Simulink models with real-time data streams
- Automate simulation workflows with cloud-based execution
- Scale complex simulations across distributed computing resources
- Maintain version control and reproducibility of simulation results
- Visualize and analyze simulation outputs in real-time dashboards
Main project components
MQTT Sources
Collect data from IoT devices and sensors using the MQTT protocol.
Connected devices should publish to MQTT brokers which are then connected to Quix, allowing rapid processing and analytics.
Phasor Measurement Unit (PMU) Data:
PMU data typically requires microsecond precision, which MQTT doesn't guarantee.
Configuration
Collecting real or simulated configuration data enables model behavior to be altered at run-time.
Downsample
Reduce the raw data rate to a manageable 10-20 Hz and filter any common noise values.
JOIN
Configuration, BMS and PMU data is joined and formatted, providing the model with exactly the data it needs to run.
MATLAB / Simulink
Upload your MATLAB and Simulink models to the Quix platform where they can be repeatedly run with the required resources and data.
Store
Publish model outputs, errors and telemetry data to a data store for later analytics and comparison against other test runs.
Technologies used
Pandas - Data manipulation and analysis
scikit-learn - Machine learning
TensorFlow - Deep learning
MATLAB - Core platform
Simulink - Model-based design
Simscape Electrical - Power systems modeling
MATLAB Parallel Server - Parallel computing
MQTT - Lightweight messaging
OPC UA - Industrial communication
Using this template
