Data-Driven Control with MATLAB and Simulink

Design and implement data-driven and AI-based control algorithms

Engineers use data-driven control algorithms in scenarios where traditional control methods may fall short. These scenarios may occur when modeling plant dynamics using first principles is difficult or impractical, or when adaptive control is necessary.

With MATLAB and Simulink, you can:

Design, simulate, and implement data-driven control techniques using AI and non-AI-based methods
Identify system dynamics or learn controller parameters directly from data using offline techniques on your desktop
Update controller parameters in real-time within embedded systems using online techniques

	Offline Techniques	Online Techniques
AI-Based	Model predictive control (MPC) with neural state-space Offline reinforcement learning Learn more	Online reinforcement learning Model reference adaptive control with AI-based disturbance model Learn more
Non-AI-Based	Traditional methods with system identification Fuzzy inference system (FIS) tuning Learn more	Active disturbance rejection control Model reference adaptive control with non-AI-based disturbance model Extremum seeking control Closed-Loop PID Autotuner Adaptive MPC with online system identification Learn more

Why MATLAB and Simulink?

Compare and combine various data-driven and traditional control techniques

Design, simulate, and implement model reference adaptive control (MRAC), active disturbance rejection control (ADRC), reinforcement learning (RL), model predictive control (MPC), and other data-driven and traditional control methods within a single environment.

Adaptive Control Design in Simulink

Closed-Loop PID Autotuner

Deep Reinforcement Learning for Control and Decision-Making Applications

Data-Driven Control Tech Talks (3 videos)

Control House Heating System Using Nonlinear MPC with Neural State-Space Prediction Model

Integrate with Simulink and Model-Based Design

Implement and test data-driven control algorithms in Simulink using pre-built Simulink blocks. Automatically generate code from your control algorithm for direct deployment on embedded hardware.

Implement PMSM Speed Control Using Active Disturbance Rejection Control

Deploy Trained Reinforcement Learning Policies

Get started with reference examples

Utilize documented references and examples for flight control, robotics, energy management, and other applications to implement data-driven control techniques without starting from scratch.

Adaptive Control Design Examples

Automated Driving Applications of Model Predictive Control

Energy Management Using Deep Learning-Based MPC (8:09)

Design Controller for Artificial Pancreas Using Fuzzy Logic