Scribd
Upload
15 views2 pages

TP3 Commande MAC - MPC 2024

Uploaded by

Ghoual Bouziane
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
15 views2 pages

TP3 Commande MAC - MPC 2024

Uploaded by

Ghoual Bouziane
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Ahmed Zabana University Center Relizane

Department of Electrical Engineering


5th year Bachelor LMD Automation and Systems (Master 2)
Module: TP Predictive/Adaptive Control
Year 2024-2025
Instructor: Bouziane GHOUAL ([email protected])

TP N°3: Practical Comparison of


Adaptive Control and Predictive Control for Trajectory Tracking
The primary goal of this TP is to compare the performance of Adaptive Control and Model Predictive Control (MPC)
in tracking a sinusoidal reference trajectory for a nonlinear first-order system. Students will explore how Adaptive
Control uses real-time error correction to adjust the system output, while MPC predicts future states and optimizes
control inputs over a defined horizon (lambda). Through simulation and analysis, the session aims to highlight the
advantages of MPC in predictive capabilities and its robustness in handling dynamic systems.
• Theoretical Reminder:

1. Adaptive Control:
Adaptive control allows the controller to adjust in real-time to handle system uncertainties. Unlike traditional
fixed-gain controllers, adaptive controllers continuously adjust their parameters to ensure optimal performance.
MRAC is a specific approach where the goal is to make the plant's behavior match that of a reference model.
The reference model represents the desired response of the system. MRAC continuously adjusts the controller
parameters based on the error between the plant output and the reference model output.
2. Predictive Control:
Model Predictive Control (MPC) computes control inputs by predicting the future system behavior and
minimizing a cost function over a finite prediction horizon. The goal of MPC is to drive the system’s output to
track a desired trajectory while considering system dynamics and control constraints.

• Methodology:
1. System Model
The system is simulated over 𝑡 ∈ [0,10] seconds with a time step of 𝑑𝑡 = 0.01 The reference trajectory is:
𝑟(𝑡) = 𝑠𝑖𝑛(0.5𝑡)
2. Adaptive Control
The controller uses real-time error 𝑒(𝑡) = 𝑟(𝑡) − 𝑥(𝑡) to compute the control input:
𝑢𝑎𝑑𝑎𝑝𝑡𝑖𝑣𝑒 = 𝐾𝑎𝑑𝑎𝑝𝑡𝑖𝑣𝑒 ∙ 𝑒(𝑡).
𝐾𝑎𝑑𝑎𝑝𝑡𝑖𝑣𝑒 = 𝑘𝑎𝑑𝑎𝑝𝑡𝑖𝑣𝑒 + 𝑔𝑎𝑚𝑚𝑎 ∗ 𝑒(𝑡) ∗ 𝑢(𝑡) ∗ 𝑑𝑡
System dynamics are updated using the nonlinear equation.

3. Model Predictive Control (MPC)


MPC optimizes future control inputs over a prediction horizon (𝜆). The cost function incorporates tracking error
and control effort:
𝜆

𝐽 = ∑[𝑄𝑒(𝑡 + 𝑗)2 + 𝑅𝑢(𝑡 + 𝑗)2 ]


𝑗=1
where 𝑄 and 𝑅 are weighting factors for error and control effort. A simplified optimization loop compute
𝑢𝑝𝑟𝑒𝑑𝑖𝑐𝑡𝑖𝑣𝑒 iteratively.

• Step-by-Step Procedure:
Step 1: Generate Desired Sinusoidal Trajectory
Define the time vector and a sinusoidal trajectory as the desired output. This trajectory will be used as a reference
that the system aims to follow.

1
❖ Question 1: Make a plot figure for this system.

Step 2: Implement Adaptive Control for Tracking our system trajectory


We use in this part the Adaptive Control for tracking the system path and to be close to this trajectory.

❖ Question 2: Make a plot figure for this control including reference trajectory to compare.

Step 3: Implement Predictive Control for Tracking Control


In this part, we use Model Predictive Control to minimize the tracking error. At each time step, we compute the
control signal by adjusting it to reduce the tracking error.

❖ Question 4:
❖ Plot the figure between ‘Reference', 'Adaptive Control', and 'Predictive Control'.
❖ Explain the figure, which control is better for tracking trajectories.
2

You might also like