Mohamed Elwalid Fadul
← All projects

Case study 07

PI Control of Dam Water Level

An academic control-engineering project in TTK4105 at NTNU, developing a closed-loop PI controller for water-level regulation in a simplified dam model under varying inflow and measurement noise.

GitHubReport
Discipline
CONTROL ENGINEERING / SYSTEM MODELLING / SIMULINK
Course
TTK4105 — Control Engineering
Project type
academic
Period
Spring 2025
Evidence
Control code + Technical report
Team
4 mechanical engineering students
My role
System modelling, PI control and simulation
Simplified reservoir and dam system with inflow, water level and controlled outflow
System illustration from the submitted report showing the modelled inflow, reservoir level and controlled outlet flow.
Reconstructed Simulink model for PI water-level control
Executable Simulink reconstruction of the reservoir mass balance, PI controller, actuator limits and feedback loop.
Nominal PI water-level regulation response
Nominal 1000-second simulation showing convergence to the 5 m reference and equilibrium outlet command.
Water-level response to piecewise-constant inflow disturbances
Seeded inflow-disturbance test with level, control signal, inflow and accumulated IAE.
Water-level response with reconstructed measurement noise
Reconstructed measurement-noise scenario; the assumed noise parameters are documented separately from the original report values.

01 / Engineering challenge

The system behind the project

Reliable water-level control is a representative process-control problem involving mass balance, feedback, actuator behaviour, external disturbances and uncertain measurements. The work connects core control theory to infrastructure and industrial fluid systems.

02 / My responsibility

What I personally worked on

Worked in a four-person mechanical-engineering team to formulate the dynamic model, develop and tune the PI controller, build the Simulink implementation and evaluate the response under disturbances and sensor noise.

Engineering focus

  • — First-order dynamic modelling
  • — PI control and pole placement
  • — Disturbance and noise response
  • — Performance evaluation using IAE

Tools and methods

MATLABSimulinkControl EngineeringSystem ModellingLaTeX

03 / Design and implementation

From concept to working system

  1. 01

    Derive a first-order dynamic model from the reservoir mass balance and linearise it around an operating point.

  2. 02

    Select a PI structure to eliminate steady-state error without unnecessarily amplifying measurement noise.

  3. 03

    Tune the proportional and integral gains using pole placement and implement the closed loop in Simulink.

  4. 04

    Test the system with varying inflow, band-limited measurement noise and the Integral of Absolute Error performance metric.

04 / Testing and outcome

What the work demonstrated

The portfolio reconstruction reproduces the documented model and controller in executable MATLAB and Simulink files. Across seeded review scenarios, the level remained between 4.895 m and 5.128 m; peak absolute error was 0.128 m for the inflow-disturbance case. These figures verify the reconstructed conceptual model, not the behaviour or safety of a physical dam.

05 / Next iteration

How I would develop it further

  • 1Introduce feedforward compensation using measured or estimated inflow
  • 2Evaluate adaptive control for changing reservoir dynamics
  • 3Add actuator dynamics, time delay and anti-windup behaviour
  • 4Extend the model with nonlinear geometry and improved sensor filtering
GitHubReport