Adaptive Control for Tank Clogging Detection

ISEF Category: Robotics and Intelligent Machines

Subcategory: Control Theory  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

A clogged pipe can fool a controller faster than you can notice the water level change. That makes this project feel like a tiny version of a real plant failure. You build a system that senses the problem, estimates the clog, and reacts before the tanks drift out of range.

What Is It?

This project studies a coupled water system, where two tanks affect each other through flow lines and valves. Think of it like two connected buckets with a pump feeding them. If one valve starts to clog, the whole system changes behavior. Your controller has to notice that shift and update its model in real time.

The key idea is adaptive control. That means the controller does not assume the system stays the same. Instead, it keeps estimating parameters, like flow resistance or effective orifice size, while the system runs. In your setup, adjustable clamps can create known restrictions, so you can compare the controller’s estimate with the real valve opening. That gives you a clear way to test whether the algorithm can detect a changing physical fault.

Why This Is a Good Topic

This is a strong science fair topic because it mixes a real engineering problem with measurable math. You can change one thing at a time, like valve restriction, pump speed, or tank coupling, and see how the estimates shift. The project connects to industrial process control, leak detection, and fault diagnosis. You will learn system modeling, sensor calibration, parameter estimation, and how to judge whether a controller works under changing conditions.

Research Questions

• How does changing valve restriction affect the accuracy of adaptive parameter estimates?
• What is the effect of pump speed on how quickly the controller detects clogging?
• Does adding a second tank improve or worsen fault detection compared with a single-tank model?
• To what extent does sensor noise change the stability of the adaptive control loop?
• Which parameter estimation method tracks a time-varying orifice diameter most accurately?
• How does the placement of the level sensor affect detection delay and estimation error?

Basic Materials

• PVC tanks or clear containers with similar geometry.
• Peristaltic pumps.
• Flexible tubing sized to the pumps and valves.
• Adjustable clamps or pinch valves.
• Water level sensors or pressure sensors.
• Microcontroller or data acquisition board.
• Breadboard and jumper wires.
• Power supply matched to the pumps.
• Measuring cup or graduated cylinder.
• Digital kitchen scale with 0.1 g accuracy.
• Ruler or meter stick.
• Waterproof marker for tank labels.
• Laptop for logging and analysis.

Advanced Materials

• Transparent acrylic tank assembly.
• Industrial-grade peristaltic pumps with controlled drive speed.
• Differential pressure sensors or ultrasonic level sensors.
• Precision flow meters.
• Solenoid valves for scripted restriction changes.
• Adjustable orifice inserts with known diameters.
• Signal conditioning board.
• Microcontroller with analog input and PWM output.
• Data acquisition system.
• Reference gauges for pressure and flow.
• Calibration weights or volumetric standards.
• Lab power supply.
• Safety spill tray.

Software & Tools

• Python: Plots time series, fits parameters, and compares adaptive control methods.
• MATLAB: Runs state-space models and controller simulations if your lab has access.
• ImageJ: Measures clamp position or visible fluid levels from recorded video.
• Excel: Organizes calibration data and basic error checks.
• Jupyter Notebook: Keeps your code, graphs, and notes in one place.

Experiment Steps

1. Define the state variables, the measured signals, and the one fault you want the controller to detect first.
2. Build a mathematical model of the coupled tanks and decide which parameters the adaptive algorithm should estimate.
3. Plan a calibration method that turns raw sensor readings into level or flow values you can trust.
4. Design a set of known restriction states using adjustable clamps or orifice inserts so you have ground truth.
5. Choose the error metric, the baseline controller, and the adaptive controller so you can compare them fairly.
6. Map out the test matrix that changes one condition at a time and separates normal variation from clogging.

Common Pitfalls

• Ignoring sensor drift, which makes a healthy system look like it is clogging.
• Using a clamp position as ground truth without checking the actual orifice diameter or flow change.
• Tuning the controller on one pump speed, then finding that the model fails at another speed.
• Letting bubbles or trapped air distort the water level reading and confuse the estimator.
• Changing several variables at once, which makes it impossible to tell whether the detection came from clogging, pump variation, or coupling changes.

What Makes This Competitive

A strong version of this project does more than show a working tank demo. You would compare at least one adaptive method against a fixed baseline, then test both under changing fault sizes and changing flow conditions. Better entries also report estimation error, detection delay, and stability, not just whether the level looked right. If you can validate your estimates against measured orifice size and show where the method breaks, your project starts to look like real control research.

Project Variations

• Test whether the same adaptive controller can detect a slow leak instead of a clog.
• Compare clamp-based restrictions with interchangeable orifice inserts to see which fault source gives cleaner parameter estimates.
• Add a disturbance, like a sudden pump speed change, and measure whether the controller can still separate disturbance from clogging.

Learn More

• MIT OpenCourseWare, Control Systems: Search MIT OpenCourseWare for lectures and notes on feedback, state-space models, and adaptive control.
• NIST Engineering Statistics Handbook: Use the sections on regression, uncertainty, and model error, found through the NIST website.
• NASA Systems Engineering Handbook: Read the control and verification sections, available through NASA technical reports and handbooks.
• IEEE Control Systems Magazine: Search for review articles on adaptive control, fault detection, and parameter estimation through your school library or PubMed-style journal access tools.
• PubMed: Search for review papers on fault detection in fluid systems and adaptive estimation methods.