🌡️ PID Temperature Control of a CSTR (MATLAB & Simulink)

I built this simulation to model how a PID controller maintains a steady temperature in a Continuous Stirred Tank Reactor (CSTR). In chemical engineering, precise temperature control is the difference between a successful reaction and a ruined batch. This project focuses on disturbance rejection—specifically how the system recovers when things go wrong.

---

🛠️ The Control Strategy

I modeled the CSTR process as a first-order system with a 10s time constant. The target was a constant 50°C setpoint.

🧠 PID Tuning (Manual Logic)

I didn't use auto-tune; I set these gains manually to understand the trade-offs:

• Kp = 2: Gets the temperature moving toward the setpoint quickly.
• Ki = 0.1: Wipes out the steady-state error so the temp actually hits 50°C and stays there.
• Kd = 0.5: Acts as a "damper" to prevent the system from overshooting the target and oscillating.

🧱 Simulink Model Breakdown

I built the feedback loop with specific blocks to mimic real-world hardware constraints:

1. Sum Block: Continuously calculates the Error ($Setpoint - Actual$).
2. Saturation Block: Crucial Step. I limited the Heat Input ($Q$) between 0 and 10. Real heaters don't have infinite power, so I capped the controller's output to keep the simulation realistic.
3. Step Block: Used to inject a -5 unit cooling disturbance at the 100-second mark to test if the controller could recover from a sudden external "shock."

---

📊 Results: Handling the Disturbance

The simulation ran for 200 seconds. Here’s what happened on the Scope:

• 0s - 50s: The temperature rose steadily and settled at 50°C with almost zero overshoot.
• 100s Mark: The cooling disturbance hit. The temperature dropped immediately.
• 100s - 150s: The PID controller "saw" the drop and ramped up the heat input. The system fought back and returned to 50°C within 30 seconds.

---

💡 What I Actually Learned

• Physical Limits: Adding the Saturation Block changed everything. It taught me that a PID controller is only as good as the hardware (the heater) it's controlling.
• Disturbance vs. Setpoint: It’s easy to stay at a setpoint in a vacuum. The real test of an engineer is how the system handles a "hit" like the cooling disturbance I added at 100s.
• Visualizing Logic: Seeing the signal flow in Simulink made the math from my MATLAB scripts much easier to troubleshoot.

---

📂 Project Structure

• cstr_pid_model.slx: The main Simulink block diagram.
• cstr_pid_control.mlx: MATLAB Live Script for gain initialization and plotting.
• scope_output.png: The final temperature vs. time graph.

⚙️ How to Run

1. Open MATLAB and run cstr_pid_control.mlx to load the variables.
2. Open the cstr_pid_model.slx file.
3. Click Run and open the Scope block to see the live response.

---

👩‍🔬 Contact

Baljit Kaur LinkedIn | bk5721333@gmail.com