An interactive MATLAB App Designer simulation of a PID-controlled autonomous vehicle longitudinal speed system. Features real-time 3D vehicle animation, live performance metrics, and adjustable controller gains all within a custom-built dark-themed dashboard UI.
The full dashboard: PID gain sliders, speed response plot, 3D vehicle animation, semicircular gauge, and live performance metrics panel.
This project was developed for MCE 532 — Control Systems Engineering at the Federal University of Technology, Minna (FUT Minna). It implements a complete closed-loop PID speed controller for a simulated autonomous vehicle modelled as a First-Order Plus Time Delay (FOPTD) plant.
The app was built entirely in MATLAB App Designer and runs as a standalone .mlapp file. The user can interactively tune PID gains via sliders or numeric fields, set a reference speed via a knob, and immediately observe the closed-loop step response, 3D vehicle animation, and computed performance metrics — all updating in real time.
- Model an autonomous vehicle's longitudinal dynamics as an FOPTD transfer function
- Implement a discrete-time PID controller using Euler integration
- Tune the controller using the Ziegler-Nichols step response method
- Build a professional interactive MATLAB GUI with real-time animation and metrics
- Evaluate controller performance: rise time, settling time, overshoot, and SSE
The vehicle's throttle-to-speed dynamics are modelled as a First-Order Plus Time Delay (FOPTD) system:
K · e^(-Td·s) 5 · e^(-0.3s)
G(s) = ─────────────── = ─────────────────
τs + 1 2s + 1
K = 5.0 (process gain — speed per unit throttle)
τ = 2.0 s (time constant — vehicle inertia)
Td = 0.3 s (transport delay — actuator response lag)
A discrete-time PID controller implemented via Euler integration:
u(t) = Kp·e(t) + Ki·∫e(t)dt + Kd·de(t)/dt
Ziegler-Nichols Tuned Gains:
Kp = 1.17
Ki = 0.57
Kd = 0.27
The continuous-time plant is discretised at dt = 0.05 s (20 Hz):
dy/dt = [-y(t) + K · u_delayed(t)] / τ → y[k] = y[k-1] + dydt · dt
Time delay implemented as a circular shift buffer of length = round(Td/dt) = 6 steps
| Panel / Component | Function |
|---|---|
| PID Gains Panel | Kp, Ki, Kd sliders (range: 0–10, 0–5, 0–2) with live numeric edit fields |
| Reference Speed Knob | Sets Vref (10–30 m/s); simulation re-runs instantly on change |
| Speed Response Plot | Live animated step response: reference (red dashed) vs output (blue) |
| 3D Vehicle Animation | Real-time 3D sedan driving down a road; speed-accurate distance tracking |
| Semicircular Gauge | Displays final controlled speed (0–30 m/s) |
| Simulation Time Slider | Read-only progress bar showing simulation playback position |
| Performance Metrics | Rise Time / Settling Time / Overshoot % / SSE — colour-coded |
| Status Lamp | Green = running, Red = stopped |
| Start/Stop Button | Toggle simulation on/off |
| Comments Panel | Displays plant parameters, controller type, and tuned gains |
| Metric | Target | Achieved |
|---|---|---|
| Rise Time | < 2.0 s | ~1.4 s |
| Settling Time | < 5.0 s | ~4.2 s |
| Overshoot | < 20 % | ~12 % |
| Steady-State Error | ≈ 0 | < 0.01 m/s |
Overshoot display is colour-coded: green (< 12%), orange (12–15%), red (> 15%)
pid-vehicle-speed-controller/
├── speed1.mlapp ← Full MATLAB App Designer app (open directly in MATLAB)
├── README.md
└── media/
└── app_screenshot.png ← Screenshot of the running app dashboard
Requirements:
- MATLAB R2021a or later
- App Designer (included in standard MATLAB — no additional toolbox required)
- Control System Toolbox (optional — not used in this version, but recommended for future extension)
Step 1 — Clone or download
git clone https://github.com/YOUR-USERNAME/pid-vehicle-speed-controller.gitStep 2 — Open in MATLAB
- In MATLAB, navigate to the cloned folder
- Double-click
speed1.mlapp— it opens in App Designer - Click Run (▶) in the App Designer toolbar
Step 3 — Use the app
- The simulation runs automatically on startup with default Z-N tuned gains
- Adjust Kp, Ki, Kd sliders to observe how gains affect the step response
- Turn the Reference Speed knob to change the target speed (10–30 m/s)
- Watch the 3D animation and response plot update in real time
- Read Rise Time, Settling Time, Overshoot, SSE from the metrics panel
- Press STOP to pause the simulation at any time
Why Euler integration instead of Simulink?
The PID loop is implemented manually using Euler's forward method within a standard MATLAB for loop. This makes every step of the controller and plant equations explicit and readable — ideal for academic understanding and demonstration.
Time delay implementation:
The transport delay Td = 0.3 s is implemented using a circular shift buffer (u_buffer). At each time step, the delayed input is taken from the tail of the buffer. This avoids fractional-delay interpolation while keeping the implementation simple and accurate at the chosen dt.
Real-time animation:
The vehicle animation uses tic/toc wall-clock timing to map real elapsed time directly to simulation indices, so the animation plays at exactly 1× real time regardless of loop speed. The 3D sedan is drawn frame-by-frame using patch() objects (body, cabin, wheels, lights) — no Simulink 3D Animation toolbox required.
- Add Simulink model alongside the App for comparison
- Implement auto-tuning (relay feedback / Åström-Hägglund method)
- Add disturbance injection (speed bump, wind) mid-simulation
- Extend to 2-DOF controller (setpoint weighting)
- Export simulation data to CSV for offline analysis
- Add lane-change lateral dynamics (2D full vehicle model)
| Detail | Info |
|---|---|
| Course | MCE 532 — Control Systems Engineering |
| Institution | Federal University of Technology, Minna (FUT Minna) |
| Department | Mechatronics Engineering |
| Level | 500 Level (Final Year) |
| Year | 2024/2025 |
| Developer | Raphael Ebubechi Efita |
This project is licensed under the MIT License — see the LICENSE file for details.
Raphael Ebubechi Efita Mechatronics Engineering | Control Systems | MATLAB | Embedded Systems Federal University of Technology, Minna, Nigeria
