Smart Irrigation Fuzzy Logic Controller

A self-contained Jupyter notebook implementing a Mamdani-type fuzzy logic controller to compute watering duration based on soil moisture and ambient temperature. Built with scikit-fuzzy and matplotlib, it includes visualizations, a rule base, partition testing, and rule conflict checks.

Repository structure

• Fuzzy_Assignment_225522M.ipynb — Main notebook with the full implementation and documentation.

Features

• Linguistic variables and membership functions:
  • Inputs: Soil moisture (0–100) — dry/moist/humid; Temperature (0–40 °C) — low/medium/high
  • Output: Watering duration (0–30 min) — short/medium/long
• 9-rule fuzzy rule base
• Inference demo for sample inputs
• Visualizations of membership functions and output decision
• Validation utilities:
  • Partition testing to ensure coverage of fuzzy sets
  • Rule conflict testing for contradictory antecedents

Requirements

• Python 3.8+
• numpy
• scikit-fuzzy
• matplotlib

Install packages on Windows (cmd):

pip install -U scikit-fuzzy numpy matplotlib

Tip: The notebook’s first cell contains a commented pip install line that you can run inside the notebook if preferred.

Getting started

1. Open the notebook in VS Code (with the Jupyter extension) or Jupyter Lab/Notebook.
2. Select a Python environment with the required packages installed.
3. Run cells from top to bottom. Plots will display inline, and the computed watering duration will be printed.

How it works

• Defines universes of discourse for inputs and output and assigns triangular membership functions.
• Encodes interpretable rules (e.g., dry soil + high temperature → long watering).
• Uses scikit-fuzzy’s control system to perform fuzzification, inference, and defuzzification.
• Provides plots and simple checks to validate the fuzzy set coverage and rule consistency.

Customization

• Tune membership function parameters to match your sensor calibration and irrigation needs.
• Extend the rule base or add new inputs (e.g., humidity, sunlight) as needed.