A Node.js implementation of a Genetic Algorithm (GA) for solving the Traveling Salesman Problem (TSP) applied to package delivery routes. Generates random delivery coordinates in a region, computes distances with the Haversine formula, and runs GA across multiple iterations to find the shortest route.

• 🗺️ Location Generation via GenerateLocation.js: Randomly generate delivery points within a bounding box (default Surabaya).
• 📐 Distance Matrix saved in location_matrix.json: Haversine distances (km) for all point pairs.
• 🧬 Genetic Algorithm in RoutePlanner.js:
  • Order-1 crossover
  • Random initial population (default size: 100)
  • Tournament selection (contest size: 5)
  • Swap mutation (rate: 0.02)
  • Fitness function = inverse total route distance
• 🔄 Multiple Runs: Repeat GA (default 10 runs) to identify the global best route.
• 📈 Output: Logs the best route sequence (point names) and its total distance in km.

1. Generate Data

   node GenerateLocation.js

   • Creates location_array.json and location_matrix.json

2. Run GA Planner

   node RoutePlanner.js

   • Prints:
     • Best Route Sequence: [PointX, PointY, ...]
     • Lowest Total Distance (km): <value>

• Point Count: Modify deliveryPointCount in GenerateLocation.js
• GA Parameters: Adjust runCount, populationSize, mutationRate, and generationCount in RoutePlanner.js

MIT License © 2025 NeaByteLab