How paulrobinson07/cte-capstone works
This is a browser-based neural-network car-driving simulation with timed generations. The main loop updates and renders cars, walls, and evolution state; cars sense track boundaries, feed distances through weighted neurons, and use outputs to steer. The user can steer manually or force a generation advance. The README is empty, so the architecture is based on the sampled JavaScript; the HTML entry wiring was not shown.
Main entry points: car.js Car Simulation · game.js Game Loop · output_neurons.js Control Neurons · wall.js Track Walls
Components 8
Simulation Loop
- Game Loop
game.js - Line Geometry
game.js
Driving World
- Car Simulation
car.js - Track Walls
wall.js
Neural Control
- Sensor Neurons
input_neurons.js - Control Neurons
output_neurons.js
Evolution
- Generation Handler
evolution_handler.js
Other components
- Player
Connections 12
- Player → Game Loop: steers / advances
game.js - Game Loop → Car Simulation: updates
game.js - Game Loop → Track Walls: updates
game.js - Game Loop → Generation Handler: updates
game.js - Car Simulation → Sensor Neurons: creates / updates
car.js - Car Simulation → Control Neurons: creates / updates
car.js - Sensor Neurons → Control Neurons: supplies weighted readings
output_neurons.js - Control Neurons → Car Simulation: provides control values
car.js - Car Simulation → Track Walls: checks ray collisions
car.js - Car Simulation → Line Geometry: calls intersection helpers
car.js - Generation Handler → Car Simulation: creates successors
evolution_handler.js - Game Loop → Track Walls: retargets focus
game.js