README.md

best.json

It contains the training data used in our research

• state
  1. first variable: angle(car between center)，1 number
  2. second variable: track, 19 numbers
  3. third variable: trackpos(shift from center), 1 number
  4. fourth variable: speed, [x, y, z], 3 numbers
  5. five variable: spin(four tyres), 4 numbers
  6. six variable: engine speed, 1 number
• action
  1. first varibale: steer angle, 1 number
  2. second variable: accelerate, 1 number
  3. third varible: brake, 1 number
• reward
  1. first variable: reward, 1 number