Robot Obstacle Avoidance Controller Based on Deep Reinforcement Learning

Table 1

Symbols and rewards terms.

Symbols
	Running cycle
	Set of actions at th cycle
	Position set of agent at th cycle
	Agent’s central coordinates at th cycle
	Reward
	Position set of the obstacle
	Position set of the track
	Position set of the end
	Constant
	Empty set
Rewards terms
Distance reward
Obstacle collision penalty
Off the track penalty
Time penalty
Obstacle avoidance reward
Speed reward