Multi-Featured Policy Gradient Supplementary and implementation for "Deep Reinforcement learning with Multiple Unrelated Rewards for AGV Mapless Navigation" accepted by T-ASE Code coming soon Supplementary material is available in Release.