Formation Control using Simplified Reinforcement Learning for Multi-agent systems with State Delay