Real-time Traffic Monitoring and Optimization in Power Plant System using Integrated Actor-Critic with Proximal Policy Optimization

In recent scenarios, real-time monitoring of Power Plant System (PPS) is critical for ensuring the reliable and efficient operations. However, the increasing volume of network traffic by the growth of big data and other technologies has led to decreased network performance. In addition, the existing methods majorly struggled with low energy utilization efficiency and inefficient monitoring delays. To overcome these issues, integrated Asynchronous Advantage Actor-Critic with Proximal Policy Optimization (A3C-PPO) is proposed for optimization of power plant systems. The clipping mechanism of PPO helps to prevent the large policy updates to reduce monitoring delays. In addition, the optimized energy allocation of the proposed A3C-PPO in real-time traffic learns an energy aware policy to balance energy utilization for improving overall efficiency. Initially, real-time traffic data is collected using multi sensors such as temperature, and pressure sensors. Then, this data transmits into the Base Band Unit (BBU) of Cloud-Radio Access Network (C-RAN) with 5G networks. After that, the resources are scheduled for power networks with A3C to enable real-time traffic monitoring. Finally, PPO is employed to optimize the resources based on clipping mechanisms by avoiding policy updates. From the results, the proposed A3C-PPO outperformed existing Dueling Deep Q-network (DQN) algorithm with decreased transmission delay of 20 % and increased energy utilization by 22% respectively