Multi-Source Communication Delays (MSCD) seriously impair the operation of Distributed Control Systems (DCS), which are essential for real-time process monitoring in industrial settings. These difficulties were revealed by a testbed analysis conducted at the Tiger Food Paste production facility. A Multi-Aware Particle Swarm Optimisation (M-PSO) technique has been developed to identify communication delays by optimising significant features such as routing speed and bandwidth in order to solve MSCD. The approach guarantees accurate delay detection in dynamic network conditions through enhancements of gradient descent and dynamic learning rate updates. Additionally, a Network Management Algorithm (NMA) has been adopted, which uses priority scheduling to reduce latency and enhance bandwidth usage. Significant improvements in delay detection and management were shown when the suggested solutions were evaluated against common QoS criteria, improving the DCS network's dependability and efficiency. With applicability in a variety of real-time control systems, this study offers a strong framework for resolving communication delays in industrial automation. Advanced machine learning models may be used in future research to increase flexibility to changing network circumstances.