In fog computing, computing resources are deployed at the network edge, which can include routers, switches, gateways, and even end-user devices. Fog computing focuses on running computations and storing data directly on or near the fog devices themselves. The data processing occurs locally on the device, reducing the reliance on network connectivity and allowing for faster response times. However, the conventional intrusion detection system (IDS) failed to provide security during the data transfer between fog nodes to cloud, fog data centres. So, this work implemented the optimized IDS in fog computing environment (OIDS-FCE) using advanced naturally inspired optimization algorithms with extreme learning. Initially, the data preprocessing operation maintains the uniform characteristics in the dataset by normalizing the columns. Then, comprehensive learning particle swarm based effective seeker optimization (CLPS-ESO) algorithm extracts the intrusion specific features by analyzing the internal patterns of all rows, columns. In addition, automatic termination-based whale optimization algorithm (ATWOA) selects the best intrusion features from CLPS-ESO resultant features using correlation analysis. Finally, the hybrid extreme learning machine (HELM) classifies the varies instruction types from ATWOA optimal features. The simulation results show that the proposed OIDS-FCE achieved 98.52% accuracy, 96.38% precision, 95.50% of recall, and 95.90% of F1-score using UNSW-NB dataset, which are higher than other artificial intelligence IDS models.