Recently, the process control network (PCN) of oil and gas installation has been subjected to amorphous cyber-attacks. Examples include the denial-of-service (DoS), distributed denial-of-service (DDoS), and man-in-the-middle (MitM) attacks, and this may have largely been caused by the integration of open network to operation technology (OT) as a result of low-cost network expansion. The connection of OT to the internet for firmware updates, third-party support, or the intervention of vendors has exposed the industry to attacks. The inability to detect these unpredictable cyber-attacks exposes the PCN, and a successful attack can lead to devastating effects. This paper reviews the different forms of cyber-attacks in PCN of oil and gas installations while proposing the use of machine learning algorithms to monitor data exchanges between the sensors, controllers, processes, and the final control elements on the network to detect anomalies in such data exchanges. Python 3.0 Libraries, Deep-Learning Toolkit, MATLAB, and Allen Bradley RSLogic 5000 PLC Emulator software were used in simulating the process control. The outcomes of the experiments show the reliability and functionality of the different machine learning algorithms in detecting these anomalies with significant precise attack detections identified using tree algorithms (bagged or coarse ) for man-in-the-middle (MitM) attacks while taking note of accuracy-computation complexity trade-offs.

In recent times, the process control network (PCN) of oil and gas installation has been subjected to amorphous cyber-attacks which include Denial-of-Service (DoS), Distributed-Denial-of-Service (DDoS), Man-in-the-Middle (MitM) attacks, and this may have been caused majorly by the integration of open network to Operation Technology (OT) as a result of low-cost network expansion. The connection of the OT to the internet for firmware updates, third-party support, or vendor interventions, has exposed the industry to attacks. The inability to detect these unpredictable cyber-attacks exposes the PCN and a successful attack can lead to devastating effects. This paper reviews the different forms of cyber-attacks in PCN of oil and gas installations and proposes the use of machine learning algorithms to monitor data exchanges between the sensors, controllers, processes, and the final control elements on the network so as to detect anomalies in such data exchanges. Python 3.0 Libraries, Deep-Learning Toolkit, MATLAB, and Allen Bradley RSLogic 5000 PLC Emulator software were used in the simulation of process control. The outcome of the experiments shows the reliability and functionality of the different machine-learning algorithms in detecting these anomalies with significant precise attack detections identified using a coarse tree algorithm.

This paper explores the use of deep learning to optimize the performance of a peer-to-peer energy system with an intelligent master controller. The goal addresses inefficiencies caused by energy seasonality by predicting hourly power consumption through a deep learning algorithm. The intelligent master controller was designed to manage the collaborative energy system, and the deep learning technique was employed as an optimization scheme to forecast power system performance for more efficient utilization. The deep learning algorithm was trained using dataset from American electric power, where consumer load data serves as input, and forecasted power serves as output. The forecasted power was then used as input to the intelligent master controller, which determines suitable power supply for generation and storage based on the predicted demand. The experiment results show promising accuracy with a root mean square error (RMSE) of 0.1819 for hourly energy consumption averaged over a year, 0.2419 for hourly energy consumption averaged over a month, 0.0662 for hourly energy consumption averaged per day, and 0.0217 for hourly energy consumption. These findings demonstrate that the system is well-trained and capable of accurately predicting the energy required by the intelligent master controller, thus enhancing the overall performance of the peer-to-peer energy system.

A new Model Reference Adaptive Control (MRAC) based on minimizing Integral Time Absolute Error (ITAE) model criterion has been designed. However, an obvious challenge to the performance of the ITAE-MRAC method is poor rise time and settling time in positioning control of dish antenna. Thus, Fuzzy Logic Controller (FLC) and Fuzzy-Proportional Derivative plus Integral (Fuzzy-PD+I) controller were designed. The primary purpose of this work is to use intelligent based multi-loop control strategy to achieve efficient positioning control for satellite antenna. Two control strategies were proposed by separately integrating ITAE-MRAC model with FLC and Fuzzy-PD+I controllers. This resulted in multi-loop control systems in which the ITAE-MRAC model forms the outer loop whose output based on the adjustment mechanism was used to adjust the control dynamic of the inner (or secondary) loop controller (FLC or Fuzzy-PD+I). The use of multi-loop makes the inner loop more stable and efficient in providing improved tracking response to setpoint signal. The control models based on PID, ITAE-MRAC, FLC, Fuzzy-PD+I, ITAE-MRAC-FLC, and ITAE-MRAC-Fuzzy-PD+I were separately applied to the dish antenna positioning system. The results obtained from the controllers indicated that the rise times and settling times were improved including good overshoots respectively with zero steady-state error in all cases. The percentage improvement over FLC and Fuzzy-PD+I revealed 30% and 4.95% reduction in rise times and 22% and 1.62% reduction in settling times using ITAE-MRAC-FLC and ITAE-MRAC-Fuzzy-PD+I, respectively. Generally, the results indicated that the use of the ITAE based MRAC model enhanced the performances of the FLC and Fuzz-PD+I, and the overall transient and steady-state response dynamic characteristics of the dish antenna positioning system.