Efficient Cyberattack Detection Methods in Industrial Control Systems.

The article deals with the issue of detecting cyberattacks on control algorithms running in a real Programmable Logic Controller (PLC) and controlling a real laboratory control plant. The vulnerability of the widely used Proportional-Integral-Derivative (PID) controller is investigated. Four effective, easy-to-implement, and relatively robust methods for detecting attacks on the control signal, output variable, and parameters of the PID controller are researched. The first method verifies whether the value of the control signal sent to the control plant in the previous step is the actual value generated by the controller. The second method relies on detecting sudden, unusual changes in output variables, taking into account the inertial nature of dynamic plants. In the third method, a copy of the controller parameters is used to detect an attack on the controller's parameters implemented in the PLC. The fourth method uses the golden run in attack detection.

Fast Nonlinear Predictive Control Using Classical and Parallel Wiener Models: A Comparison for a Neutralization Reactor Process.

The Wiener model, composed of a linear dynamical block and a nonlinear static one connected in series, is frequently used for prediction in Model Predictive Control (MPC) algorithms. The parallel structure is an extension of the classical Wiener model; it is expected to offer better modeling accuracy and increase the MPC control quality. This work discusses the benefits of using the parallel Wiener model in MPC. It has three objectives. Firstly, it describes a fast MPC algorithm in which parallel Wiener models are used for online prediction. In the presented approach, sophisticated trajectory linearization is performed online, which leads to computationally fast quadratic optimization. The second objective of this work is to study the influence of the model structure on modeling accuracy. The well-known neutralization benchmark process is considered. It is shown that the parallel Wiener models in the open-loop mode generate significantly fewer errors than the classical structure. This work's third objective is to validate the efficiency of parallel Wiener models in closed-loop MPC. For the neutralization process, it is demonstrated that parallel models demonstrate better control quality using various indicators, but the difference between the classical and parallel models is not significant.

Forgery Cyber-Attack Supported by LSTM Neural Network: An Experimental Case Study.

This work is concerned with the vulnerability of a network industrial control system to cyber-attacks, which is a critical issue nowadays. This is because an attack on a controlled process can damage or destroy it. These attacks use long short-term memory (LSTM) neural networks, which model dynamical processes. This means that the attacker may not know the physical nature of the process; an LSTM network is sufficient to mislead the process operator. Our experimental studies were conducted in an industrial control network containing a magnetic levitation process. The model training, evaluation, and structure selection are described. The chosen LSTM network very well mimicked the considered process. Finally, based on the obtained results, we formulated possible protection methods against the considered types of cyber-attack.

Beyond the quadratic norm: Computationally efficient constrained nonlinear MPC using a custom cost function.

A new approach to nonlinear Model Predictive Control (MPC) is discussed in this work. A custom user-defined cost function is used in place of the typically considered quadratic norm. An approximator of the cost function is applied to obtain a computationally simple procedure and linearization of two trajectories is carried out online. The predicted output trajectory of the approximator and the predicted trajectory of the manipulated variable, both over the prediction horizon, are repeatedly linearized online. It yields a simple quadratic programming task. The algorithm is implemented for a simulated neutralization benchmark modeled by a neural Wiener model. The resulting control quality is excellent, identical to that observed in the MPC scheme with nonlinear optimization. Validity of the described MPC algorithms is demonstrated when only simple box constraints are considered on the process input variable and in a more demanding case when additional soft limitations are put on the predicted output. Two structures of the approximator are compared: polynomial and neural; the advantages of the latter one are shown and stressed.

Digital Twins in the Practice of High-Energy Physics Experiments: A Gas System for the Multipurpose Detector.

The digital twins technology delivers a new degree of freedom into system implementation and maintenance practice. Using this approach, a technological system can be efficiently modeled and simulated. Furthermore, such a twin offline system can be efficiently used to investigate real system issues and improvement opportunities, e.g., improvement of the existing control system or development of a new one. This work describes the development of a control system using the digital twins methodology for a gas system delivering a specific mixture of gases to the time-of-flight (ToF) multipurpose detector (MPD) used during high-energy physics experiments in the Joint Institute for Nuclear Research (Dubna, Russia). The gas system digital twin was built using a test stand and further extended into target full-scale installation planned to be built in the near future. Therefore, conducted simulations are used to validate the existing system and to allow validation of the planned new system. Moreover, the gas system digital twin enables testing of new control opportunities, improving the operation of the target gas system.

Computationally Efficient Nonlinear Model Predictive Control Using the L1 Cost-Function.

Model Predictive Control (MPC) algorithms typically use the classical L2 cost function, which minimises squared differences of predicted control errors. Such an approach has good numerical properties, but the L1 norm that measures absolute values of the control errors gives better control quality. If a nonlinear model is used for prediction, the L1 norm leads to a difficult, nonlinear, possibly non-differentiable cost function. A computationally efficient alternative is discussed in this work. The solution used consists of two concepts: (a) a neural approximator is used in place of the non-differentiable absolute value function; (b) an advanced trajectory linearisation is performed on-line. As a result, an easy-to-solve quadratic optimisation task is obtained in place of the nonlinear one. Advantages of the presented solution are discussed for a simulated neutralisation benchmark. It is shown that the obtained trajectories are very similar, practically the same, as those possible in the reference scheme with nonlinear optimisation. Furthermore, the L1 norm even gives better performance than the classical L2 one in terms of the classical control performance indicator that measures squared control errors.