Trends in Advanced Intelligent Control, Optimization and Automation

An appearance of utter in oscillating structures is an endemic phenomenon. Most common causes are vibrations induced by the moving flow of a gas which is interacting with the structure. Typical examples include: turbulent jets, vibrating bridges, oscillating facial palate in the onset of apnea.

In this paper, we study the problem of state and control signal constraints in discrete-time sliding mode control. We introduce a sufficient condition for finite time convergence of the representative point to the sliding hyperplane, while respecting imposed restrictions. We propose a new control strategy based on the reaching law approach.

Unstable plant considered in the paper consists of an integrator and 1st order component with positive pole. PID controller is used, so the feedback system becomes of 3rd order. Root-locus design method is applied which for such system gives analytic expressions for controller settings. If weak control action is required, the design provides better responses than conventional one. Unstable ships, particularly oil tankers, require weak control.

In the paper two different control systems for nonlinear multi-input multi-output MIMO plants are compared and discussed. Their main objective is to reduce the influence of the plant inputs and outputs. A multi-controller control structure, which contains a set of a dynamic decoupling controllers is compared with a synthetized online nonlinear model predictive controller. Pros and cons of both methods are discussed and presented in series of simulations of control of a selected nonlinear MIMO plant. Te paper ends with some final remarks on a practical implementation of decoupling methods for a nonlinear MIMO plants.

Offset-free model predictive control (MPC) algorithms for nonlinear state-space process models, with modeling errors and under asymptotically constant external disturbances, is the subject of the paper. A brief formulation of the MPC formulation used is first given, followed by a brief remainder of the case with measured state vector. The case with process outputs measured only and thus the necessity of state estimation is further considered.The main result of the paper is the presentation of a novel technique with process state estimation only, despite the presence of deterministic disturbances. The core of the technique is the state disturbance model used for the state prediction. It was introduced originally for linear state-space models and is generalized to the nonlinear case in the paper. This leads to a simpler design without the need for decisions of disturbance structure and placement in the model and to simpler (lower dimensional) control structure with process state observer only. A theoretical analysis of the proposed algorithm is provided, under applicability conditions which are weaker than in the conventional approach of extended process-and-disturbance state estimation. The presented theory is illustrated by simulation results of a nonlinear process.

The paper presents an approach to dumping of the pendulum during dynamic stabilization in an arbitrary angle. The rapid oscillations of the pendulum’s suspension point cause that created effective potential has a local minimum which guarantees the stability of the pendulum. The external disturbances bring an additional energy into the system and the pendulum increases the amplitude of oscillations around its equilibrium position. The aim of this paper is to describe the damping method of this excess oscillations during dynamic stabilization of the pendulum.

Example of dual control of linear uncertain system have been presented. The control task with short horizon (N=2) were solved using dynamic programming. It was shown that the optimal solution is ambiguous, the cost function is non-convex and has many local minima. Optimal control depends in a discontinuous manner on the initial conditions. It was also observed that active learning occurs only when the uncertainty of the initial state exceeds a certain threshold. In this case, the amount of information transmitted from sensor to the controller is much greater than in the case of passive learning.

The subject of this paper is a comparison of two control strategies of an inverted pendulum on a cart. The first one is a linear-quadratic regulator (LQR), while the second is a state space model predictive controller (SSMPC). The study was performed on the simulation model of an inverted pendulum, determined on the basis of the actual physical parameters collected from the laboratory stand AMIRA LIP100. It has been shown that the LQR algorithm works better for fixed-value control and disturbance rejection, while the SSMPC controller is more suitable for the trajectory tracking task. Furthermore, the system with SSMPC controller has smoother changes in the control signal, that can be beneficial for an actuator, while LQR controller may generate adverse, rapid changes in the control signal.

From the control viewpoint, 3D crane is a dynamic, nonlinear and multidimensional electromechanical system. In this paper, five control systems using a set–point weighted PID controllers (modified controllers) are designed. These structures and properties are presented. Optimization process of controllers settings based on integral performance indices is made. Simulation tests of control systems are presented. Comparison of integral indices for control systems using classical PID controllers and modified PID controllers on physical model is presented.

In the paper a Lyapunov matrices approach to the parametric optimization problem of time delay system with a PID-controller is presented. The value of integral quadratic performance index is equal to the value of Lyapunov functional for the initial function of the time delay system. The Lyapunov functional is determined by means of the Lyapunov matrix. The example of the inertial system with time delay and with a PID-controller is presented.

The first order sensitivity analysis is performed for a class of optimal control problems for time lag parabolic equations in which retarded arguments appear in the integral form with h ∈ (0, b) in the state equations and with k ∈ (0, c) in the Neumann boundary conditions. The optimality system is analyzed with the respect to a small parameter. The directional derivative of the optimal control is obtained as a solution to an auxiliary optimization problem. The control constraints for the auxilary optimization problem are received.

In this paper a method of solving optimal control problems of systems described by Fredholm type integral equations was described. Special attention was given to difficult problems with point constraints imposed on the state of the system.

Observers for descriptor electrical circuits by the use of the shuffle algorithm are proposed. Necessary and sufficient conditions for the existence of the observers are given. The effectiveness of the proposed method is demonstrated on a numerical example.

The paper discusses the subject of estimation of potential financial benefits achievable with the rehabilitation (modification or tuning) of the control system. This issue appears almost in any process improvement initiative giving arguments for control upgrades. The subject exists in literature for several years with well established the same limit approach. The procedure is based on the assumption of Gaussian properties of the considered variable reflected in its histogram. Review of industrial data shows frequent situations when process variables are of different character featuring long tails. Such properties are well described by α–stable distributions. This paper presents extension of the method on such general probability density functions family. The analysis is illustrated with the simulation and industrial data examples.

Time-varying communication delays, data packet dropping, and other network-induced phenomena are inherent in networked control systems. Their presence can deteriorate noticeably control quality and narrow stability margins forcing designers to adopt conservative controller tuning. Control quality drop can be to some extent remedied by employment of network packet buffering or queuing, finite horizon state estimation, and continual time delay identification methods applied to networked control loops. The paper presents a modular structure for a networked control system where the named techniques are deployed in a network node located on the actuator site. A case study for a DC motor servomechanism and a periodic trajectory tracking problem is given. Simulation results are provided.

The paper focuses on essential elements of the control systems that are currently used in semi-automated cars and those that will be used in fully automated cars named as autonomous cars. Autonomous car is a vehicle that is capable of sensing its environment and navigating without human input. The highest level of automated driving assumes that all activities done by the driver can be replaced by a suitable control system. In this context, the autonomous car can be regarded as a control system working in a closed-loop setting. The desired value for the defined control system is a vehicle state specified in the destination point where the car should be placed starting a ride from a given initial state. The car in automated driving mode affects other cars in traffic, elements of road infrastructure and other road users, such as pedestrians, cyclists, animals, etc. Set of sensors provides data on the environment surrounding the car in the area defined by their field of view. These data after initial pre-processing are used for detecting objects surrounding the vehicle. Data from different types of sensors are combined with each other in order to increase the confidence level of the objects detected in the vehicles neighbourhood. Having the list of detected objects a vehicle environmental model is created, which is used to analyse the current situation and then plan the trajectory of the vehicle to the destination state. Besides the vehicle environmental model an important role in fully automated cars plays the driver model by means of which can be determined the characteristics for the actuators in such a way that the dynamics of the vehicle is as close as possible to the situation in which the driver manually guides the vehicle. Models of vehicle kinematics and dynamics are essential to tune controllers that are responsible for determining the trajectory of the vehicle and its stability properties. It should be emphasised that due to the nature of the automotive industry a fully automated driving will not happen overnight, but as technology develops. Control systems already on the market are gradually developed and their functionalities will work towards taking overall control of the vehicle.

The paper analysis the stability property of a series of cars following each other and functioning in adaptive cruise control mode. The adaptive cruise control mode controls the acceleration and deceleration of the car in order to maintain a set speed or to avoid a crash. Such series of cars can be mathematically represented by an equivalent system consisting of a set of masses, springs and dampers. Using this representation, the dynamic behaviour of the cars in a chain can be described by a matrix-vector differential equation of second-order. The paper analyses this model and formulates stability conditions.

Control systems in autonomous vehicles can be considered as distributed embedded software systems where independent microprocessor systems communicate together using different communication protocols. Typical autonomous driving functionality is then realised by several microprocessors communicating with each other. Quality assurance and safety standards combined with increasing complexity and reliability demands make the development of such systems challenging. In order to assure the required quality and compliance with safety standards, a formal and methodical approach for testing and verification is required. The paper presents a proposal of such approach for verification and testing of control systems in the automotive applications covering active safety, advanced driver assistance and autonomous driving systems. The main focus of this approach is black-box testing and includes test design, implementation and execution.

A laboratory Anti-Lock Brake System (ABS) is examined. The architecture of the ABS system is shown. The real-time experiments related to a control action to stabilize the slip at a certain level are taken into consideration. A new slip control algorithm differs from the previously used approaches. It is based on the measured current of the braking DC motor. The experimental results collected in the real-time for an old (relay) and new (current based) slip control are compared.

Several problems may arise when multiple trains are to be tracked using two IP camera streams. In this work, real-life conditions are simulated using a railway track model based on the Pomeranian Metropolitan Railway (PKM). Application of automatic clustering of optical flow is investigated. A complete tracking solution is developed using background subtraction, blob analysis, Kalman filtering, and a Hungarian algorithm. In total, six morphological, convolutional and non-linear filtering methods are compared in sixty-three combinations. Accuracy and performance of the system are evaluated, and the obtained results are analysed and commented on.

The problem of improving driving safety for vehicles equipped with magnetorheological (MR) dampers is considered. It is proposed to control the MR dampers using the clipped LQ (Linear Quadratic) control which can be regarded as the two-dimensional Skyhook algorithm. The strategy is applied to a half-car model with four degrees of freedom, oriented on roll dynamics. Here, control algorithm optimised with respect to minimisation of roll vibrations is considered. Simulation experiments were performed assuming the model is subjected to impulse excitation generated as a torque applied in the centre of gravity, that is equivalent to a manoeuvre in the form of a sudden turning. The quality of the algorithm was validated using the RMS-based performance index and the results confirm the effectiveness of the proposed solution for the semi-active suspension.

The paper presents review of key aspects associated with object detection and tracking algorithms in sensing systems for automotive applications covering active safety, advanced driver assistance and autonomous driving systems. The role of discussed systems is to establish location and monitor relative movement of objects and environment with main focus on identification of vulnerable road users. The object detection algorithms allow to classify and differentiate both static and dynamic objects in the vehicle’s surrounding. The analysis will be focused on selection of most suitable platform for development of optimized and safe vulnerable road user detection system.

There are typically two main control loops with PI controllers operating at each turbo-generator set. In this paper a model predictive controller QDMC for the steam turbine is proposed - instead of a typical PI controller. The QDMC controller utilize a step-response model for the controlled system. This model parameters are determined, based on the simplified and linear model of turbo-generator set, which parameters are identified on-line with RLS algorithm. It has been found that the proposed QDMC controller realize the reference trajectories of the effective power and the angular velocity, and damp the electromechanical oscillations with satisfactory quality in comparison to the typical PI and DMC controllers.

Due to the presence of right-half plane zeroes in the transfer function of a nonminimum-phase plant, conventionally designed feedforward controller becomes unstable. Main focus of this article is put on a comparison of the Extended Bandwidth Zero Phase Error and recently proposed fixed-structure approximate inverse feedforward control design methods applicable to the nonminimum-phase systems. Comparison of the techniques presented in the paper is based on the frequency analysis of a closed-loop error transfer function, followed by simulation examples and experimental validation on a laboratory setup with a double-drum coiling machine. Both simulation and experimental results showed the superiority of trajectory tracking obtained by the fixed-structure approximate inverse method under the assumed conditions.

Our aim is to discuss briefly methods of using cameras in control systems. Then, we concentrate on a new approach to iterative learning control (ILC) for nonlinear repetitive production processes. Finally, we propose the methodology of applying a camera for tuning ILC and illustrate it by the example of a multilayer system for laser power control in selective laser melting (SLM).

This paper describes an approach for tracking objects in 3D scene using Stereo Vision System with the control of cameras gaze and vergence. The goal of camera positioning mechanism is to keep cameras fixed on a common visual target. Mechanical linkage between cameras ensures separate control of vergence and gaze angles and the same distance from the left and the right camera to the fixated world point. Keeping the tracking target fixated causes that object lies near horopter- a surface with zero disparity. It means that stereo images of this object have a narrow range of disparities and makes it possible to use stereo algorithm that accepts only a limited range of disparities. On the other hand, most of the stereo matching algorithms need fully rectified input images or at least accurate camera calibration parameters. Side effect of such active control of cameras positioning is a continuous degradation of calibration and needs re-calibration or estimation of stereo parameters for each new position.

The paper addresses the issue of the state estimation problem of a class of discrete-event systems. The receding-horizon approach is employed to solve above problem. The system and its variables are described within the (max,+) algebra. Thus making possible to incorporate robustness within the overall framework. The paper also shows the transformation of the interval cost function into the scalar one, and hence, making the computational procedure trackable within the quadratic programming framework. Resulting in interval estimates of the system state, which can be used for both fault diagnosis and control purposes.

It is possible to improve sound insulation of devices and machinery from the environment by appropriately controlling vibrations of their casings. When implementing this approach, there is a need to select efficient locations of actuators on each of the casing walls. A common solution is to employ an optimization algorithm to find favourable arrangement according to a given objective. One of the essential parameters of this process is a number of actuators to be optimized, but most often it is assumed a priori, without any proper considerations.The aim of this paper is to give an insight into the problem of defining the number of actuators for active noise reduction implementations. The optimal value depends on the particular application, its mechanical limitations, costs, considered frequency band to be controlled, etc.However, a general approach for analysis and decision making can be formulated. As an exemplary structure, a light-weight device casing is considered. The relationships between the number of actuators, the considered frequency band, and obtained values of the optimization index are given.

Programming a control example by CPDev tool using LD, FBD, and SFC graphical languages of IEC 61131-3 standard is presented. Another example demonstrates the use of data structures and arrays of function blocks in ST textual language. The two parts indicate that CPDev provides now essential functionalities specified in the standard. Remarks how the tool relates to established engineering environments such as CoDeSys, STEP 7, Control Builder F, and ISaGRAF are also given.

This paper describes a system for code auto-generation of Model Predictive Control algorithms for Multiple-Input, Multiple-Output processes. Transcompiler – the main part of the system – generates C code of the algorithm, basing on MATLAB code, which contains definition of both algorithms and its parameters. The resulting code is optimised for microcontrollers in terms of memory and computational power necessary for on-line calculation of the optimal values of manipulated variables. This approach may decrease the development time of prototype controllers and lower their cost. Tests are conducted using STM32 microcontroller for simulated processes and results are demonstrated and described.

This paper describes implementation of the Dynamic Matrix Control (DMC) algorithm performed on an Altera Field Programmable Gate Array (FPGA) with the Cyclone IV chip. The DMC algorithm is implemented in its analytical (explicit) version which requires computationally simple matrix and vector operations in real time, no on-line optimisation is necessary. The test-bench application is prepared for fast comparison between C and HDL versions of code. A large number of independent logic cells can provide multi-parallel operations to achieve very fast operations. As a result, the algorithm may be used for controlling very fast dynamic processes characterised by sampling periods of millisecond order. Preliminary results of real experiments are demonstrated. The discussed control structure provides possibility to fast change of algorithm.

This paper describes the process of implementation of DMC algorithm in numerical version in embedded environment. The study focuses on the dependences between algorithm parameters and memory and performance requirements. As a result of the investigation the algorithm modifications are presented to achieve compromise between the fast calculations and the memory requirements.

The real-time simulator of nuclear reactor basic processes (neutron kinetics, heat generation and its exchange, poisoning and burning up fuel) build in a network environment is presented in this paper. The client-server architecture was introduced, where the server is a powerful computing unit and the web browser application is a client for user interface purposes. The challenge was to develop an application running under the regime of real-time, with a high temporal resolution, in an environment which is not a native real-time. The problem of a real-time operation taking into account the variable time of calculations and a communication latency was solved using the developed mechanism of step length adaptation. Results of multiple studies of a numerical compliance with the reference simulator proved correctness of the developed application.

In this paper a hardware-software implementation of an object tracking system, which uses a moving camera is described. Selected algorithms: mean-shift, particle filter, KLT and so-called tracking by detection were analysed and evaluated. Particular attention was paid to the effectiveness of fast moving object tracking and the ability to implement the algorithm in a heterogeneous computing system. The selected solution was implemented in the Zynq SoC (System on Chip) device from Xilinx. Object position was used to control two servomotors, which constituted a pan-tilt mounting of the camera. Additionally, object position prediction was realised using a Kalman filter. The proposed system is able to process a 1280 × 720 @ 60 fps video stream in real time and track moving objects.

In the article we present an implementation of the system for automatic calculation of relevant descriptors of the glass melting process from the image of the glass surface. The purpose of the system is an automatic analysis of the process course in real time and providing information about batch amount if different zones of the furnace, the symmetry of batch distribution and the symmetry of temperature distribution. The system has been developed in Matlab and implemented with consideration of industrial requirements.

In the work there is considered an NP-hard flexible job shop problem. Its solution lies in allocation of operations to machines and determination of the sequence of their execution. There is also a method of construction of approximate algorithms presented, based on the idea of descent search, determining the allocation of operations. What is more, there were computational experiments conducted to investigate the correlation between the size of the neighborhood and the quality of solutions determined by the algorithm.

This paper describes implementation of the Generalized Predictive Control (GPC) algorithm on an STM32 microcontroller with the ARM Cortex M7 core. The algorithm is implemented in its analytical (explicit) version which requires computationally simple matrix and vector operations in real time, no on-line optimisation is necessary. As a result, the algorithm may be used for controlling very fast dynamic processes characterised by sampling periods of millisecond order. Results of real experiments are demonstrated for two example processes.

A new modeling technique leading to a simple and reliable dynamic model of multi-source friction is used together with adaptive control algorithm. The model is applied in the electric linear drive system. We give detailed information about preparing the proposed model and the control algorithm. We also investigate effects of the fuzzy model’s complexity on the system performance. The presented experiments prove the usefulness of the proposed approach.

The relationships between the reachability of positive standard and fractional discrete-time and continuous-time linear systems are addressed. It is shown that: 1) The fractional positive discrete-time and linear systems are reachable in one step if and only if the corresponding positive standard system is reachable in one step; 2) If the positive standard discrete-time linear system with single input is unreachable, then the corresponding fractional positive system is also unreachable; 3) The fractional positive continuous-time linear system is reachable if and only if the corresponding continuous-time positive standard system is reachable.

Fractional order linear descriptor systems with finite memory are studied. The formula for trajectory of such system is given. The Lyapunov-Krasovskii approach is used to analyze the stability of the considered systems.

The paper is intended to present a possibility of modeling a heat transfer process in one dimensional plant with the use of new, non integer order, discrete models. The proposed models have the form of hybrid transfer functions, containing both integer and non integer order parts. The integer order part represents a pure delay, the non integer order part is modeled with the use of CFE approximant. The proposed models were compared to step responses of real experimental plant with the use of typical MSE cost function. Results of experiments show that a good accuracy can be achieved for relatively low order of model.

In this paper the fractional model of the human arm is presented. Proposed approach is an attempt to consider the muscle damping properties in the simplest form possible. As the base for our simulations the equation of motion based on Lagrange formalism was used. In order to obtain fractional properties we propose using fractional-order derivatives instead of integer-ones. This simplification creates opportunity for an easy implementation and comparison with commonly used models. The core of the presented method is solving this nonlinear equation. The preliminary results was shown. The simplest model possible was analyzed. We considered only two DOF (degrees of freedom) planar model without joint limitations and properly distributed masses. This experiment is to show the damping properties of the fractional model.

Many systems appearing in practice exhibit a hybrid nature, that is, a coupling between continuous dynamics and discrete events. Special cases of such systems are those in which the linear subsystems are switched according to time or according to state and/or input. In the paper a method for approximation of some kinds of switched fractional linear systems is proposed and their stability is analyzed.

The relationship between the controllability of standard and fractional linear stationary discrete-time addressed. It is shown that the fractional linear discrete-time control system is controllable if and only if the corresponding linear standard discrete-time system is controllable.

In this work, a selected type of non-integer order system is examined. The conditions for asymptotic stability are formulated with use of modified Michailov theorem. A numerical example for n=3 was also presented.

The bounds on the mobile robot curvature of motion and path curvature continuity constraints usually result either from mechanical construction limitations or practical motion smoothness requirements. Most path planning primitives compatible with those constraints force planning algorithms to utilize costly numerical methods for computation of maximal path curvature or positional path constraints verification. In this paper a novel path primitive is proposed, which can be concatenated with the line and circle segments to form a path with bounded curvature such that its perfect realization by a unicycle robot guarantees continuous time-derivative of its curvature of motion. Satisfaction of prescribed curvature bounds and positional path constraints resulting from obstacles in the environment is formally guaranteed using explicit analytic formulas presented in the paper. It is shown that the proposed approach yields an arbitrarily precise $$ {{\mathbb{G}}^3} $$-continuous approximation of the Reeds-Shepp paths. Presented analysis is further utilized to formulate the global path planning problem in a continuous domain as a tractable optimization problem. Computational effectiveness of the proposed method has been additionally verified by quantitative comparison of constraint satisfaction checking speed with the $$ {\eta^3} $$-splines.

Multi-robot task allocation is a well known and widely researched decision-making problem that is difficult to solve in reasonable time even for small instances. Additional complexity is added by the fact that the parameters of the system may change over time, which happens either by external stimuli or by the task execution itself. One of the common causes behind these changes is the movement of executors or tasks. This paper tackles a problem of multi-task, multi-robot allocation in a such an environment. Formulated and solved is a specific decision-making problem. Performed is an experimental comparison of a dedicated solution algorithm with known methods for the more general Multidimensional Knapsack and Covering problem. Empirical evaluation illustrates that a dedicated approach is competitive and often necessary, as the general approach proves to be too slow.

Robot control systems structures based on agents are presented. Agents are classified into 8 categories, with the embodied agent being the most general one. Out of those agents diverse control systems are built. Single/multi-robot systems are considered, where robots can have single or multiple effectors. The presentation of the subject relies on already implemented systems.

We present a new specialized evolutionary algorithm for the global path planning for mobile robots.We assume that multiple criteria are involved, notably the energy consumption, travel time and movement cost on dangerous areas, as well as constraints like the location obstacles, a limited minimal robot’s turning angle, a maximal energy (related to the capacity of batteries) and maximal time of travel. The new algorithm involves some new problem specific evolutionary operations. The simulation results using the V-REP platform involve obstacles, different surfaces and dangerous areas. The results are very encouraging, much better than those obtained using the traditional meta-heuristics, notably the widely used genetic algorithm. Moreover, a novel explicit involvement of energy consumption as a key factor, provides much new insight from both a theoretical and even more practical points of view.

In this paper trajectory for approaching during emergency STS transfer operation with oil spill is considered as a sequence of navigation manoeuvres in specific navigational environment. The designed way points - ship positions and speed are determined as reference values to support navigator in decision making during steering and to mitigate the risk of collision which mostly results from exceeding the speed limit of approaching. To prevent this, the values of position and speed in each way points are determined with respect to specific manoeuvring procedure during STS approaching and by taking into account constraints resulting from ship’s manoeuvre of stopping and speed deceleration performance included in manoeuvring booklet. Additional constraints result from STS transfer operation guide and navigation practise. The task of trajectory planning is defined as optimization process to minimize trajectory length, course alteration and maximize safety.

In the paper a three-stage, semirecursive scheme for the Wiener-Hammerstein system identification is proposed. The algorithm combines both parametric and nonparametric strategies and allows to recover linear and nonlinear subsystems directly from the noisy inputoutput data. As to the nonlinearity, the main idea is based on the recursive kernel censoring of measurements, while linear dynamics are recovered by a special kind of deconvolution. Efficiency of the obtained estimates is justified by numerical example.

The paper presents a new method of active identification by the use of nth order Strejc model with time delay. The model approximates the dynamics of multi-inertial system based on its step response. The basic and inconvenient version of identification method for this model was published by Strejc in 1959. Different results of research on this model were published in other works. In almost all these methods a key role is played by the graphic procedures for setting coordinates of the inflection point in step response of the system, drawing a tangent line in this point and finally determining the specific intervals given by this line. Based on the Strejc table one can then determine the hypothetical order and the time constant of the model. The reasons for the model approximation errors in this method are the inaccuracy of location of the inflection point, the procedure of tangent line drawing and the main idea that the step response of a real system and the model response are equal only in the one point. The method presented in this work, is based on the condition that the normalized step responses of the system and of the nth order Strejc model must be equal in two chosen points spaced from each other that will guarantee a good matching of the whole step characteristics and hence a good approximation model. As it turned out, the simple analytical formulas can be received which enable fast and simple determination of Strejc model parameters (for any specifying order n≥1). The most important features of this method are lack of procedure for the location of the inflection point and the tangent line drawing procedure, as well as the non-assumption in advance of the order of the model (in contrast to the Strejc method). The presented method is suitable for easy implementation in PLC controllers with self-tuning.

The paper presents the method of qualitative modeling of industrial process in the form of cause-and-effect graph that directly takes into account fault influence on process variables. Selected applications of that graph are briefly characterized. Its usefulness in alarm analyzes, designing of process diagnostics and HAZOP safety analyses is described.

The paper presents the straw combustion process in the biomass boiler in terms of its dynamics. Straw combustion takes place in a specially adapted chamber of the boiler. The analyzed biomass boiler has also a water jacket, which stores heat received from the burning straw. The thermal energy stored in the water jacket is used to heat the building.Author shows the original model of the straw combustion process which is based on the assumption that the loaded straw has the shape of a cube, and it still maintains this shape during combustion. In this model, it was also assumed that the combustion speed is proportional to the current burning straw surface, air mass flow and burning speed factor, which is characterized for a particular type of straw. The paper presents a mathematical model of the batch-fired straw boiler. That model was simulated in MATLAB/Simulink. The model time data was matched to the real time data of combustion thus making identification of some system parameters.

The paper analyses the problem of experimental identification (frequency response), modelling and optimization of the towing tank wavemaker in the Scilab/Xcos environment.The experimental identification of the objects (the towing tank wave-maker placed in the hydrodynamic laboratory of the CTO S.A. Ship Design and Research Centre (CTO)) and the implementation of the models in the simulation environment, enable to perform:1.tuning of the cascade PID controller (Astrom-Hagglund relay method),2.checking the stability (Routh-Hurvitz criterion, Nyquist criterion) of the wavemaker with tuned controller,3.evaluating the regulatory quality and simulating of work of the optimized (Ziegler-Nichols method) system of the wavemaker.The above works and the achieved results are further described and presented in the content of this paper.

We focus our attention on two stable models of nonlinear dynamic systems (external dynamic approach): NFIR (nonlinear finite impulse response) and simple version of NARX (nonlinear autoregressive model with external inputs), and their linear counterparts. The main idea investigated in the paper is to project the vector of past inputs u _n ’s onto random directions drawn uniformly from the unit sphere, (instead of estimated) and select only those projections that are relevant for proper neural networks based models.

In this paper the process model of steam superheating system, localized inside a power boiler; and the advanced, cascade control system which is used to maintain a steam temperature are proposed. Model and control loop have been implemented and investigated in Distributed Control System and in Matlab as well.

This paper describes model identification of an active magnetic levitation laboratory process. Magnetic levitation is a nonlinear, open-loop unstable and time-varying dynamical system whose modelling and control are difficult. Theoretical model takes into account set of parameters which are difficult to identify. The proposed approach assumes discrete-time form of the model the parameters of which are tuned using nonlinear optimisation methods. The obtained model is more accurate and can be directly used in model-based control algorithms. Results of real experiment are demonstrated for the INTECO MLS2EM laboratory set-up.

The paper gives an overview of the infrastructure of the electric power control system. It describes the basic threats caused by unauthorized actions and provides examples of the energy management system infrastructure vulnerabilities and their possible consequences. An architecture of a system that offers an effective protection of the Industrial Control Systems from cyber threats is described in some details. The test scenarios used for the system verification and the results of experiments performed both in Supervisory Control and Data Acquisition testbeds and in a operational power control substation are also presented.

The problem of measurement effort distribution for detection of the abnormal state of distributed parameter system monitored with sensor network is considered. The measurement strategy is formulated in terms of maximizing the power of parametric hypothesis test related to the nominal system state. Then, using communication schemes based on the class of so-called gossip algorithms a computational procedure for optimizing the measurement effort over the sensor network is proposed. Finally, the presented fault detection approach is verified on the example of convective-diffusion process.

Article presents a comparison of process anomaly detection in nuclear power plant steam turbine using combination of data driven methods. Three types of faults are considered: water hammering, fouling and thermocouple fault. As a virtual plant a nonlinear, dynamic, mathematical steam turbine model is used. Two approaches for fault detection using one class and two class classifiers are tested and compared.

In this paper a path planning algorithm for the ship collision avoidance is presented. Tested algorithm is used to determine close to optimal ship paths taking into account changing strategy of dynamic obstacles. For this purpose a path planning problem is defined. A specific structure of the individual path and fitness function is presented. Principle of operation of evolutionary algorithm and based on it dedicated application vEP/N++ is described. Using presented algorithm the simulations on close-to-real sea environment is performed. Tested environment presents the problem of avoiding one static obstacle representing island and two dynamic objects representing strange ships. Obtained results proof that used approach allows to calculate efficient and close-to-optimal path for marine vessel in close-to-real time.

The paper deals with the problem of sensor fault-tolerant control for non-linear MIMO systems. The proposed strategy is based on fault estimation strategy. The $$ {\mathcal{H}}_\infty $$ approach is used to design the observer. Subsequently, the control strategy for the system with faulty sensors is proposed. The fault-tolerant controller is designed in such a way as to achieve the $$ {\mathcal{H}}_\infty $$ performance and tolerate predefined sensor faults. The final part shows an illustrative example with an implementation to a twin-rotor system.

In the paper very preliminary idea of fault tolerant industrial robot is presented. It is noticed that, the most natural way of fault tolerance - redundancy, is not appropriate in case of the robotics area. Using on the production line redundant robots (treated overall as the actuators) is unprofitable from economic point of view. Moreover, redundancy of the actuators (seeing as the parts of the robot) is mostly impossible from construction point of view. Hence, the solution of fault tolerance is offered in the way of adequate control system. In the first chapter problem statement is done and the general idea of solving it is presented. Following, idea of control system is briefly mentioned. In the next unit simple graphs presents the robot’s workspace in selected case of the fault. The wider idea of fault tolerance control in case of industrial manipulator is given in the summary.

Safety and security aspects consist of two different group of functional requirements for the control and protection systems. It is the reason why the analyses of safety and security shouldnt be integrated directly. The paper proposes extension of the currently used methods of functional safety analyses. It can be done with inclusion of the level of information security assigned to the technical system. The article addresses some important issues of the functional safety analysis, namely the safety integrity level (SIL) verification of distributed control and protection systems with regard to security aspects. A method based on quantitative and qualitative information is proposed for the SIL (IEC 61508, 61511) verification with regard of the evaluation assurance levels (EAL) (ISO/IEC 15408) and the security assurance levels (SAL) (IEC 62443).

Diagnostic knowledge about chosen technical classes of objects can be effective gained by analyzing Internet webpages. In this paper for analyzing these data is proposed the data granulation method. Information granules are mathematical models describing data aggregates. Data aggregates are connected with each other and described by the Fuzzy Description Logic. It is presented that this data granulation model can be used to sharpen the diagnostic knowledge.

A new development stage of cybernetics (the so-called cybernetics 2.0) - as a science on general regularities of systems organization and control – is discussed. A fruitful combination of organization and control within cybernetics 2.0 would give a substantiated and efficient answer to the primary question of activity systems engineering: how should control systems for them be constructed? Actually, this is a “reflexive” question related to second-order and even higher-order cybernetics. Mankind has to learn to design and implement control systems for complex systems (high-technology manufacturing, product life cycle, organizations, regions, etc.), similarly to the existing achievements in technical systems engineering!

We consider a fuzzy variant of the permutational flow shop scheduling problem in which uncertain values of operations durations are represented by fuzzy numbers. Since even deterministic problem is strongly NP-hard, we propose a tabu search based metaheuristics with the new fuzzy blocks properties for neighborhood searching acceleration.

In this paper the industrial platform for rapid prototyping of intelligent real-time monitoring and diagnostic system was proposed. Its architecture is ready to utilize advanced computational intelligence methods, especially devoted to novelty detection such as autoassociative neural network, local outlier factor, one-class support vector machines, or to solve multiclass classification problems. The rapid prototyping tool set based on Matlab/Simulink and industrial automation equipment was described in details. As an example of the use of the proposed platform, CNC milling tool head mechanical imbalance online prediction system was described.

The paper concerns an original evolutionary music composition system. On the basis of available solutions, we have selected a finite set of music features which appear to have a key impact on the quality of composed musical phrases. Evaluation criteria have been divided into rule-based and statistical sub-sets. Elements of the cost function are modeled using a Gaussian distribution defined by the expected value and variance obtained from an analysis of recognized music pieces. An evolutionary algorithm, considering a reference sequence of chords as an input, is created, implemented and tested. The results of a sampling survey (poll) proves that the melodies generated by the system arouse the interest of a listener.

An unconditionally stable finite difference discretization motivated by the well-known Crank–Nicolson method is used to develop an Iterative Learning Control (ILC) design for systems whose dynamics are described by a fourth-order partial differential equation. In particular, a discrete in time and space model of a deformable rectangular mirror, as an exemplar application, is derived and used in the ILC design. Finally, the feasibility of the new ILC design is confirmed by numerical simulations.

This paper develops an Iterative Learning Control (ILC) design applied to an example, active long ladder circuits, of spatially interconnected systems. The design is based on writing the dynamics as those of an equivalent standard differential linear system and then converting to the discrete domain. For this reason, a standard linear systems equivalent model is developed and then discretized. A numerical case study is also given to illustrate the new design.

This paper addresses the design of learning filters for a class of iterative learning control (ILC) schemes. In particular, the paper develops a method for the design of finite impulse response (FIR) filters to approximate the inverse of the dynamics resulting from the feedback controller design. The filter design is linear matrix inequality (LMI) based and guarantees convergence of the ILC scheme. Also application of the generalized Kalman-Yakubovich-Popov (KYP) lemma allows the inclusion of finite frequency range performance specifications. Finally, a simulation study illustrate the effectiveness of the new design procedure.

This paper presents an example of adaptive fuzzy control design for an unstable plant with a pole changing location in the right half-plane. The design procedure utilizes the Nyquist and the circle stability criteria that can be graphically tested using Nyquist plots. It is assumed that the function of the fuzzy controller is a nonlinearity described by a sector condition. An adaptation mechanism ensures that during adaptation this function stays in a safe sector.

The task for detection of atypical elements is one of the fundamental tasks of contemporary data analysis, finding applications in numerous problems in practically all areas of sciences and engineering. As an example, in the classic approach of automatic control, e.g. fault detection problems, the appearance of an unusual value of a vector describing a system’s technical state may testify to the occurrence of a malfunction. This paper presents a procedure for the detection of atypical elements, understood in the sense that they happen rarely. Particularly, in the case of multimodal distributions with more distant factors, such an approach allows atypical elements to be located not only in peripheral regions, but also potentially inside, between modes. The outcome indicating whether an examined observation should be classed as atypical is defined here in fuzzy and intuitionistic fuzzy forms.

The paper presents a novel, transistor level, implementation of selected fuzzy set operators suitable for fuzzy control systems realized in low-power hardware. We propose a fully digital, asynchronous realization of basic fuzzy logic (FL) functions, such as the bounded sum, bounded difference, bounded product, bounded complement, fuzzy logic union (MAX) and fuzzy logic intersection (MIN). All of the proposed operators has been implemented in the CMOS TSMC 180nm Technology and verified by means of transistor level simulations in Hspice environment. The proposed structures of the FL functions can easily be scaled to any signal resolutions.

In the paper we study some control properties of a two compartmental model of response to anticancer treatment which combines antiangiogenic and cytotoxic drugs and takes into account multiple delays in control. Mode precisely we formulate sufficient local controllability conditions for semilinear systems which result from the use of Hakufeldt et al. model, endowed with two control variables representing antiangiogenic modality, combined with chemotherapy which contain delays related to PK/PD effects and some clinical recommendations e.g. normalization of vascular network. Then the optimized protocols of the combined therapy for the model, considered as solutions of an optimal control problem for dynamical systems with delays in control, are found using necessary conditions of optimality and numerical calculations . Structural sensitivity of considered control properties and optimal solutions are also discussed.

Glucose regulation system is one of the most important problems in ICUs. Nowadays either frequent testing and insulin shots or automatic insulin pumps are used to avoid hyperglycemia even for patients without history of diabetes. As it is impossible to directly measure the glucose level, there is a need for estimation tools ensuring appropriate quality of measurements.In this paper the application of Hybrid Newton Observer is proposed. We use Intensive Care Units Minimal Model in order to estimate the value of sugar level. Three possible approaches to calculating the Jacobian are considered as this task consists the main issue in Newton observer implementation. The analysis is then confirmed with simulation.

Malignant melanomas are the most deadly type of skin cancers. Early diagnosis is a key for successful treatment and survival. The paper presents the system for supporting the process of diagnosis of skin lesions in order to detect a malignant melanoma. The paper describes the development process of an intelligent system purposed for the diagnosis of malignant melanoma. Presented system can be used as a decision support system for primary care physicians and as a system capable of self-examination of the skin with usage of dermatoscope. The system utilizes computational intelligence methods for proper classification of the dermoscopic features extracted from the medical images. The paper also proposes the extension of the well know ABCD method used for malignant melanoma diagnosis. The proposed system is tested on 126 and trained on 80 skin moles and the obtained results are very promising.

Functional neurosurgery is used for treatment of conditions in central nervous system that arise from its improper physiology. One of the possible approaches is Deep Brain Stimulation (DBS). In this procedure a stimulating electrode is placed in desired brain’s area to locally affect its activity. Among others, DBS can be used as a treatment for dystonia, depression, obsessive-compulsive disorder (OCD) and Parkinson’s Disease (PD). In this paper authors focus on application of classifiers in Deep Brain Stimulation (DBS) for Parkinson’s Disease (PD). In neurosurgical treatment of the Parkinson’s Disease the target is a small (9 x 7 x 4 mm) deeply in brain situated structure called Subthalamic Nucleus (STN). The goal of the Deep Brain Stimulation is the precise permanent placement of the stimulating electrode within target nucleus. As this structure poorly visible in CT or MRI it is usually stereotactically located using microelectrode recording. Several microelectrodes are parallelly inserted into the brain and then in measured steps advanced towards expected location of the nucleus. At each step, usually from 10 mm above expected center of the STN, the neuronal activity is recorded. Because STN has a distinct physiology, the signals recorded within it also present specific features. By extraction certain attributes from recordings provided by the microelectrodes, it is possible to construct a binary classifier that provides useful discrimination. This discrimination divides the recordings into two classes, i.e. those registered within the STN and those registered outside of it. From this it is known which microelectrodes and at which depths have passed through the STN and thus a physiological map of its surrounding is made.

The paper presents utilization of one of the latest tool from the group of Machine learning techniques, namely Deep Convolutional Neural Networks (CNN), in process of decision making in selected medical problems. After the survey of the most successful applications of CNN in solving medical problems, the paper focuses on the very difficult problem of automatic analyses of the skin lesions. The authors propose the CNN structure and the way to cope with the insufficient number of learning data. The research was carried out and validated on the data base of over 10000 images. The efficiency of the proposed approach reaches 84%.

Quality of university education influences the economic competitiveness in knowledge-based societies. Unfortunately the curricula in engineering are often too slow to responding to the actual research and industrial needs, therefore fail in shaping the careers of tomorrow. The European Commission supports EU countries and higher education institutions in modernizing education programmes to provide graduates with university degrees and employable skills. This paper presents the experience gained during realization of specially designed international European Master on Advanced Robotics – EMARO+ (Erasmus+ program, former Erasmus Mundus program) supported by the European Commission. EMARO+ accepts only the best applicants from all over the World. The paper presents the program structure and the candidate selection criteria. The program is compared with other higher degree studies in robotics, available to international students. Examples of thesis are discussed. The current research trends in robotics are summarized to illustrate the needs of modern curricula in robotics. Short discussion of the key-factors influencing the quality of internationalization in the teaching process are considered.

Software projects have become an integral part of undergraduate and graduate engineering programs at technical universities. This trend in engineering education is driven by the increasingly usage of computer technology in almost every aspect of daily life. Based on the teaching experience in the last few years, it has been found out, that a surprising number of software related projects conducted by students at technical universities had been delivered late. Such situations usually stand for lower rating for the project, retest, or even prevent being promoted to the next study level. The result of the analysis has shown that lack of effort estimation or inaccurate estimation can be considered as the main reasons preventing successful and on-time project completion. In the paper, a course is proposed that covers required topics for understanding effort estimation concept. The majority of the course is based on the case studies of software engineering practices. The course is completed by a teacher guide to manage student software projects.