Information Systems Architecture and Technology: Proceedings of 39th International Conference on Information Systems Architecture and Technology – ISAT 2018

In presented paper, an analysis of model order reduction (MOR) techniques applied to steel beams filled with a composite material is presented. This research concerns specific construction solutions used in technological machines. The analyzes concern three reduction methods: Guyan reduction also referred as static condensation, Craig-Bampton reduction and Kammer reduction. These techniques are applied to matrix equations describing steel beams filled with a composite material model, established by the finite element method (FEM). The article contains information about preparation of the full model and model parameters identification process. To verify FEM model quality its results are compared to experimental modal analysis results. The analysis compares and contrasts the MOR techniques by considering the nature of the individual algorithms and analyzing results of numerical example. The comparison of reduced models computational time at subsequent stages have also been made.

Between the recent advances in manufacturing engineering stands Tailored Forming, a process chain that produces massive hybrid material components through the use of different forming techniques. The motivation behind such a process is the achievement of higher performance parts, such as lightweight or local integrated functions. Thereby, new restrictions take place in the design of these parts, requiring the implementation of a suitable multi-material design methodology to attend user requirements. One of these new challenges is the design of the joining zone between the two metals, which presents limited controllability during the manufacturing process. With this objective, here is proposed the use of a Case Based Reasoning (CBR) system as design method. For that, a parametric model is created and, through an interface between CAD and finite element systems, a solution space is generated and analyzed, forming the first case-base. A comparison analysis of these results is executed, bringing valuable information for the current research. At the end, a similarity method is implemented in order to propose the most suitable solution among all variations based on specified requirements. With that, this tool will support the user on the creation of new cases and the machine learning process on storing the knowledge.

Oscillators form a very important part of RF circuitry. Several oscillator designs exist among which the Colpitts oscillator have gained widespread application. In designing Colpitts oscillator, different methods have been suggested in the literature. These ranges from intuitive reasoning, mathematical analysis, and algorithmic techniques. In this paper, a new meta-heuristic Bat Algorithm (BA) is proposed for designing Colpitts oscillator. It involves a combination of BA and Artificial Neural Network (ANN). BA was used for selecting the optimum pair of resistors that will give the maximum Thevenin voltage while ANN was used to determine the transient time of the optimized pairs of resistors. The goal is to select, among the several optimized pairs of resistors, the pair that gives the minimum transient response. The results obtained showed that BA-ANN gave a better transient response when compared to a Genetic Algorithm based (GA-ANN) technique and it also consumed less computational time.

In this research paper we investigate the procedure design and the implementation in a real time MATLAB SIMULINK R2017a simulation environment of an accurate adaptive observer state estimator. The effectiveness of the observer state estimator design is proved through intensive simulations performed to estimate the state-of-charge of a lithium-ion rechargeable battery integrated in a hybrid electric vehicle Battery Management System structure for a particular Honda Insight Japanese car. The state-of-charge is an essential internal parameter of the lithium-ion battery, but not directly measurable, thus an accurate estimation of battery state-of-charge becomes a vital operation for the Battery Management System. This is the main reason that motivates us to find the most suitable state-of-charge estimator in terms of estimation accuracy, fast convergence and robustness to the possible changes in the state-of-charge initial value, to the temperature effects on the battery, changes in the battery internal resistance and nominal capacity.

In the article the influence of the basic functions used to represent a polynomial from the current spline link on the approximating properties of the semi-local smoothing spline, proposed by D.A. Silaev, is studied. When constructing splines of this type, a recurrence formula that binds the group of coefficients of a polynomial from a previous spline link with a similar group of polynomial coefficients from the current spline link, is used. Silayev D.A. were studied the properties of the spline using only a power basis.It is shown that the study of the magnitude of the eigenvalues of the matrix of stability, which is used in the algorithm for constructing the investigated spline, is not enough to predict the approximation properties of this spline. The accuracy of the approximation is also significantly influenced by the number of conditionality of the matrix, which is a block from the traditional matrix of the least squares method. It is shown that the transition to polynomials, which are presented in the form of certain Hermite polynomials, is expedient. When using semi-local splines, the number of spline units decreases in comparison with the interpolation spline on the same grid. But a significant reduction in the description of the spline does not lead to a marked deterioration in the accuracy of the solutions of boundary value problems solved with the help of splines of the investigated species. The obtained theoretical results are confirmed at solving practical problems.

An extending approach to the concept of the product of context and ambiguous. In this article, we present the concept of overlap product, where contextual words are inserted among code words and strings reduced by the common overlapping context. Thus, the concept of code on the basis of overlap product is created, also called O–code. The initial results on the properties and conditions for decoding are the basis to establish an effective algorithm for testing of O–codes with the complexity of n3.

This work is presenting new properties of vertices of a dimensional unit cube obtained after mutually unambiguous (bijective) transformation of these vertices of a cube into a tree. Some of the presented properties were obtained with the Newton symbol based on an extended definition.

Advanced driving assistance systems (ADAS) could perform basic object detection and classification to alert drivers for road conditions, vehicle speed regulation, and etc. With the advances in the new hardware and software platforms, deep learning has been used in ADAS technologies. Traffic signs are an important part of road infrastructure. So, it is very important task to detect and classify traffic signs for autonomous vehicles. In this paper, we firstly create a traffic sign dataset from ZED stereo camera mounted on the top of Racecar mini autonomous vehicle and we use Tiny-YOLO real-time object detection and classification system to detect and classify traffic signs. Then, we test the model on our dataset in terms of accuracy, loss, precision and intersection over union performance metrics.

Medical research is not only expensive but also time-consuming, what can be seen in the queues, and then after the waiting time for the analysis of the effects obtained from tests. In the case of computed tomography examinations, the end result is a series of the described images of the examined object’s shape. The description is made on the careful observation of the results.In this work, we propose a solution that allows to select images that are suspicious. This type of technique reduces the amount of data that needs to be analyzed and thus reduces the waiting time for the patient. The idea is based on a three-stage data processing. In the first one, key-points are located as features of found elements, in the second, images are constructed containing found areas of images, and in the third, the classifier assesses whether the image should be analyzed in terms of diseases. The method has been described and tested on a large CT dataset, and the results are widely discussed.

The singular value decomposition (SVD) is an important tool for matrix computations with various uses. It is often combined with other methods or used within specific procedures. The text briefly introduces the SVD and lists its important features and selected elements of the SVD theory. In addition, the text deals with two important issues related to the field of artificial intelligence with extensive practical use. The first is face recognition analysis in relation to face representation using principal component analysis (PCA) and the second is fractional order singular value decomposition representation (FSVDR) of faces. The presented procedures can be used in an efficient real-time face recognition system, which can identify a subject’s head and then perform a recognition task by comparing the face to those of known individuals. The essence of the procedures, way of their application, their advantages and shortcomings, and selected results are presented in the text. All procedures are implemented in MATLAB software.

At present, school-age children are regularly exposed to a significant number of negative factors during their school time. The impact of these factors lead to the produce of an organism’s response, called stress. Regular stress can in turn cause a deterioration in the health of children.In this paper we propose a new approach to the use of physical indicators that are tracked using a fitness bracelets. In this work, we obtain and analyze the student’s physical activity indicators dynamics. The relationship between the physical condition and the level of psychological loading of schoolchildren during school lessons is analyzed. At the result was developed model for determining stress conditions in school-age children, and a web-service for analyzing and assessing the psychophysiological stress of school-age children is implemented.

The paper presents the results of a comparative study of the artifact subspace re-construction (ASR) method and two other popular methods dedicated to correct EEG artifacts: independent component analysis (ICA) and principal component analysis (PCA). The comparison is based on automatic rejection of EEG signal epochs performed on a dataset of motor imagery data. ANOVA results show a significantly better level of artifact correction for the ASR method. What is more, the ASR method does not cause serious signal loss compared to other methods.

The article presents some results of dynamical objects identification technology based on coincidence matrixes of templates and tested objects’ amplitude-phase images (APIm) calculated with discrete Hilbert transforms (DHT). DHT algorithms are modeled on basis of isotropic (HTI), anisotropic (HTA), generalized transforms – AP-analysis (APA) and the difference (residual) relative shifted phase (DRSP-) images to calculate the APIm. The identified objects are recognized as members of classes modeled with 3D templates – images of different types airplanes rotated in space. The dynamic anisotropic properties of APIm causes the increasing of sensitivity to circular angle rotation and make possible effective classification of tested objects at DHT domains. Methods to objects and templates matching accuracy increasing are based on calculations and correlation of intra- and inter-classes coincidence matrixes.

Assessment of scientific achievements of scientists is difficult because the science is divided into scientific domains and disciplines. The classification is not a partition, so very often disciplines are related to a few scientific domains. The paper presents the method of calculating scientists’ contributions to science, which are based on the number of articles published in journals connected to disciplines which are, in turn, related to scientific domains. The application of fuzzy relations and their composition simplifies the problem of describing these connections. The idea of the scientific contribution unit and the usage of the optimistic fuzzy aggregation norm allows calculating the scientific contribution of each scientist. Since levels of scientific contributions belong to the interval [0,1], there is a possibility to prepare rankings of scientists. The example of the application of this method is supported by the result of the estimation of scientific achievement by the real scientist.

A flow production system with concurrently executed supply chains providing material handling/transportation services to a given set of workstations is analyzed. The considered streams of split and merge supply chains representing all the stages at which value is added to a manufacturing product (including the delivery of raw materials and intermediate components are scheduled under constraints imposed by the solution to an associated milk-run vehicle routing problem. A declarative model of the investigated milk-run delivery principle makes it possible to formulate a vehicle routing and scheduling problem, the solution to which determines the route, the time schedule, and the type and number of parts that different trucks must carry to fulfill orders from various customers/recipients. The goal is to find solutions that minimize both vehicle downtime and the takt time of the production flow. The approach proposed allows to view the above trade-off-like problem as a constraint satisfaction problem and to solve it in the Oz Mozart constraint programming environment.

The topic of unmanned aerial vehicle (UAV) routing is transitioning from an emerging topic to a growing research area with UAVs being used for inspection or even material transport as part of multi-modal networks. The nature of the problem has revealed a need to identify the factors affecting the energy consumption of UAVs during execution of missions and examine the general characteristics of the consumption, as these are critical constraining factors in UAV routing. This paper presents the unique characteristics that influence the energy consumption of UAV routing and the current state of research on the topic. This paper provides the first overview of the current state of and contributions to the area of energy consumption in UAVs followed by a general categorization of the factors affecting energy consumptions of UAVs.

The paper presents a case study which illustrates a possible application of Agile approach to crisis management. The proposal of such a merger was done in another paper, here its implementation to epidemic crisis management is described. It is shown that the nature of activities during the epidemic crisis suits the Agile philosophy fairly well and that the application of Agile frameworks to epidemic crisis management may substantially increase its efficiency, mainly due to the communication patterns which are required by the Agile approach.

The paper deals with a problem of an emergency power supply in the case of a blackout. It needs to be decided which of the power consuming devices should stay active and under what modes of operation. For this purpose a decision making problem with multiple criteria such as cost, systems operation time and priority usage, was formulated. It was assumed that the information about the recovery time is given by an expert in the form of certainty distributions. Then the results were provided under the assumption that the planned execution time is not shorter than the estimated recovery time with a given certainty threshold. This methodology is illustrated with a computational example.

The purpose of this paper is to present the concept of modules and interfaces for a hybrid simulation model that forecasts demand for healthcare services on the regional level. The interface, developed with the Visual Basic for Application programming tools for spreadsheets, enables comprehensive planning of simulation experiment for the combined model that operates based on two different simulation paradigms: continuous and discrete-event. This paper presents the capabilities of the developed tools and discusses the results of the conducted experiments. The cross-sectional age-gender specific demographic parameters describing population of two subregions of Lower Silesia were calculated based on historical data retrieved from Central Statistical Office databases. We demonstrated the validity of the developed interface. The model correctly responded to the seasonal increased intensity of patients arrivals to healthcare system.

Medium-term electric energy demand forecasting is becoming an essential tool for energy management, maintenance scheduling, power system planning and operation. In this work we propose Generalized Regression Neural Network as a model for monthly electricity demand forecasting. This is a memory-based, fast learned and easy tuned type of neural network which is able to generate forecasts for many subsequent time-points in the same time. Time series preprocessing applied in this study filters out a trend and unifies input and output variables. Output variables are encoded using coding variables describing the process. The coding variables are determined on historical data or predicted. In application examples the proposed model is applied to forecasting monthly energy demand for four European countries. The model performance is compared to performance of alternative models such as ARIMA, exponential smoothing, Nadaraya-Watson regression and neuro-fuzzy system. The results show high accuracy of the model and its competitiveness to other forecasting models.

Unmanned Aerial Vehicles (UAV) routing is transitioning from an emerging topic to a growing research area and one critical aspect of it is the energy consumption of UAVs. This transition induces a need to identify factors, which affects the energy consumption of UAVs and thereby the routing. This paper presents an analysis of different parameters that influence the energy consumption of the UAV Routing Problem. This is achieved by analyzing an example scenario of a single UAV multiple delivery mission, and based on the analysis, relationships between UAV energy consumption and the influencing parameters are shown.

The paper is devoted to a quantitative approach to support one of the most important problem solution in the area of defense planning process - armed forces (AF) structure optimization. The MUT team, taking part in Polish Strategic Defense Review has proposed a set of methods and tools to support the analyses for the evaluation of required capabilities of Polish Armed Forces in predicted security environment. The set of methods and tools presented in the paper is limited to AF structure optimization problem. The idea of optimization and particular components of the conflict model and methods of solving the problems are presented in the sequence of steps. The structure of AF is fixed for defined threat scenarios under the financial constraints or without such limitations. The measures of combat power of weapon systems for different participants of probable conflict and some important parameters like terrain or type of operation factors (multipliers) are defined and presented. The experimental results of the allocation process are based on the hypothetical conflict evaluation.

This paper explores the problem of the externalisation of expert knowledge in order to build a new procedure for a given work place. In the first stage of our approach, the processes carried out by experts in the company are verbally described by them and recorded. Next, using Google Cloud Speech Api technology, the voice recording is converted into text. In the second stage, based on the set of the text, a Finalised Word Dictionary is created. In the third stage the steps constituting the procedure are distinguished and finally the procedure for a given work place, based on the expert’s vocalised report is created. The proposed application can be a useful tool for the acquisition of new explicit knowledge within an enterprise.

Hazards are present in all workplaces and can result in serious injuries, short and long-term illnesses, or death. In this context, management of safety is essential to ensure the occupational health of workers. Aiming to assist the safety management process, especially in industrial environments, a Cognitive Vision Platform for Hazard Control (CVP-HC) is proposed. This platform is a Cyber Physical system, capable of identifying critical safety behaviors overcoming the limitations of current computer vision systems. In addition, the system stores experiential knowledge about safety events in an explicit and structured way. This knowledge can be easily accessed and shared and may be used to improve the user/company experience as well as to understand the company safety culture and to support a long term change process. The CVP-HC is a scalable yet adaptable system capable of working in a variety of video analysis scenarios whilst meeting specific safety requirements of companies by modifying its behavior accordingly. The proposed system is based on the Set of Experience Knowledge Structure (SOEKS) and Decisional DNA (DDNA).

The ability to manage Big Data (BD) is regarded as the main driving force behind organization’s development, as well as maintaining its place in the market and of its innovative success. However, when we look on the number of organizations that actually use BD, it is relatively limited for the moment. It is mainly a lack of appropriate approaches, methods and tools targeted at the use of BD, but above all, there is a low level of knowledge about the importance of BD for business. This paper: (i) provides an overview of the context of BD in contemporary business; (ii) discusses the key concepts which underlay the concept of BD, (iii) examines selected methods and tools which are available for BD analysis. Based on the above, the paper then uses the job portals for the ICT market as a case to propose and to demonstrate a specific approach for BD analysis. The results from this study may be valuable for academics and ICT professionals, who search for appropriate methods and tools to manage job portals. This study provides also a value for employees who want to better know about the requirements and needs of the ICT job market as well as for employers who want to recruit more well-oriented candidates.

High utility itemset mining (HUIM) is a popular and important mining task in recent years. The problem is considered computational expensive in terms of execution time and memory consumption. Many algorithms have been proposed to solve this problem efficiently. In this paper, we propose a parallel approach for mining HUIs, which utilizes the modern multi-core processors by splitting the search space in to disjointed sub-spaces, assign them to the processor cores and explore them in parallel. Experimental results show that the proposed algorithm outperformed the original state-of-the-art HUIM algorithm EFIM in terms of execution times and have comparable memory usage.

Storage of indexing structures in the Vector Space Model (VSM) form has a number of advantages. In the case when text documents are considered, the indexing structure states the Term-By-Document (TBD) matrix. Its size is proportional to the product of the indexed documents number and the keywords number. In the case of large text documents databases, the size of the indexing structure is a serious limitation. Too large TBD matrix may not be able to be stored in memory or the process of searching for documents may take too much time. The article presents a methodology that allows to reduce the size of the large TBD matrix. The operation performed on the TBD matrix is the Singular Value Decomposition (SVD). It allows to transform the original indexing structure vectors into a space with fewer dimensions. As a result of the operation, keywords used in the indexing process are generalized. This is a desirable effect, methods for generalizing the keywords are called the Latent Sematic Indexing (LSI) methods. Despite the undeniable advantages of the SVD decomposition, it has a big disadvantage. Its computational complexity is O(n3). In practice, this prevents the application of the method to a large indexing structure. The methodology presented in the article assumes the use of the Epsilon decomposition in order to divide the original TBD matrix into parts before the reduction process. The proposed modification allows the use of the SVD decomposition for the indexing structure of any size.

We consider a problem of identifying an unknown distribution of a random variable based on its single observation. We present known results of constructing a minimax decision rules for one-dimensional case equivalent to a problem of equitable optimal partitioning of a measurable space. An example of finding a minimax decision rule for two-dimensional case is given.

In this paper, we propose a decision-making model based on the architecture of a three-layer perceptron with diagonal weighted synaptic connections between the neurons of the input, the latent and the original layers. The evolution of the model is carried out as a task of adaptation of the neural network, which consists of procedures for correction of the number of synaptic connections between the neurons of the input hidden and output layers due to the diagonalization of the matrices of synaptic connections in the basis of the input vector vectors. It is shown that the time of decision making in the diagonalized three-layer neural network is smaller in comparison with the time in the non-diagonalized.

Probabilistic Learning Task is a game that serve psychiatrists and psychologists to measure some cognitive abilities of people having various cognitive disorders. Mathematical models together with machine learning techniques are routinely used to summarize large amount of data produced by players during the game. Parameters of mathematical models are taken to represent behavioral data gathered during the game. However, there is no study of reliability of those parameters available in literature. We investigate how much one can trust the values of models parameters. We proposed a specific method to assess reliability of models parameters, that makes use of the game sessions of human players and their virtual counterparts.

Mutation testing is the most effective technique for assessing the quality of test suites, but it is also very expensive in terms of computational costs. The cost arises from the need to generate and execute a large number of so called mutants. The paper presents and evaluates a machine learning approach to dealing with the issue of limiting the number of executed mutants. The approach uses classification algorithm to predict mutants execution results for a subset of the generated mutants without their execution. The evaluation of the approach takes into consideration two aspects: accuracy of the predicted results and stability of prediction. In the paper the details of the evaluation experiment and its results are presented and discussed. The approach is tested on four examples having different number of mutants ranging from 90 to over 300. The obtained results indicate that the predicted value of the mutation score is consistently higher then the actual one thus allowing for using the results with high confidence.

The paper presents a modification of a dynamic programming approach employed for decision rules optimization with respect to their length. There are two aspects taken into account: (i) consideration on the length of approximate decision rules and (ii) consideration on the size of a directed acyclic graph constructed by the modified algorithm.

Refugee crisis has escalated into a leading crisis in recent years, including Europe since 2015 after the massive refugee and migrant sea arrivals in the Mediterranean. Therefore, its socio-economic and environmental impact requires complex decision making for the delivery of effective humanitarian aid operations. Refugee settlement and shelter is an operations sector where the application of multi-criteria decision making (MCDM) methods seems appropriate. Additionally, the range of involved decision makers as well as their relationships can be addressed using a multi-agent system (MAS). Different decision making fuzzy methods have been proposed in the literature which can be used by the agents in order to address refugee siting. The purpose of this paper is to perform a comparative analysis of two such methods, namely, hierarchical fuzzy TOPSIS and fuzzy axiomatic design approach used in a MAS for refugee siting. The comparative study has been done by evaluating operating temporary sites in Greece and the obtained results reflect the current situation.

The study focuses on two operators of a genetic algorithm (GA): a crossover and a mutation in the context of machine learning of fuzzy logic rules. A decision support system (DSS) is placed in a simulation environment created in accordance with the complex adaptive system (CAS) concept. In a multi-agent CAS system, the learning classifier system (LCS) paradigm is used to develop a learning system. The aim of the learning system is to discover binary rules that allow an agent to perform efficient actions in a simulation environment. The agent’s objective is to make an effective decision on which order, from the set of the awaiting orders, should be transferred into a production zone next. The decision is based on the fuzzy logic system response. In the conducted study, two input signals and one output signal of the fuzzy logic system are considered. The concept of the presented fuzzy logic system affects the construction of rules of a specific agent. The paper focuses on the problem of coding the agent’s rules and modification of the coding by the GA.