Trajectory optimization of industrial robot arms using a newly elaborated “Whip-lashing” method
The application of the Industry 4.0′s elements—e.g., industrial robots—has a key role in the efficiency improvement of manufacturing companies. In order to reduce cycle times and increase productivity, the trajectory optimization of robot arms is essential. The purpose of the study is the elaboration of a new “whip-lashing” method, which, based on the motion of a robot arm, is similar to the motion of a whip. It results in achieving the optimized trajectory of the robot arms in order to increase velocity of the robot arm’s parts, thereby minimizing motion cycle times and to utilize the torque of the joints more effectively. The efficiency of the method was confirmed by a case study, which is relating to the trajectory planning of a five-degree-of-freedom RV-2AJ manipulator arm using SolidWorks and MATLAB software applications. The robot was modelled and two trajectories were created: the original path and path investigate the effects of using the whip-lashing induced robot motion. The application of the method’s algorithm resulted in a cycle time saving of 33% compared to the original path of RV-2AJ robot arm. The main added value of the study is the elaboration and implementation of the newly elaborated “whip-lashing” method which results in minimization of torque consumed; furthermore, there was a reduction of cycle times of manipulator arms’ motion, thus increasing the productivity significantly. The efficiency of the new “whip-lashing” method was confirmed by a simulation case study.

Automated Learning of Hungarian Morphology for Inflection Generation and Morphological Analysis
The automated learning of morphological features of highly agglutinative languages is an important research area for both machine learning and computational linguistics. In this paper we present a novel morphology model that can solve the inflection generation and morphological analysis problems, managing all the affix types of the target language. The proposed model can be taught using (word, lemma, morphosyntactic tags) triples. From this training data, it can deduce word pairs for each affix type of the target language, and learn the transformation rules of these affix types using our previously published, lower-level morphology model called ASTRA. Since ASTRA can only handle a single affix type, a separate model instance is built for every affix type of the target language. Besides learning the transformation rules of all the necessary affix types, the proposed model also calculates the conditional probabilities of the affix type chains using relative frequencies, and stores the valid lemmas and their parts of speech. With these pieces of information, it can generate the inflected form of input lemmas based on a set of affix types, and analyze input inflected word forms. For evaluation, we use Hungarian data sets and compare the accuracy of the proposed model with that of state of the art morphology models published by SIGMORPHON, including the Helsinki (2016), UF and UTNII (2017), Hamburg, IITBHU and MSU (2018) models. The test results show that using a training data set consisting of up to 100 thousand random training items, our proposed model outperforms all the other examined models, reaching an accuracy of 98% in case of random input words that were not part of the training data. Using the high-resource data sets for the Hungarian language published by SIGMORPHON, the proposed model achieves an accuracy of about 95-98%.

Applying Expert Heuristic as an a Priori Knowledge for FRIQ-Learning
Many Reinforcement Learning methods start the learning phase from an empty, or randomly filled knowledge-base. Having some a priori knowledge about the way as the studied system could be controlled, e.g. in the form of some state-action control rules, the convergence speed of the learning process can be significantly improved. In this case, the learning stage could start from a sketch, from a knowledge-base formed based upon the already existing knowledge. In this paper. the a priori (expert) knowledge is considered to be given in the form state-action fuzzy control rules of a Fuzzy Rule Interpolation (FRI) reasoning model and the studied reinforcement learning method is restricted to be a Fuzzy Rule Interpolation-based Q-Learning (FRIQ-Learning) method. The main goal of this paper is the introduction of a methodology, which is suitable for merging the a priori stateaction fuzzy control rule-base to the initial state-action-value function (Q-function) representation. For demonstrating the benefits of the suggested methodology, the a priori knowledge-base accelerated FRIQ-Learning solution of the “mountain car” benchmark is also discussed briefly in the paper

Two-Echelon Vehicle Routing Problem with Recharge Stations
In the age of the fourth industrial revolution, versatile solutions for manufacturing and logistics systems are required to increase the utilization, reliability, flexibility and cost efficiency. The diversified customers’ demands present manufacturing systems with new challenges (Tamás, 2017). Matrix production may become a suitable solution through configurable production cells, the transfer of parts and tools using automated guided vehicles (AGVs) and the separation of logistics from production. The in-plant supply of manufacturing and assembly cells in matrix production is based on autonomous electric vehicles. The design and operation of autonomous electric vehicles represent a special type of vehicle routing and scheduling problems because vehicles must be recharged and therefore the solutions of traditional routing problems cannot be used to optimize the operation of AGVs. This fact was the motivation for writing this paper. After this introduction, the remaining parts of the paper are divided into five sections. Section 2 presents a literature review to summarize the research background. Section 3 presents the model framework and mathematical model of two-echelon vehicle routing problem with recharge stations. Section 4 presents the used heuristic solutions, while section 5 presents the results of the numerical analysis. Conclusions and future research directions are discussed in the last section.

Optimization of Multi-Depot Periodic Vehicle Routing Problem with Time Window
Logistics is one of today's most important industries. It is important to store and transport raw materials, products cost-effectively. The article presents a specific delivery task, the Multi-Depot Periodic Vehicle Routing Problem with Time Window. In case of the problem several customers must be visited and must satisfy their demands. The vehicles start their route from one of the several depots, visit some customers and return to the depot from which they started their route. The customers have time window, which means that they must be visited within a predefined time interval. The periodic keyword means that the customers must be visited not once, but periodically. This means, that a periodic time is given, and the number of visits of each customer within this periodic time are also known in advance. The goal is the minimization of the length of the route. This problem is solved in this paper with construction and improvement algorithms. The presented construction algorithms are the Nearest Neighbor, Insertion Heuristics and Greedy algorithm. The presented improvement algorithms are the Firefly Algorithm, Harmony Search, Particle Swarm Optimization, Simulated Annealing and Tabu Search algorithms. Based on the test results the improvement of construction algorithms gave better performance than improving randomly generated solutions.

Economic, Social Impacts and Operation of Smart Factories in Industry 4.0 Focusing on Simulation and Artificial Intelligence of Collaborating Robots
Smart Factory is a complex system that integrates the main elements of the Industry 4.0 concept (e.g., autonomous robots, Internet of Things, and Big data). In Smart Factories intelligent robots, tools, and smart workpieces communicate and collaborate with each other continuously, which results in self-organizing and self-optimizing production. The significance of Smart Factories is to make production more competitive, efficient, flexible and sustainable. The purpose of the study is not only the introduction of the concept and operation of the Smart Factories, but at the same time to show the application of Simulation and Artificial Intelligence (AI) methods in practice. The significance of the study is that the economic and social operational requirements and impacts of Smart Factories are summarized and the characteristics of the traditional factory and the Smart Factory are compared. The most significant added value of the research is that a real case study is introduced for Simulation of the operation of two collaborating robots applying AI. Quantitative research methods are used, such as numerical and graphical modeling and Simulation, 3D design, furthermore executing Tabu Search in the space of trajectories, but in some aspects the work included fundamental methods, like suggesting an original whip-lashing analog for designing robot trajectories. The conclusion of the case study is that—due to using Simulation and AI methods—the motion path of the robot arm is improved, resulting in more than five percent time-savings, which leads to a significant improvement in productivity. It can be concluded that the establishment of Smart Factories will be essential in the future and the application of Simulation and AI methods for collaborating robots are needed for efficient and optimal operation of production processes.

On the neutrality of two symmetric TSP solvers toward instance specification
Currently, many computationally difficult problems can be solved using very efficient methods. Some of these state-of-the-art methods have online implementations. The research question addressed herein is how sensitive are such implementations if the input data are preprocessed in a specific manner? The symmetric traveling salesman problem (sTSP), which is an NP-hard problem with many real-life applications is studied. The proposed method includes systematic transformation using rotations and reflections of the vertex order of sTSP instances. This model was used for investigating the neutrality of Concorde [1] (currently the best exact sTSP solver) and Lin–Kernighan implementation [2], both from NEOS [3] (the state-ofthe-art collection of online tools in computational optimization).

Defining rough sets using tolerances compatible with an equivalence
We consider tolerances T compatible with an equivalence E on U, meaning that the relational product E ∘ T is included in T. We present the essential properties of E-compatible tolerances and study rough approximations defined by such E and T. We consider rough set pairs (XE, XT), where the lower approximation XE is defined as is customary in rough set theory, but XT allows more elements to be possibly in X than XE. Motivating examples of E-compatible tolerances are given, and the essential lattice-theoretical properties of the ordered set of rough sets {(XE, XT)∣X ⊆ U} are established.

MeasApplInt - a novel intelligence metric for choosing the computing systems able to solve real-life problems with a high intelligence
Intelligent agent-based systems are applied for many real-life difficult problem-solving tasks in domains like transport and healthcare. In the case of many classes of real-life difficult problems, it is important to make an efficient selection of the computing systems that are able to solve the problems very intelligently. The selection of the appropriate computing systems should be based on an intelligence metric that is able to measure the systems intelligence for real-life problem solving. In this paper, we propose a novel universal metric called MeasApplInt able to measure and compare the real-life problem solving machine intelligence of cooperative multiagent systems (CMASs). Based on their measured intelligence levels, two studied CMASs can be classified to the same or to different classes of intelligence. MeasApplInt is compared with a recent state-of-the-art metric called MetrIntPair. The comparison was based on the same principle of difficult problem-solving intelligence and the same pairwise/matched problem-solving intelligence evaluations. Our analysis shows that the main advantage of MeasApplInt versus the compared metric, is its robustness. For evaluation purposes, we performed an illustrative case study considering two CMASs composed of simple reactive agents providing problem-solving intelligence at the systems’ level. The two CMASs have been designed for solving an NP-hard problem with many applications in the standard, modified and generalized formulation. The conclusion of the case study, using the MeasApplInt metric, is that the studied CMASs have the same real-life problems solving intelligence level. An additional experimental evaluation of the proposed metric is attached as an Appendix.

Cloud workload prediction based on workflow execution time discrepancies
Infrastructure as a service clouds hide the complexity of maintaining the physical infrastructure with a slight disadvantage: they also hide their internal working details. Should users need knowledge about these details e.g., to increase the reliability or performance of their applications, they would need solutions to detect behavioural changes in the underlying system. Existing runtime solutions for such purposes offer limited capabilities as they are mostly restricted to revealing weekly or yearly behavioural periodicity in the infrastructure. This article proposes a technique for predicting generic background workload by means of simulations that are capable of providing additional knowledge of the underlying private cloud systems in order to support activities like cloud orchestration or workflow enactment. Our technique uses long-running scientific workflows and their behaviour discrepancies and tries to replicate these in a simulated cloud with known (trace-based) workloads. We argue that the better we can mimic the current discrepancies the better we can tell expected workloads in the near future on the real life cloud. We evaluated the proposed prediction approach with a biochemical application on both real and simulated cloud infrastructures. The proposed algorithm has shown to produce significantly (∼ 20%) better workload predictions for the future of simulated clouds than random workload selection.

An Algorithm using Context Reduction for Efficient Incremental Generation of Concept Set
The theory of Formal Concept Analysis (FCA) provides efficient methods for conceptualization of formal contexts. The methods of FCA are applied mainly on the field of knowledge engineering and data mining. The key element in FCA applications is the generation of a concept set. The main goal of this research work is to develop an efficient incremental method for the construction of concept sets. The incremental construction method is used for problems where context may change dynamically. The paper first proposes a novel incremental concept set construction algorithm called ALINC, where the insertion loop runs over the attribute set. The combination of object-level context processing and ALINC is an object level incremental algorithm (OALINC) where the context is built up object by object. Based on the performed tests, OALINC dominates the other popular batch or incremental methods for sparse contexts. For dense contexts, the OINCLOSE method, which uses the InClose algorithm for processing of reduced contexts, provides a superior efficiency. Regarding the OALINC/OINCLOSE algorithms, our test results with uniform distribution and real data sets show that our method provides very good performance in the full investigated parameter range. Especially good results are experienced for symmetric contexts in the case of word clustering using context-based similarity.

String Transformation Based Morphology Learning
There are several morphological methods that can solve the morphological rule induction problem. For different languages this task represents different difficulty levels. In this paper we propose a novel method that can learn prefix, infix and suffix transformations as well. The test language is Hungarian, and we chose a previously generated word pair set of accusative case for evaluating the method, comparing its training time, memory requirements, average inflection time and correctness ratio with some of the most popular models like dictionaries, finite state transducers, the tree of aligned suffix rules and a lattice based method. We also provide multiple training and searching strategies, introducing parallelism and the concept of prefix trees to optimize the number of rules that need to be processed for each input word. This newly created novel method can be applied not only for morphology, but also for any problems in the field of bioinformatics and data mining that can benefit from string transformations learning.

Investigating the piece-wise linearity and benchmark related to Koczy-Hirota fuzzy linear interpolation
Fuzzy Rule Interpolation (FRI) reasoning methods have been introduced to address sparse fuzzy rule bases and reduce complexity. The first FRI method was the Koczy and Hirota (KH) proposed "Linear Interpolation". Besides, several conditions and criteria have been suggested for unifying the common requirements FRI methods have to satisfy. One of the most conditions is restricted the fuzzy set of the conclusion must preserve a Piece-Wise Linearity (PWL) if all antecedents and consequents of the fuzzy rules are preserving on PWL sets at {lpha}-cut levels. The KH FRI is one of FRI methods which cannot satisfy this condition. Therefore, the goal of this paper is to investigate equations and notations related to PWL property, which is aimed to highlight the problematic properties of the KH FRI method to prove its efficiency with PWL condition. In addition, this paper is focusing on constructing benchmark examples to be a baseline for testing other FRI methods against situations that are not satisfied with the linearity condition for KH FRI.

Some Considerations and a Benchmark Related to the CNF Property of the Koczy-Hirota Fuzzy Rule Interpolation
The goal of this paper is twofold. Once to highlight some basic problematic properties of the KH Fuzzy Rule Interpolation through examples, secondly to set up a brief Benchmark set of Examples, which is suitable for testing other Fuzzy Rule Interpolation (FRI) methods against these ill conditions. Fuzzy Rule Interpolation methods were originally proposed to handle the situation of missing fuzzy rules (sparse rule-bases) and to reduce the decision complexity. Fuzzy Rule Interpolation is an important technique for implementing inference with sparse fuzzy rule-bases. Even if a given observation has no overlap with the antecedent of any rule from the rule-base, FRI may still conclude a conclusion. The first FRI method was the Koczy and Hirota proposed "Linear Interpolation", which was later renamed to "KH Fuzzy Interpolation" by the followers. There are several conditions and criteria have been suggested for unifying the common requirements an FRI methods have to satisfy. One of the most common one is the demand for a convex and normal fuzzy (CNF) conclusion, if all the rule antecedents and consequents are CNF sets. The KH FRI is the one, which cannot fulfill this condition. This paper is focusing on the conditions, where the KH FRI fails the demand for the CNF conclusion. By setting up some CNF rule examples, the paper also defines a Benchmark, in which other FRI methods can be tested if they can produce CNF conclusion where the KH FRI fails.

Fitness Landscape Analysis and Edge Weighting-Based Optimization of Vehicle Routing Problems
Vehicle routing problem (VRP) is a highly investigated discrete optimization problem. The first paper was published in 1959, and later, many vehicle routing problem variants appeared to simulate real logistical systems. Since vehicle routing problem is an NP-difficult task, the problem can be solved by approximation algorithms. Metaheuristics give a “good” result within an “acceptable” time. When developing a new metaheuristic algorithm, researchers usually use only their intuition and test results to verify the efficiency of the algorithm, comparing it to the efficiency of other algorithms. However, it may also be necessary to analyze the search operators of the algorithms for deeper investigation. The fitness landscape is a tool for that purpose, describing the possible states of the search space, the neighborhood operator, and the fitness function. The goal of fitness landscape analysis is to measure the complexity and efficiency of the applicable operators. The paper aims to investigate the fitness landscape of a complex vehicle routing problem. The efficiency of the following operators is investigated: 2-opt, order crossover, partially matched crossover, cycle crossover. The results show that the most efficient one is the 2-opt operator. Based on the results of fitness landscape analysis, we propose a novel traveling salesman problem genetic algorithm optimization variant where the edges are the elementary units having a fitness value. The optimal route is constructed from the edges having good fitness value. The fitness value of an edge depends on the quality of the container routes. Based on the performed comparison tests, the proposed method significantly dominates many other optimization approaches.

Fuzzy Rule Interpolation and SNMP-MIB for Emerging Network Abnormality
It is difficult to implement an efficient detection approach for Intrusion Detection Systems (IDS) and many factors contribute to this challenge. One such challenge concerns establishing adequate boundaries and finding a proper data source. Typical IDS detection approaches deal with raw traffics. These traffics need to be studied in depth and thoroughly investigated in order to extract the required knowledge base. Another challenge involves implementing the binary decision. This is because there are no reasonable limits between normal and attack traffics patterns. In this paper, we introduce a novel idea capable of supporting the proper data source while avoiding the issues associated with the binary decision. This paper aims to introduce a detection approach for defining abnormality by using the Fuzzy Rule Interpolation (FRI) with Simple Network Management Protocol (SNMP) Management Information Base (MIB) parameters. The strength of the proposed detection approach is based on adapting the SNMP-MIB parameters with the FRI. This proposed method eliminates the raw traffic processing component which is time consuming and requires extensive computational measures. It also eliminates the need for a complete fuzzy rule based intrusion definition. The proposed approach was tested and evaluated using an open source SNMP-MIB dataset and obtained a 93% detection rate. Additionally, when compared to other literature in which the same test-bed environment was employed along with the same number of parameters, the proposed detection approach outperformed the support vector machine and neural network. Therefore, combining the SNMP-MIB parameters with the FRI based reasoning could be beneficial for detecting intrusions, even in the case if the fuzzy rule based intrusion definition is incomplete (not fully defined).

Fuzzy Automaton as a Detection Mechanism for the Multi-Step Attack
The integration of a fuzzy system and automaton theory can form the concept of fuzzy automaton. This integration allows a discretely defined state-machine to act on continuous universes and handle uncertainty in applications like Intrusion Detection Systems (IDS). The typical IDS detection mechanisms are targeted to detect and prevent single-stage attacks. These types of attacks can be detected using either a common convincing threshold or by pre-defined rules. However, attack techniques have changed in recent years. Currently, the largest proportion of attacks performed, are multi-step attacks. The goal of this paper is to introduce a novel detection mechanism for multi-step attacks built upon Fuzzy Rule Interpolation (FRI) based fuzzy automaton. In that respect, the FRI method instruments the fuzzy automaton to be able to act on a not fully defined state transition rule-base, by offering interpolated conclusion even for situations which are not explicitly defined. In the suggested model, the intrusion definition state transition rule-base is defined using an open source fuzzy declarative language. On the multi-step attack benchmark dataset introduced in this paper, the proposed detection mechanism was able to achieve 97.836% detection rate. Furthermore, in the studied examples, the suggested method was able not only to detect but also early detect the multi-step attack in stages, where the planned attack is not fully elaborated and hence less harmful. According to these results, the IDS built upon the FRI based fuzzy automaton could be a useful device for detecting multi-step attacks, even in cases when the intrusion state transition rule-based is incomplete. The early detection of multi-step attacks also allows the administrator to take the necessary actions in time, to mitigate the potential threats.

Fuzzy Behavior Description Language: A Declarative Language for Interpolative Behavior Modeling
The behavior-based system (BBS) is a hierarchical structure built upon behavior components, behavior coordination and behavior fusion. The goal of this paper, is to recall the concept of the interpolative fuzzy behavior-based system and to introduce a declarative language especially designed for supporting its implementation and configuration into embedded applications. The suggested Fuzzy Behavior Description Language (FBDL) aids the definition of fuzzy rule-based systems and their connections to form behavior components and behavior coordination as fuzzy state-machines. The suggested language also assists the fuzzy rule definition with variable consequent, to help the creation of behavior fusion functions. For simplifying the definition of hierarchical rule-bases, the structure of rule-base dominancy is also introduced in the FBDL. According to the suggested embedded application concept, the FBDL code, as a parameter configuration, can directly "run" on a built in fuzzy state machine controller, called "FRI Behavior Engine". This case the behavior of the agent controlled by the FRI Behavior Engine, can be directly modified by changing the FBDL code, without reprogramming other parts of the agent controller software.