Complex Dynamic Objects Research Articles

Development of a method for assessing the state of dynamic objects using a population algorithm

The object of the study is complex dynamic objects with a hierarchical structure. A method for assessing the state of dynamic objects using a population algorithm is proposed. The study is based on the snake optimization algorithm for finding a solution to the state of dynamic objects with a hierarchical structure. For training snake agents (SA), evolving artificial neural networks are used. The originality of the method lies in using additional advanced procedures that allow you: – to determine the initial position of SA, taking into account the type of uncertainty by using a correction factor for the degree of awareness of the state of the initial situation in relation to the object of analysis; – to take into account the initial velocity of each SA, which allows studying complex functions; – to ensure the universality of SA food location search strategies, which allows classifying the type of data to be processed; – to adjust the SA velocity by adjusting the ambient temperature, which allows priorizing the search for a solution in a certain plane; – to explore the solution spaces of functions described by non-typical functions, using exploitation mode procedures; – to flexibly adjust the transition from the SA fighting mode to the mating mode using the food saturation coefficient; – to replace individuals unsuitable for search using the SA fertility rate; – to conduct a simultaneous search for a solution in different directions, by changing the ambient temperature and adjusting the food saturation coefficient. Modeling showed a 13–19 % increase in data processing efficiency by using additional improved procedures.

Development of a method for assessing the state of dynamic objects using a combined swarm algorithm

The object of the study is complex dynamic objects. The subject of the study is the decision-making process in the problems of managing complex dynamic objects. A method of assessing the state of dynamic objects using a combined swarm algorithm is proposed. The research is based on a combined swarm algorithm - for finding a solution to the state of dynamic objects with a hierarchical structure. To train the individuals of the combined swarm algorithm (CSA), evolving artificial neural networks are used, and to select the best in the combined swarm algorithm, an improved genetic algorithm is used. The originality of the method is: – in taking into account the type of uncertainty and noise of data during the operation of the combined swarm algorithm due to the use of appropriate correction factors; – in the implementation of adaptive strategies for the search for food sources due to setting appropriate search priorities; – in taking into account the presence of a predator while choosing food sources by the flock agents of the combined swarm algorithm, which allows excluding unwanted search areas; – in the additional consideration of the available computing resources of the state analysis system of complex dynamic objects while determining the maximum permissible parameters of the combined swarm algorithm; – in the possibility of changing the search area and speed of movement by separate individuals of the flock of the combined swarm algorithm; – in determining the best individuals of the flock of the combined swarm algorithm using an improved genetic algorithm; – in training knowledge bases, carried out by training the synaptic weights of the artificial neural network, the type and parameters of the membership function, the architecture of individual elements and the architecture of the artificial neural network as a whole. The method makes it possible to increase the efficiency of data processing at the level of 14–20 % by using additional improved procedures. The proposed method should be used to solve problems of evaluating complex dynamic objects

Development of a method for managing technical systems using a bio-inspired algorithm

Today’s management solutions depend precisely on the successful solution of optimization problems, which are discontinuous, undifferentiated and multimodal. One of the approaches to increase the efficiency of solving optimization problems is bio-inspired algorithms. The object of the study is complex dynamic objects. The subject of the study is the decision-making process in the problems of managing complex dynamic objects. A management method using a bio-inspired algorithm is proposed. The research is based on the goose algorithm – for finding a solution to the state of dynamic objects with a hierarchical structure. Evolving artificial neural networks are used to train goose agents (GA) and an advanced genetic algorithm is used to select the best ones in the combined swarm algorithm. The originality of the proposed method lies in setting GA taking into account the uncertainty of the initial data, improved global and local search procedures. Also, the originality of the study lies in determining GA food locations, which allows choosing the priority of search in a given direction. The next element in the originality of the study is the ability to determine the indicators of guard GA, which allows adjusting the amount of time during which the GA flock will be located. Another original element of the study is the determination of the initial velocity of each GA. This makes it possible to optimize the speed of conducting exploration by each GA in a certain research direction. The method allows increasing the efficiency of data processing at the level of 10–12 % by using additional improved procedures. The proposed method should be used to solve problems of evaluating complex dynamic objects

Development of an evaluation method using a combined cat swarm optimization algorithm

The object of the study is complex dynamic objects with a hierarchical structure. The problem solved in the study is to increase the efficiency of decision-making while ensuring the given reliability. The subject of the study is the process of decision-making in management problems using an improved cat swarm optimization algorithm (CSO), an improved genetic algorithm and evolving artificial neural networks. The proposed method, due to additional and improved procedures, allows you: – to take into account the type of uncertainty of the initial data for setting CA for the local search procedure; – to implement adaptive strategies for finding food sources by CA; – to take into account the experience of the most authoritative CA while conducting local and global search; – to take into account the available computing resources of the state analysis system of complex dynamic objects and determine their required amount for involvement; – to take into account the CA search priority; – to determine the best CA using an improved genetic algorithm; – to conduct training of knowledge bases, which is carried out by training the synaptic weights of the artificial neural network, the type and parameters of the membership function, and architecture of individual elements and the architecture of the artificial neural network as a whole; – to avoid the local extremum problem by using the jump procedure. The proposed method was tested on the example of solving the problem of determining the composition of an operational group of troops (forces) and elements of its operational structure. An example of using the method showed an increase in the efficiency of data processing at the level of 14–19 % by using additional improved procedures. The proposed approach should be used to solve the problems of evaluating complex and dynamic processes characterized by a high degree of complexity

Synthesis of the model of management of complex dynamic objects taking into account the events of their security

The rapid development of complex, decentralized, non-linear technical structures - robotic means urgently requires the creation of an optimal algorithmic support for an automatic situational control system of such dynamic objects, taking into account the possibility of increasing the safety of their operation. This will be a guarantee, and as a result, a significant increase in the efficiency and quality of the tasks assigned by the specified technical structures. For the practical implementation of this task, it is advisable to comprehensively consider the nonlinear model of the processes of changing the state of a complex dynamic object. It is advisable to take into account the possibility of operational automatic compensation of dangerous incidents. Such a model will become the basis for the synthesis of nonlinear synergistic situational laws of management of these structures. The difference of the proposed approach is the consideration of the influence of intensive variations of incident flows in the state management laws of nonlinear dynamic objects. Emphasis on promising areas of research, namely: the application of the obtained results to justify the requirements for the design characteristics of control systems and their algorithmic support from the point of view not only of increasing their safety of operation, but also of ensuring the specified performance indicators of a wide range of possible tasks. One of these tasks is the provision of departmental communication (for the collection, processing, storage, protection of information and its operational transmission) in the case of the use of dynamic objects as mobile aerial platforms (unmanned aerial systems (UAVs)) for the placement of special electronic communications devices.

Synthesis of the neuro-fuzzy regulator with genetic algorithm

Real-acting objects are characterized by the presence of various types of random perturbations, which significantly reduce the quality of the control process, which determines the use of modern methods of intellectual technology to solve the problem of synthesis of control systems of structurally complex dynamic objects, allowing to compensate the influence of external factors with the properties of randomness and partial uncertainty. The article considers issues of synthesis of the automatic control system of dynamic objects by applying the theory of intelligent control. In this case, a neural network based on radial-basis functions is used at each discrete interval for neuro-fuzzy approximation of the control system, allowing real-time adjustment of the regulator parameters. The radial basis function is designed to approximate functions defined in the implicit form of pattern sets. The neuro-fuzzy regulator's parameter configuration is accomplished using a genetic algorithm, enabling more efficient computation to determine the regulator's set parameters. The regulator's parameters are represented as a vector, facilitating their application to multidimensional objects. To determine the optimal tuning parameters of the neuro-fuzzy regulator, characterized by high convergence and the possibility of determining global extrema, a genetic algorithm was used. The effectiveness of the neuro-fuzzy regulator is explained by the possibility of providing quality control of the dynamic object under random perturbations and uncertainty of input data.

Chemical complexity challenge: Is multi‐instance machine learning a solution?

AbstractMolecules are complex dynamic objects that can exist in different molecular forms (conformations, tautomers, stereoisomers, protonation states, etc.) and often it is not known which molecular form is responsible for observed physicochemical and biological properties of a given molecule. This raises the problem of the selection of the correct molecular form for machine learning modeling of target properties. The same problem is common to biological molecules (RNA, DNA, proteins)—long sequences where only key segments, which often cannot be located precisely, are involved in biological functions. Multi‐instance machine learning (MIL) is an efficient approach for solving problems where objects under study cannot be uniquely represented by a single instance, but rather by a set of multiple alternative instances. Multi‐instance learning was formalized in 1997 and motivated by the problem of conformation selection in drug activity prediction tasks. Since then MIL has found a lot of applications in various domains, such as information retrieval, computer vision, signal processing, bankruptcy prediction, and so on. In the given review we describe the MIL framework and its applications to the tasks associated with ambiguity in the representation of small and biological molecules in chemoinformatics and bioinformatics. We have collected examples that demonstrate the advantages of MIL over the traditional single‐instance learning (SIL) approach. Special attention was paid to the ability of MIL models to identify key instances responsible for a modeling property.This article is categorized under: Data Science > Chemoinformatics Data Science > Artificial Intelligence/Machine Learning

SELECTION OF THE SET OF ALLOWABLE VALUES OF THE VARIABLE PARAMETERS OF THE STABILIZER OF A COMPLEX DYNAMIC OBJECT

Topicality. One of the most important tasks in the application of computational methods for the parametric synthesis of controllers of complex dynamic objects is the task of determining the set of permissible values of the variable parameters of the controller, where the target function is calculated based on the solutions of the mathematical model of the disturbed motion of the dynamic object with its subsequent minimization. The purpose of the work is to construct the set of permissible values of variable parameters of the stabilizer of a complex dynamic object when applying the algorithmic combined method of parametric synthesis of stabilizers of complex dynamic objects, the essence of which is the direct calculation of the integral quadratic functional on the solutions of a closed dynamic system with subsequent finding of its global minimum through a sequential combination of two algorithms – the Sobol grid algorithm and the Nelder-Mead algorithm. Results. With the help of the Sobol grid construction algorithm, the starting point of the computational process is brought to the node of the Sobol grid, which is located in the small vicinity of the point of the global minimum. At the second stage of optimization, the found Sobol grid node is selected as the starting point for applying the Nelder-Mead method, which is implemented by the Optimization Toolbox software product of the MATLAB package or the Minimize software product of the MATHCAD package and leads the computational process to the point of the global minimum. Conclusion. The paper proves a theorem according to which the stability region of a closed system of the first approximation can be taken as such a set, and also gives an example of a solution to the problem of parametric synthesis of the stabilizer of the car's course stability system during its emergency braking.

Development of methods for identifying the state of various dynamic objects

Artificial intelligence technologies are actively used to solve both general and highly specialized tasks. In the process of assessing (identifying) the condition of complex and heterogeneous objects, there is a high degree of a priori uncertainty regarding their condition and a small amount of initial data describing them. The trends of armed conflicts of the last decades and the regularities of the development of information systems, convincingly indicate the need to change approaches to the collection of information from various sources and their analysis. There is a constant transformation of the forms of information presentation and the order of storage and access to various types of data. The problem of integrating disparate sources of information collection into a single information space is also not fully resolved. That is why the issue of improving the efficiency of assessing the state of complex and heterogeneous dynamic objects is an important and urgent issue. The objects of research are heterogeneous dynamic objects. The subject of the research is the identification of the state of heterogeneous dynamic objects. In the research, the method of identifying the state of heterogeneous dynamic objects was developed. The novelty of the proposed method consists in: ‒ taking into account the degree of uncertainty about the state of a heterogeneous dynamic object; ‒ taking into account the degree of data noise as a result of distortion of data characterizing the state of a heterogeneous dynamic object; ‒ reducing computing costs while assessing the state of heterogeneous dynamic objects; ‒ the possibility of performing calculations with source data that are different in nature and units of measurement. It is advisable to implement the mentioned method in specialized software, which is used to analyze the state of complex technical systems and make decisions.

Modeling of multi-factory dependences in complex control systems by SUGENO fuzzy knowledge base

The creation of intelligent control systems based on soft computing for complex real-time systems is considered. Usually, the control of many objects and technological processes is performed by a human operator based on instructions and personal experience. Due to the uncertainty and incompleteness of information about the parameters of the object, the impossibility of their direct measurement and the natural diversity of the environment, the use of classical mathematical methods for the automation of control is impossible. Adaptive methods are used to solve the management problem under conditions of uncertainty. The most effective is the implementation of fuzzy control based on production rules, which does not require knowledge of the object model. The concept of fuzzy control is widely used because of its ability to operate according to conditions. The object is evaluated using fuzzy rules from the knowledge base that create a fuzzy decision domain. The stage of structural identification of the system when modeling an unknown dependency by Sugeno's fuzzy knowledge base is presented, which allows implementing a control algorithm based on a neural network. To take into account the maximum number of influencing factors, the hierarchical structure of the fuzzy control system is applied. Granulation by certain properties of the object and the environment at the level of input variables allows to reduce the dimensionality of the data vector and the number of rules of the knowledge base, making it possible to build fuzzy control systems with an unlimited number of inputs. According to this model, the decision-making process is divided into two levels. The first level reduces the amount of information to define the situations to be processed in the second level by the Sugeno knowledge base. A preliminary fuzzy model of control of complex dynamic objects with incomplete determination is built.

Estimation of pressure effect on the crystallization process control based on a generalized theoretical model of a crystallizing metal

It is installed, that on aspects such as automation and control, liquid and/or crystallizing metal is a complex dynamic object, whose properties correspond to rheological body properties. Conclusions were made that the casting structure formation during crystallization proceeds under the influence of two main factors — heat removal to the mold wall and metallostatic pressure. Theoretical studies shown that the deformation of cast metal is carried out in order to correct defects but does not give this metal greater strength. It is established, by using the results of computational and theoretical studies that for efficient control of the crystallization process, it is necessary to calculate the solidification rate, and crystallization control can be reduced to control the metastable state interval. Keywords aluminum alloys, control, pressure, crystallization theory, mathematical model, heat flow, cooling rate

MODAL SYNTHESIS OF THE OBSERVER OF THE STATE VARIABLES OF THE RIVER DREDGER AUTOMATION SYSTEM

The purpose of the work is to apply automation technologies and digitalization of the process of assessing the immeasurable variables of the state of the river dredger automation system when controlling high-tech software and hardware equipment with an electric drive, in order to increase the productivity of the vessels of the technical fleet when performing dredging operations on waterways. River dredgers are complex dynamic control objects, the performance of which in the conditions of digital transformation can be significantly increased through the use of new software and hardware, the algorithmization of which is based on neural network technologies and learning processes. To increase the efficiency and reliability of the operation of such facilities with the prospect of transition to unmanned control of the energy systems of dredgers, it is necessary to increase the information content of the complex of sensors containing information about the object state. In this regard, the use, along with the existing ones, of mathematical information sensors, which are observers, is proposed in the paper. On the example of controlling a precision electric drive, an algorithm for synthesizing a dynamic system with an observer, in which all components of the state variables vector are restored at the output coordinate, is considered. The proposed algorithm for the synthesis of the state vector evaluator is implemented in the form of a program compiled in the MATLAB codes, and it is demonstrated by the example of calculating a three-dimensional dynamic observer for the automation system of a river dredger. The obtained values of state variables and errors of their recovery, as well as the results of modeling the dynamic reactions of the automation system of the river dredger, correspond to the specified speed and required stability of the electric drive of the river dredger during its transition to a steady state under new initial conditions.

Problems of constructing logical-linguistic models of complex management dynamic objects

Objective. The aim of the study is to solve the problems of constructing logicallinguistic (fuzzy) models for controlling the trajectory movement of complex dynamic objects.Method. To build logical-linguistic (fuzzy) control models, theories of fuzzy sets and algorithms of L. Zadeh were used.Result. The process of constructing logical-linguistic (fuzzy) models for controlling the trajectory movement of complex dynamic objects using the theory of fuzzy sets and fuzzy algorithms is given. A marine mobile object (MMO) operating in an uncertain environment was used as a control object in the work. It is shown that logical-linguistic (fuzzy) MPO control models can be built taking into account the rich practical experience of a navigator who does not have special knowledge, expressed by them in a qualitative form for building fuzzy control models.Conclusion. Logicallinguistic control models make it possible to ensure the movement of a marine moving object along a given trajectory with a quality of the controlled process sufficient for practical purposes under conditions of various external disturbances and drift of the parameters of the controlled object. The use of fuzzy control algorithms can significantly reduce the cost of computer time compared to the optimal search algorithm with an acceptable deterioration in the quality of the process and ensure the implementation of the control process in real time.

Application of the neural network approach in the control sys-tem of a complex dynamic object

Application of the neural network approach in the control sys-tem of a complex dynamic object

Model and method for representing complex dynamic information objects based on LMS-trees in NoSQL databases

The article analyzes the existing approaches to the description of large dynamic information objects in the construction of Automated control systems. Introduced and defined the concept of a ComplexDynamical Information Object. A comparative analysis of the temporal complexities of tree-like structures is carried out and the optimal one for working with ComplexDynamical Information Objectis selected. Most modern automated control systems use various approaches to describe automation objects for their operation. Under the automation object, we mean functional objects that are described in the form of structural models that reflect the properties of physical objects. Thus, for optimal work with complex dynamic information objects, we have developed our own model and method for describing the LMS-tree (Log-structured merge-tree), with the ability to split and store down to elementary levels. One of the features of our approach to describing objects is the presence of tree-like levels -the so-called “leaves”, by which we mean special tree elements that expand the description of the tree structure of a particular tree level. The minimal elements of the leaves of the tree –“veins”-are details, that is, elementary information elements. A leaf is a combination of “veins”(details) according to certain characteristics, which provide extended information about the level of the tree object. An atomic-level descriptor is a multiple NoSQL database field (array) where the tree level number is the index of the database array. This approach allows you to retrieve and group objects according to the element level of the tree definition.

СПЕЦИАЛЬНОЕ МОДЕЛЬНО-АЛГОРИТМИЧЕСКОЕ И ПРОГРАММНОЕ ОБЕСПЕЧЕНИЕ ПРОАКТИВНОГО УПРАВЛЕНИЯ ГРУППОВЫМ ПОВЕДЕНИЕМ РОБОТОТЕХНИЧЕСКИХ СРЕДСТВ

The paper describes the proactive control of robots group behavior using Behavior-Based Systemmodels, where the intellect is formed by a physical entities behavior. The observed complex is anarray of distributed agents functioning in real time under disturbances. John Boyd’s OODA loopmodel is used to describe the control system work cycle of such network object. The input data of thecontrol task are a planning horizon, a group action scenario, an array of agents and their possibleelementary operations, a set of scenarios using restrictions and a quality indicator of the controlproblem solution. The output data is the distribution plan of agents in space and time to realize thescenario under restrictions. The developed technology predicts the environmental disturbances. Thecomplex predictive modeling methodology for a self-organized robots group control is used withlogical-dynamic models. One of the key advantages of developed combined models, methods, algorithmsand software is the possibility to coordinate analytical and simulation control models of complexdynamic objects and their logical-algebraic analogs and models based on intelligent informationtechnology. This coordination is on the conceptual, model-algorithmic, information and softwaredetailing levels. The special language for modeling, planning, proactive monitoring and controltask description is also developed. This language can be used for dialog interaction, calculationplanning and data mining too. The proposed method main advantage is the non-isolated, but integratedsolving of robotics configuration (reconfiguration) modeling, planning and management withthe structural dynamics proactive control common problem solution.

3D Point Clouds Data Super Resolution-Aided LiDAR Odometry for Vehicular Positioning in Urban Canyons

LiDAR odometry algorithms are extensively studied for vehicular positioning. However, achieving high-precision positioning using low-cost 16-channel LiDAR in urban canyons remains a challenging problem due to the limited point cloud density from low-cost LiDAR and excessive dynamic surrounding objects. To fill this gap, this paper proposes enriching sparse 3D point clouds to denser clouds via a novel deep learning-based superresolution (SR) algorithm before its utilization in 3D LiDAR odometry. We validate the effectiveness of the proposed method using the KITTI dataset and a challenging dataset collected in an urban canyon (with complex environmental structures and dynamic objects) of Hong Kong. We conclude that significantly denser point clouds are achieved with considerable accuracy. In addition, significantly improved performance of 3D LiDAR odometry is obtained in the evaluated dataset collected in an urban canyon of Hong Kong.

Information technologies in the management of technical systems - development of the engineering education

The current level of production development requires the training of qualified specialists capable of solving complex scientific and technical problems in the field of technical systems management using rapidly changing information technologies. Consequently, the training of professional personnel should provide for the needs of developers of advanced control systems for complex dynamic objects. This problem is successfully solved at the Department of Automation and Control Processes (ACP department) in ETU “LETI”. Students training is carried out in two complementary directions “Management in technical systems” and “Information systems and technologies”. Long-term experience proves the advisability of choosing such an approach in the context of rapidly developing technologies. The report contains studies based on the application of data analysis methods and models that dynamically integrate the information necessary for adaptive management of the process of advanced specialists training. The dataset includes: actual employers’ data, production needs, on the one hand, data obtained during the educational process, on the other hand. At the same time, the data set includes data that enable students to assess the effectiveness of the educational services provision, the level of professional training in the areas of specialty. Comprehensive analysis of data integrated in a unified environment, based on the methods of intelligent systems, analytics and knowledge engineering, proves the validity of choosing this approach in the learning process.

Refined intelligence theory: artificial intelligence regarding complex dynamic objects

A new round of scientific and technological change and industrial transformation is emerging. Based on deep learning and big data, a third achievement of artificial intelligence is represented by AlphaGo and other typical application scenarios. When dealing with complex dynamic objects, the traditional artificial intelligence based on statistical linear dynamic modeling experiences bottlenecks related to interpretability, generalization, and reproducibility. It is urgent that a new generation of artificial intelligence theory be established that is based on complexity and multi-scale analysis, which we refer to as refined intelligence. To deal with the non-linearity of complex systems, we constructed a systems learning theory of embedded domain knowledge and a mathematical physical mechanism that is accurate and intelligent at three performance levels: complex data perception, complex system refined construction, and complex intelligent behavior analysis. Specifically, we built a scientific data system with interpretability, including embedded spatiotemporal characteristics and mathematical laws, through complex data perception. We also built a multi-level, multi-scale, and interpretable artificial intelligence dynamic learning model that can deal with the nonlinear relationship of complex logic based on complex system-defined construction. We developed a new theory and method for interpreting and controlling the evolution of artificial-intelligence-oriented behavior and global dynamic analysis based on complex intelligent-behavior analysis. We applied the above refined intelligence theory to the proposed method for crowd intelligence with crowd entropy, and found that it can measure complexity and provide effective guidance regarding system stimulation and aggregation behavior.

Design of Nonlinear Control of Gas Turbine Engine Based on Constant Eigenvectors

A gas turbine engine represents a complex dynamic control object. Its characteristics change depending on the state of the environment and the regimes of its operation. This paper discusses an algorithmic approach to the design of a nonlinear controller, based on the concept of constant eigenvectors and analytical design of the control system. The proposed design method makes it possible to ensure the stability and the required quality of transient processes at different acceleration modes. In this case, the constancy of the matrix of the canonical basis of the closed-loop control system is assumed, which guarantees stability. The design of a neural network dynamic model of a gas turbine engine based on a neural network approximator with one input and multiple outputs is considered. An example of the design of a nonlinear controller for a gas turbine engine is considered, the neural network model of which is given in the state space. The application of neural network approximation of controller coefficients is presented.