Anatoly Levchenkov

Intelligent Electric Vehicle Motion and Crossroad Control

The purpose of this work is to organize coordination between intelligent trams and intelligent traffic lights. The main idea is to realize electric transport movement without stops excepting stops for taking passengers. The supposition is the effective decrease of electric energy charges causing with acceleration and breaking cycles. Following methods of control are proposed: intelligent agent system, and negotiations between these agents with the superagent as coordinator to solve possible conflicts. Intelligent agents are created using Web-technologies: a database and the appropriate programming languages that allow to realize negotiation easy and effectively

Algorithm for optimal energy consumption of UAV in maritime anti-collision tasks

This paper specifies the problem formulation of research on energy consumption optimization for using of robotized unmanned aerial vehicles (UAV) in transport safety tasks. UAV are proposed as an auxiliary maritime safety component, allowing visually detect maritime units, which are not connected to global Automatic Identification System (AIS). To ensure this task solution it is necessary to maximize the time in the air for UAV, therefore the minimization of energy consumption is topical. The paper provides the general formulation of maritime safety task, definition of a multi-objective target function, including energy minimization and safety criteria. A control object for energy consumption minimization is defined, an algorithm for task solution is developed and a numerical example is described.

Evolutionary algorithms and dynamic parameters for public electric transport modeling

This work is based on research in a field of intelligent agent systems, negotiation algorithm solving tasks of energy saving, optimal electric vehicle control and transport flow control in traffic jam. Main goal of research is energy saving for public electric transport. Mathematical model and evolutionary algorithm is proposed in the paper to solve multi-criteria optimization task minimizing idle time and electric energy used by public electric transport and maximize average speed of the flow in traffic jam. Paper presents a computer experiment to test proposed mathematical model and workability of evolutionary algorithm. The specific dynamic model of city transport system is created and results of evolutionary optimization are simulated.

Fuzzy-logic algorithm of UAV hardware configuration assessment for flight time and lift capacity improvements

This paper is based on authors previous research in a field of UAV optimal energy consumption [1]. The article justifies the necessity to keep track of changes of all the parameters in working quadcopters. The hardware selection algorithm and calculations connected with it will be presented as well as a model for choosing a more suitable choice will be offered.

Immune algorithm and intelligent devices for schedule overlap prevention in electric transport

This study is dedicated to optimize the electric transport traffic by adjusting timetables using evolutionary algorithms and scheduling theory. Developed multiple criteria objective functions energy consumption criterion is used for the assessment of adjusted schedules. The scheduling results of immune algorithm is compared with results of genetic algorithm by providing better energy saving effect and eliminating of schedules overlapping. This article also discusses one of the possible energy savings - proper planning of transport operations. It is considered on the railway irregular transportation operations example.

Multiple Criteria Decision Support for Heating Systems in Electric Transport

In this paper the solution to the problem of optimal control of climate parameters in public electric transport is proposed. Optimization of mechatronic system control is provided by minimization of electric energy consumption and maximization of passengers’ comfort level. We propose to solve this task using artificial intelligence and progressive multiple criteria decision making methods. The popular Nelder-Mead multiple criteria decision making method (Nelder and Mead 1965) is applied. This method makes it possible to find a minimal value for the target function. In this case there is a dependence of minimal electric energy consumption on maximal comfort level. Our modelling and investigation is based on a typical architecture of heating ventilation and air conditioning system with a traditional application of AC induction motors for driving both a compressor and a fan of the conditioner. Special interest and further development is devoted to intelligent heating systems, allowing more flexible regulation of the system’s compressor and fan operation, and, therefore, improvement of efficiency and energy saving.

Modelling of Artificial Neural Network Controller for Electric Drive in Virtual Laboratory

In this paper authors presents modelling of neural network controller to control speed of DC drive. Main tasks of this research are to use artificial neural network for speed control of DC drive in virtual laboratory.The feed-forward back-propagation neural network is used for controller. Levenberg-Marquardt back-propagation algorithm is proposed as a training method. Neural network is trained to maintain rotation speed of DC drive in defined interval.Results of modelling show the possibility to use neural network controller for speed control of DC drive.

Convolutional neural network in turn recognition tasks for electric transport safety

This study is dedicated to solve a turn recognition task by Convolutional neural network (CNN) application. System proposed in this paper is a part of the big research aimed at system, for the safe transportation process, development. Road turn is an important parameter in road circumstances analysing, collision possibility calculating and decision making for the accident prevention. System structure for turn recognition process and CNN algorithm with training is proposed in this paper. The experimental results validate the effectiveness of proposed algorithm. Experiments show that CNN correctly defines turn after training is done.

Reliability improvement of embedded train anti-collision system by multi-processor technology

The paper is based on authors previous research results and experience. This paper describes the service oriented system architecture (SoA) for train anti-collision system (TACS) using embedded multiprocessor system-on-chip (MPSoc). The improved configuration of previously developed system is proposed, necessary conditions for successful work are defined, and the advantages of using multi-processor system structure are presented.

Modelling And Simulation Of Public Transport Safety And Scheduling Algorithm.

The main objective of the transport operations is a safe transportation process with minimal energy consumption. There are various methods for gaining these tasks. This paper discusses one of the possible problem solving proper planning of public transport operations. The main goal of the research is to develop the adaptive algorithms for transport control and optimization. The main task of the target function is to minimize total downtime at intermediate stations. The specific unique Web-based computer model was developed. It uses Web database for simulation data storage and processing. Simulation results shows the workability of the developed algorithm.