Abstract
Contents
- Introduction
- 1. Theme urgency
- 2. Goal and tasks of the research
- 3. A review of research and development
- 4. Development of mathematical model
- 5. Methods and models for solving the problem
- Conclusion
- References
Introduction
Locomotive — propelled railway vehicles designed for non-powered cars traction. Locomotive crew is a group of workers on the railway, which will undertake to service the locomotive, train safe conduct, implementation timetable, cost-effective use of energy resources.
1. Theme urgency
The problem of workplace automation locomotive crew is relevant because its solution will improve the safety of train conducting composition, minimize the cost of fuel and energy resources, minimize the deviation from the timetable.
2. Goal and tasks of the research
The aim of master's thesis is creating a set of management practices locomotive using fuzzy logic rules that will automate a number of decisions on management of the locomotive.
Decision support device will assess the situation and make recommendations for action to control the locomotive. The final decision will take the locomotive crew.
To achieve this goal it is necessary to perform the following tasks:
- Domain analysis, existing methods and models of decision-making;
- Preparation using the algorithm on expert assessments driver in different situations;
- Development of a fuzzy knowledge base;
- Developing an algorithm fuzzy inference;
- Checking the accuracy and determination its effectiveness of the model.
The object of research is the process of locomotive management.
The subject of research is the use of fuzzy logic rules in the management of the locomotive
3. A review of research and development
In general, the problems of automation control of locomotive considered in Nikiforov B., Golovin V. Kutyeva Y. [7], as well as in P Larionov P., Belyy D.[10]. Regulations control the locomotive is considered in the source [8].
Decision support system for managing transport are discussed in the following papers.
Sherstuk V., Ben A. — Hybrid intellectual decision support system for ship management [12]. The paper presents the structure and operation of intellectual decision support system for ship management, which is based on precedent decision models. Intellectual decision support system is a hybrid intelligent system having subsystem based reasoning precedents, rules-based and model-based.
Kotov O. — Automated multi-purpose locomotive control system [11]. This paper considers the multi-functional automated system management and security of locomotives, developed by the Research Institute of locomotives and track machines.
Decision support systems, which are based on fuzzy logic device, described in the following papers.
Polkovnikova N., Kureichik V. — Development of expert system model based on fuzzy logic [13]. In this paper considers a model of an expert system for fault identification of dynamic object in the field.
Krieger L. — Intellectual Decision Support System in the management of public transport facilities [14]. The paper presents an intellectual decision support system in the management of public transport facilities, built on the basis of fuzzy situational modeling.
4. Development of mathematical model
It's considered the problem decisionmaking for the locomotive crew
– a signal received from the i-th situation;
– many additional signals measured at the i-th situation and necessary for decision making;
f – function that takes the decision;
– a plurality of output signals (solution offered by the system).
One of the typical situations is getting red light (). Events () will be the current speed of the train, the train brake pipe condition, the distance to the traffic lights, rails condition, characteristic of trains, etc. – decision to transfer the control arms in a different position, more optimal for the given situation.
System schematically shown in Figure 1.
Figure 1 – Schematic representation of the system
(Animation: 3 frames, 11 kilobytes)
5. Methods and models for solving the problem
To enable the intelligent fleet management plays a major role knowledge base. Knowledge representation is a formalization and structuring of knowledge, with which it shows the major characteristic features [4]: internal interpretability, structuring, connectivity, semantic metric, activity.
To formalize and knowledge representation in memory of information systems there are a number of models that can be structured as follows:
- logic models;
- production models;
- network models;
- frame-based models;
- mathematical models;
- model with fuzzy inference rules.
The basis of logical models of knowledge representation based on the notion of a formal system as a quartet:
T – set of basic symbols theory of M (eg, letters of the alphabet);
P – set of syntactic rules by which characters are constructed from basic formulas;
A – the set of formulas constructed, consisting of axioms;
F – inference rules that define a set of relations between well-formed formulas.
Production model is a model that allows you to submit proposals in the form of knowledge, called productions, such as if (condition) then (action)
. Under the condition (antecedent) is understood a sentence-sample, for search of the knowledge base, and under the action (consequent) – operations performed if successful results (they may be intermediaries acting more as conditions and terminal or target, trailing the system). Most often, such a conclusion on the basis of knowledge is direct (data from the search target) or reverse (for the purpose of its confirmation – data).
Semantic network is a directed graph whose vertices represent some of the concepts and the arcs – the relationship between them. Thus, the semantic network reflects the semantics of the domain in the form of concepts and relations [6].
Frame model is a systematic psychological model of human memory and consciousness. A frame is a data structure for representing a stereotyped situation. With each frame associated information of different kinds [4].
Approach using fuzzy inference involves the use of expert knowledge about the object to be submitted in the form of rules expressed in natural language. In describing the object used linguistic variables that define the state of the object.
Linguistic variable is defined by a tuple
– name of the linguistic variable, reflecting an object or parameter study domain; T – set of its values or terms representing names of fuzzy variables, the domain of each of which is a set of U; G – syntactic procedure describing the process of the formation of the new set T, meaningful for this decision of linguistic variables; M – semantic procedure allowing each attribute to the new value of that produced by the procedure G, some semantics by forming the corresponding fuzzy set, ie display the new value in the fuzzy variable [5].
Further formalization procedures aimed at getting fuzzy sets that determine the parameters of the control object. Further calculation is made with the use of operations on fuzzy sets (AND
, OR
, NOT
), as well as the operations of taking minimum, maximum. The last step is the inverse transformation management, resulting in a fuzzy set in the real value of the output. Generic algorithms fuzzy inference algorithms are Mamdani and Sugeno.
Conclusion
Master's thesis is devoted to the actual problem automation of workplace locomotive crew.
The studies analyzed the locomotive control automation techniques, a review of existing decision support systems for transport management, a review of existing models of data, and selected model based on fuzzy production rules. This model makes it possible to implement the system of intellectual functions, based on the analysis of incomplete information about the domain, moreover, thanks to the continuity of membership function appear advantages in processing speed.
In the future will develop a decision support system for managing a locomotive. It will provide security reference traction rolling stock, will minimizes consumption of fuel and energy resources in the area path, will minimizes the deviation from the timetable.
At the time of writing this essay master's thesis is not yet complete. Final completion: December 2014. Full text of the work and materials on the topic will be available from the author or his manager after that date.
References
- Дробинский В.А. П.М. Егунов Как устроен и работает тепловоз – М.: Транспорт, 1980. – 370 с.
- Автоматическая локомотивная сигнализация частотного типа усиленной помехоустойчивости АЛС-ЕН/В. М. Лесяков, Д. В. Шалягин і ін. – М.: Транспорт, 1990. – 176 с.
- Фреймовая модель представления знаний // Дьяконов В. П., Борисов А. В. Основы искусственного интеллекта. Смоленск, 2007.
- Олкконен Е. А. Модели представления знаний в языковых интеллектуальных обучающих системах // Прикладная математика и информатика : труды Петрозаводского государственного университета. 1997. № 6. С. 168–182.
- Обработка нечеткой информации в системах принятия решений / А.Н. Борисов, А.В. Алексеев, Г.В. Меркурьев и др. – М.: Радио и связь, 1989. – 394 с.
- Гаврилова Т.А., Хорошевский В.Ф. Базы знаний интеллектуальных систем. Учебник. СПб.: Питер, 2001.
- Никифоров Б. Д., Головін В. И., Кутыев Ю. Г. Автоматизація керування рухом поїздів. – М.: Транспорт, 1985. – 324 с.
- Керування поїздом: рекомендації машиністові локомотива й диспетчерові по забезпеченню безпеки руху. – Ростову-на-Дону: РИИЖТ, 1998. – 184 с.
- Лисенков В.М. Безпека технічних засобів у системах керування рухом поїздів. – М.: Транспорт, 1992. – 192 с.
- Ларионов П.С., Білий О.К. Автоматизована система для ведення рухомого состава//Локомотив. 1997. №7.
- Котов О. Автоматизированная многофункциональная система управления локомотивом, СТА, 1998.
- Шерстюк В.Г., Бень А. П. Гибридная интеллектуальная СППР для управления судном, Искусственный интеллект, Херсонский государственный морской институт, 2008.
- Полковникова Н. А., Курейчик В. М. Разработка модели экспертной системы на основе нечеткой логики, Известия ЮФУ Технические науки, 2014.
- Кригер Л. С. Интеллектуальная система поддержки принятия решений при управлении движением общественного транспорта, Вестник Астраханского государственного технического университета, 2012.
- Заде Л. А. Понятие лингвистической переменной и его применение к принятию приемлимых решений: Пер. с англ. – М.: Мир, 1976.
- Леоненков А. В. Нечеткое моделирование в среде Matlab и fuzzyTech – СПб.: БХВ-Петербург, 2005 – 736 с.
- Fuzzy Logic: Четкие решения нечеткой логики [Электронный ресурс]. – Режим доступа: Reference.
- Нечеткая логика в системах управления [Электронный ресурс]. – Режим доступа: Reference.
- Решение задач с применением нечеткой логики [Электронный ресурс]. – Режим доступа: Reference.
- Энциклопедия ЖД [Электронный ресурс]. – Режим доступа: Reference.