Automatic modeling and fault diagnosis of timed concurrent discrete event systems / Stefan Schneider.

Format:

Book

Author/Creator:

Schneider, Stefan, author.

Language:

English

Subjects (All):

Discrete-time systems.

Physical Description:

1 online resource (184 pages)

Edition:

1st ed.

Place of Publication:

Berlin : Logos, [2015]

Summary:

Long description: The productive operation of machines and facilities is of great economic importance for industrial companies. In order to achieve high productivity, unscheduled production downtimes induced by faults need to be minimized. In this work, an approach for modelbased fault diagnosis of timed concurrent Discrete Event Systems is proposed that can contribute to this aim. The models are automatically determined by timed identification and partitioning. These approaches allow for efficient modeling of large and complex industrial systems with concurrent behavior requiring only little system knowledge. The work explains the theoretical and practical aspects of the presented approaches and gives a detailed evaluation based on a laboratory manufacturing system.

Contents:

Intro

1 Introduction

1.1 Motivation

1.2 Discrete Event Systems (DES)

1.3 Bosch Mechatronics System (BMS)

1.4 Contribution

1.5 Organization

2 Automatic Modeling and Fault Diagnosis Challenges

2.1 Fault Diagnosis of DES

2.1.1 Terms

2.1.2 Concepts

2.2 Model-based Fault Diagnosis Challenges

2.3 Modeling Challenges

3 Identification of Timed Distributed DES Models

3.1 Preliminaries

3.2 Timed Modeling

3.2.1 Internal and External Behavior

3.2.2 Timed Model and Languages

3.3 Identification of Timed Models

3.3.1 Time Identification Approach

3.3.2 Timed Identification Algorithm

3.3.3 Precision and Completeness Properties

3.3.4 Identification Parameters

3.4 Timed Distributed Modeling

3.5 Identification of Timed Distributed Models

3.5.1 Timed Distributed Identification Approach

3.5.2 Precision and Completeness Properties

3.5.3 Discussion on Shared I/Os

3.6 Identification of Timed Distributed BMS Models

3.6.1 Data Collection

3.6.2 Timed Distributed Identification

4 Partitioning of DES Models

4.1 Preliminaries

4.2 Causal Partitioning

4.2.1 Distance and Causality

4.2.2 Causal Partitioning Algorithm

4.3 Optimal Partitioning

4.3.1 Optimization Approach

4.3.2 Optimal Partitioning Algorithm

4.4 Partition Synthesis

4.5 Partitioning of BMS Models

4.5.1 Causal Partitioning

4.5.2 Optimal Partitioning

4.5.3 Partition Synthesis

5 Fault Detection and Isolation using Timed Distributed DES Models

5.1 Preliminaries

5.2 Evaluation

5.3 Timed Fault Detection

5.4 Timed Fault Isolation

5.4.1 Preliminaries

5.4.2 Deadlock Behavior

5.4.3 Early and Late Behavior

5.5 Extension to Timed Distributed Models

5.5.1 Overview

5.5.2 Distributed Evaluation

5.5.3 Distributed Fault Detection and Isolation.

5.6 Fault Detection and Isolation of the BMS

5.6.1 Online Fault Diagnosis Implementation

5.6.2 Fault Scenarios

5.6.3 Model Validation

6 Related Works - Analysis and Comparison

6.1 Modeling of Timed DES

6.1.1 Language Model

6.1.2 Automaton Model

6.1.3 Petri Net Model

6.1.4 Discussion

6.2 Identification of DES Models

6.2.1 Preliminaries

6.2.2 Identification of Logical Models

6.2.3 Identification of Timed Models

6.2.4 Discussion

6.3 Automatic Modeling of Concurrent DES

6.3.1 Preliminaries

6.3.2 Probabilistic Data Mining

6.3.3 Model Optimization

6.3.4 Discussion

6.4 Model-based Fault Diagnosis of DES

6.4.1 Preliminaries

6.4.2 Fault Diagnosis using Logical Models

6.4.3 Fault Diagnosis using Timed Models

6.4.4 Discussion

7 Conclusion

7.1 Summary

7.2 Further Work

8 Kurzfassung in deutscher Sprache (extended summary in German).

Notes:

Description based on print version record.

PublicationDate: 20150530

ISBN:

3-8325-9498-1