Vetronics Institute 2^nd Annual Workshop Series             

Links to Abstracts and Presentations

Automotive Fault Diagnosis: Distributed Agent Diagnostic System

Yi Lu Murphey

Department of Electrical and Computer Engineering

The University of Michigan-Dearborn

Dearborn, MI 48128-1491

yilu@umich.edu

313-593-5420(phone)

313-593-9967(fax)

Abstract

In this workshop, we describe a novel diagnostic architecture, DDAS(Distributed Diagnostics Agent System) developed for vehicle fault diagnosis. The DDAS consists of signal diagnostic agents, and vehicle fault agents. Each signal diagnostic agent can be trained to be used for OBD(On board Diagnostics) and responsible for the fault diagnosis of one particular signal using either a single signal or multiple signals depending on the complexity of signal faults. Each fault diagnostic agent can be trained to detect any particular type of vehicle faults such as vacuum leak, injector fail, idlerolling, etc. Each agent is developed using a common framework that involves signal segmentation, automatic signal feature extraction and selection, machine learning and case based reasoning(CBR). All the diagnostic agents can concurrently execute their tasks, some agents possess information concerning the cause of faults for other agents, while other agents merely report symptoms. Together these signal agents present a full picture of the behavior of the vehicle under diagnosis to the vehicle diagnostic agent. DDAS provides three levels of diagnostics decisions: signal segment fault, signal fault and vehicle fault. DDAS is scalable and versatile. DDAS has been implemented for fault detection of Electronic Control Unit(ECU) signals and experiment results will be presented and discussed in the workshop.

Return to Titles

ON-BOARD DIAGNOSTIC II (OBD II) FOR LIGHT MEDIUM DUTY VEHICLES

 Subra Ganesan, Professor,

Department of Computer Science and Engineering,

Oakland University, Rochester, MI 48309

Phone: (248) 370 2206

Fax: (248) 370 4625

Email: Ganesan@oakland.edu

Web: www.secs.oakland.edu/~ganesan

Abstract:

Environmental Protection Agency (EPA) has mandated emission standards. Emission related malfunctions must be diagnosed and repaired not only to reduce emissions but to eliminate the risk of further damage to the vehicle. Computers and sensors in the vehicle must perform diagnostics and operational adjustments to maximize combustion and reduce emissions. Diagnostics of failed component and collecting freeze –frame-data for intermittent failures is very much needed for military vehicle repair and maintenance.  This tutorial provides basic concepts of OBD requirement, principles of the major automotive components related to OBD, brief introduction to various sensors, actuators needed for OBD, advanced developments, references and military applications. Research areas in the above topics and military applications will be emphasized.

Return to Titles

DSP in EMBEDDED SYSTEM

Subra Ganesan, Professor,

Department of Computer Science and Engineering,

Oakland University, Rochester, MI 48309

Phone: (248) 370 2206

Fax: (248) 370 4625

Email: Ganesan@oakland.edu

Web: www.secs.oakland.edu/~ganesan

Abstract:

This TUTORIAL emphasizes design of embedded systems using Digital Signal Processing microprocessors, and special DSP FPGA chips. Topics covered include, DSP microprocessor architecture, advanced instructions, addressing modes, interrupt, system design considerations, interfacing serial and parallel I/O, memory structure, arithmetic manipulations, software development tools, multiple DSP processor system design, and embedded system applications. Applications include automotive, multimedia, and wireless communications. Performance measurement, benchmarking and DSP system simulation, testing and debugging.  Research areas in the above topics and military applications will be emphasized.

Return to Titles

Tutorial on Cultural Algorithms

Wayne State University

Cultural Algorithms are computational models of Cultural Evolution. They consist of two basic components, a population space, and a belief space. The two components interact by means of a Vote-Inherit-Promote or VIP protocol. A variety of paradigms have been used to model the population component including Genetic Algorithms, Genetic Programming, Evolution Strategies, and Evolutionary Programming, Cellular Automata, and Multi-agent systems among others. Likewise the knowledge acquired by the problem solving activities of the population can be stored in the belief space in the form of production rules, semantic networks, version spaces, among others.

As such, Cultural Algorithms represent a general framework for producing hybrid evolutionary systems that integrate evolutionary search and symbolic reasoning together. Cultural Algorithms are particularly useful for problems whose solution requires extensive domain knowledge. The tutorial starts by providing a brief motivation for the Cultural Algorithm framework. Next the basic choices for configuring Cultural Algorithms are presented and the motivation for selected particular configurations will be discussed. The tutorial will conclude with example applications from a variety of application areas including function optimization, knowledge based system design and maintenance, software engineering, and the modeling of complex social systems.

Return to Titles 

Electric and Hybrid Vehicles for Military Applications

Prof. Chris Mi

University of Michigan - Dearborn

Department of Electrical and Computer Engineering

4901 Evergreen Road, Dearborn, MI 48128

E-mail: chrismi@umich.edu

Tel: (313) 583-6434

Fax: (313) 593-9967

Abstract

The material of this workshop comes from the instructor’s teaching experience of a graduate course “Electric Vehicles” at the University of Michigan - Dearborn.  In this 4 hour workshop, we will discuss both electric and hybrid electric vehicles suitable for military applications. We will start the session with system engineering and impact of electric and hybrid vehicles.  Simulation tools such as Advisor, dSPACE, and Matlab/Simulink will be introduced. We will investigate the principle, selection, and design issues of major components needed for an electric vehicle, such as motors, batteries, and controllers. This workshop is intended for electrical, mechanical, vehicle manufacturing and automotive system engineers who would like to extend their knowledge in electric and hybrid vehicle systems and components.

Return to Titles

Vetronics In European AFVs and An Assessment Tool

Mr Stuart Dowling

Royal Military College Of Science

Engineering Systems Department

Shrivenham SN6 8LA, England

s.dowling@rmcs.cranfield.ac.uk

Tel. 44 (0)1793 785629 or 785475

Fax. 44 (0)1793 783192

Abstract

The European and more particularly the UK view of current vetronics developments are reviewed. A review of current UK and European vehicle programs is used to set the scene. CANBUS and other COTS data bus and network standards are likely to find their way into European AFV designs from the experience gained in high end automobiles such as the Mercedes S Class and the Citroen C5. Currently TTP and DSI are being investigated as potential COTS transfer technologies.  The MRAV, FRES, FCLV and Bowman programmes have highlighted the need for a way to assess vetronics architecture and fit. The tool to do this is being developed by the author as PhD study. The methodology to do this, is one of systems integration versus impact on performance matching. The project is at an early stage where the vetronics building blocks and architecture choices are being reviewed and evaluated. Data bus, network, power distribution, video distribution and computer system architectures are all being fed into a matrix. The human factors and concepts such as crew station design are also being evaluated. The hoped for result is a tool which can score the vehicle as a system without using opinion based scoring. The ultimate aim is to link this tool to some costing data and produce a cost / effectiveness evaluation of potential vehicle vetronic fits.

 Return to Titles

Intelligent Agent Architecture and Applications

Gautam B. Singh, Ph.D.

Associate Professor

Department Computer Science & Engineering

Oakland University, Rochester, MI 48309

Phone: (248) 370-2129, Fax: (248) 370-4625

E-Mail: singh@oakland.edu

This workshop aims at providing a hands-on experience for developing intelligent agents. Principles of agent knowledge representation, agent architectures, agent communication languages and frameworks, machine learning in agents are amongst the topics covered. Theoretical ideas are presented with supporting applications and examples. The workshop will be presented in the following four modules.

Module 1: Knowledge Representation in Agents:

Module 2: Multi-agent systems and society of agents:

Module 3: Distributed decision making:

Module 4: Learning in multi-agent systems

Return to Titles

Introduction to Physical System Modeling Using Bond Graphs

Taehyun Shim, Assistant Professor

Mechanical Engineering Department

The University of Michigan-Dearborn

Tel. 313-593-5127

Email: tshim@umich.edu

With a fast growing computing power, the mathematical model has been more extensively used in the development of dynamic system from concept to commercialization in recent years. The characteristic features of these models must reflect some of important features of the real systems. It is impossible to predict the realistic behavior of the real system if the model is too simple. On the other hand, it is very difficult to analyze the individual effect if the model is too complex such as it contains all the detailed components of the model.  The development of the mathematical model is not trivial especially when the model is complex and coupled with different energy domains (electro-mechanical, hydraumechanical,and pneumaic-mechanicsl, etc.). The bond graphs are a domain-independent and a concise pictorial representation of dynamic behavior of physical system. This means that systems from different domains are described in the same way so that it is a powerful tool for modeling engineering system with different energy domains. This workshop will discuss the basic theory of the bond graphs and some of its application in various engineering systems.

Return to Titles

Advanced Vehicle Systems Research Program

Department of Electrical Engineering

Texas A&M University

College Station, Texas      77843

Phone: 979-845-7582

Email: ehsani@ee.tamu.edu

Military and civilian vehicles are moving towards more electrification, in response to the increasing demands for multi-function missions, fuel consumption and emissions reduction, and dual use electrical and electronic components. Consequently, the vehicle electric load is increasing rapidly. For military vehicles, these electrical loads include the loads for cabin climate conditioning, vehicle control and actuation, actuation by wire (X by wire), sensors, reconnaissance, communications, weapons, and electric traction. The electric power system consists mainly of power generation, energy storage and power management and distribution, in a complex system. Satisfactory development of this system, to meet the performance and efficiency requirements, demands a great effort on different levels. These are methodical design of the system architecture, proper selection of components (power generating, energy storage, power management devices, etc.) and power management strategies.

This short course will introduce the above issues and consists of the following topics:  

1. Electric Power and Energy Requirement Analysis

2. Vehicle Electric Power System Architecture

3. Electric Power Management.

4. Available and Potentially Available Energy and Power Sources

5. Electric Drives

Return to Titles

Reconfigurable Computing

Dinesh Bhatia

Department of Electrical Engineering

Erik Jonsson School of Engineering and Computer Science

The University of Texas at Dallas

PO Box 0688

Richardson TX 75083

dinesh@utdallas.edu

Abstract

Reconfigurable computing can be thought of as an ability to repeatedly configure a machine to perform different and varying functions. The term reconfigurable is broad and can be applied to many scenarios. Custom crafting of computing functions in hardware greatly speeds up the execution. However, enhanced performance does not come without the loss of flexibility and inexpensively. Typically, a custom computing machine is desirable for performance driven applications for which generic computing platforms may not be sufficient for quick execution. Reconfigurable computing offers high performance, inexpensive and highly flexible method of executing information processing applications. A reconfigurable custom computing machine makes use of some form of reconfigurable logic that can be changed and configured as demanded by an application. Reconfigurable logic/hardware can exist as statically reconfigurable where final configuration is pre-decided and once the machine has been configured, it maintains the same configuration for the entire life of the application. Field Programmable Gate Arrays (FPGAs) that rely on SRAM based (or its variants) programming technologies have a capability of being configured/ reconfigured "on the fly". This opens a new avenue that allows hardware to transmogrify from one form to another (and in-circuit!). Hardware with such a capability is called as dynamically reconfigurable logic. The presentations in the workshop will be conducted such that, both, novice engineers and researchers as well as experts in the area can benefit the most. It will be an interactive lecture driven format. The workshop will begin with a brief introduction to reconfigurable computing, clearly highlighting the differences offered by a more general microprocessor based platforms. It will also provide in-depth analysis of state of the art technologies that facilitate reconfigurable computing. In order to make reconfigurable commercially viable and useful for defense applications, it is essential to have proper software environments for mapping, programming, and implementing user applications. A significant portion of this workshop will focus on this aspect of design. This will also provide deeper insight into the class of applications that can benefit the most by reconfigurable computing. We will also analyze characteristics of applications that are particularly suitable for reconfigurable computing, and bottlenecks and shortcomings of reconfigurable computing.

Return to Titles

A Tutorial On Condition Monitoring and Fault Diagnosis of Electric Machines and Fault Tolerant Electric Machinery

Hamid A. Toliyat

Texas A&M University

College Station, TX 77843-3128

Advanced Electric Machines & Power Electronics Laboratory

Department of Electrical Engineering

Texas A&M University

College Station, TX 77843-3128

Tel: (979) 862-3034

Fax: (979) 845-6259

E-mail: Toliyat@ee.tamu.edu

A major goal of the TARDEC is research toward combat readiness and future war fighting superiority of army forces. The proposed tutorial is geared toward a highly reliable design and implementation of fully electric/hybrid traction systems for future tanks combatants. The proposed fault tolerant variable speed motor drives are not just limited to the traction systems, but they can also be applied to the auxiliary systems, such as power steering, air conditioning, etc. The proposed tutorial consists of two half-day topics:

(1) Condition Monitoring and Fault Diagnosis of Electric Machines, and

(2) Design of Fault Tolerant Electric Motor and Generators for Variable Speed Drive Applications.

Return to Titles

Vision-based automatic driving of road vehicles: the experience of the ARGO and RAS projects

Prof. Alberto Broggi

Dip. di Ingegneria dell'Informazione,

Universit_a di Parma

Parco Area delle Scienze 181/A,

I-43100 Parma, Italy

Phone: +39-0521-905707

Fax: +39-0521-905723

E-Mail: broggi@ce.unipr.it { Web: www.ce.unipr.it/broggi

Abstract

The presentation will focus on vision-based (daylight and infrared) techniques for the perception of outdoor environments aimed at recovering 3D information for the automatic driving of autonomous vehicles.  Among the research activities ongoing at the Dipartimento di Ingegneria dell'Informazione (University of Parma, Italy), the ARGO Project will be presented, together with the current cooperation with TACOM, and some results of technology transfer in extreme environments. The ARGO project is aimed at developing a prototype of an autonomous vehicle able to automatically drive in standard roads. A vehicle, equipped with sensing, processing, and actuating capabilities, was tested it in real road, weather, environment, and tra_c conditions with a 2000+ km automatically-driven tour in Italy. The presentation will describe the project and its main results, focusing on the main functionalities of the system: lane detection, obstacle detection, vehicle detection, as well as pedestrian detection, which is the theme of our current joint research with TACOM.

The presentation will also discuss the results obtained with the application of these techniques to the automatic driving of tracked vehicles in extreme environments (South Pole). The RAS (Surface Antartic Robot) vehicle is, in fact, equipped with a track detection software for the localization of tracks left on ice/snow by ahead vehicles; this functionality was tested in the last Italian scienti_c mission to Antarctica in early 2002. The presentation will be followed by a live demonstration of the software. Prerecorded image sequences will be processed in real-time by a lap-top PC; a user interface will allow to change processing parameters on-the-y, making it possible to alter the algorithms' sensitivity and experience the system behavior even in extreme environments.

Return to Titles

Universal Micro System: Reconfigurable, Real-Time, and High Performance System-on-chip Solution for Embedded Space

Vipin Chaudhary

Director, Institute for Scientific Computing

Wayne State University, Detroit, MI 48202

Phone: 313 577 0605; Fax: 313 578 5839

Email: vipin@wayne.edu

Abstract

Increasingly, today’s embedded and system-level chips are multiprocessors (Texas Instruments, Motorola, IBM, Infineon, NEC, Toshiba, Philips, etc.). Moreover, these multiprocessors are usually cores of RISC (Reduced Instruction Set Architecture) and DSP (Digital Signal Processing) engines. The Universal Micro System (UMS) from Cradle Technologies, Inc. is a single-chip computing platform that combines the classic microprocessor, micro-controller, and digital signal processor chip families. UMS permits a broad range of streaming applications (like high-speed secure networking, imaging, graphics, multimedia, digital video & communications) to be created entirely in software. This permits tremendous design economies through simplified design, software re-use and support from third-party software-solution providers. In addition, the UMS architecture yields immediate 1-to-2 orders of magnitude increase in performance, and a similar reduction in cost, and design effort over existing approaches. Another unique aspect of UMS is its programmable I/O that enables any I/O to be done entirely in software. For the first time, software development methodologies can be utilized and still guarantee real time performance. Solutions can now be realized much faster because intellectual property will be developed, reused, and recombined, much more efficiently and cost effectively. In the short term, this reduces time to market significantly, and solves many of the integration problems that exist today. In the long run, multiple products and product lines can be developed quickly and at a fraction of the cost. In a nutshell, UMS can be looked as a reconfigurable processor that is cheaper, higher performing, and easier to use than FPGAs.  In this workshop, we will cover the architecture, programmability, and tools for UMS. A cycle accurate simulator will be available for every Windows workstation for hands-on interactive experience and exercises with UMS. A UMS development board will also be used for highlighting the power of UMS.

Return to Titles

FAULT RESILIENT ELECTROMECHANICAL ENERGY CONVERSION DEVICES FOR DRIVE BY WIRE APPLICATIONS

Dr. Steven D. Pekarek & Dr. Babak Fahimi

University of Missouri- Rolla

Department of Electrical & Computer Engineering

1870 Miner Circle

Rolla, MO 65409-040

Tel: 573-341-4543

Fax: 573-341-6671

Abstract

“Drive by wire” is considered a major step towards successful implementation of an unmanned vehicle. In addition, it promises to improve the performance and efficiency of automobiles. Fulfillment of these objectives depends upon proper operation in various parts of the electromechanical energy conversion devices; namely, control electronics, sensors, power electronics drivers, and electric machines. Specifically, these components can fail. When they do, the control system must be able to compensate for the failure and continue to operate, to the best extent possible. In the development of a motor controller, one of the first design decisions is focused upon choosing between various control alternatives and deciding upon which one to implement subject to the requirement of a particular application. The approach taken in the case of “Drive by wire” applications differs significantly in that it treats controller performance together with controller fault tolerance. Rather than developing a nominal controller based upon a single control technique, the system needs to adaptively change its configuration as well as its control strategy in the event of a failure in order to attain the best performance given the complement of remaining hardware. The proposed seminar explores various steps in the development of fault tolerant motor drive solutions. This includes a systematic review of detection and diagnosis of failures in induction and synchronous machines, power electronics inverter, and sensory systems. We will then introduce the methodologies for automatic reconfiguration of control hardware and software. This seminar is designed to provide a physically insightful version of the topic for its audience.

Return to Titles

Memory Motors – A New Class of High Torque Density, Extremely Wide Speed Range and High Efficiency Electric Machines for Application as Engine Starter - Power Source in Military Vehicles

Vlado Ostovic

HTW - Hochschule fuer Technik und Wirtschaft des Saarlandes

Goebenstr. 40

66117 Saarbruecken

GERMANY

e-mail: Ostovic@htw.uni-sb.de

Phone: 011-49-6201-509300

Fax: 011-49-6201-507715

Abstract

By its power rating and efficiency impact, the integrated starter/generator (ISG) is among the most important of all electric auxiliary drives in next generation vehicles. A starter/generator replaces both the DC starter and the AC alternator. Yet, at present, there is no solution with adequate performance for both functions. A novel high torque density and high efficiency class of electric machines for ISG combined drive in military vehicles will be presented in this workshop. The proposed approach is based upon a new type of electric machines [patent pending] called memory motors due to their ability to vary the level of magnetization and save its last state in rotor magnets. A memory motor can be built either as a pole-changing permanent magnet (PCPM), or a variable flux permanent magnet (VFPM) machine. In both cases the magnetization of magnets can be simply varied by a short current pulse, without need for permanent demagnetizing current as is the case in internal permanent magnet (IPM) machines with flux weakening. The demagnetizing current flows through stator winding(s), and is superimposed to the stator phase current(s). A pole- changing memory motor is the only AC machine besides squirrel cage induction machine that can have two or more synchronous speeds when fed from a constant frequency source. A variable flux memory motor incorporates low losses of a conventional permanent magnet machine together with flux controllability of a wound rotor synchronous machine, resulting in a high starting torque, extremely wide speed range operating machine with exceptional efficiency at all modes of operation. Both machine types utilize AlNiCo magnets which have the best thermal properties among all types of permanent magnets (residual flux density change as low as 2% per 100o C, coercive force change as low as 2% per 100o C, operating temperature up to 550o C) The objective of the workshop is to illustrate the feasibility of the two new permanent magnet machines as ISG in military vehicles. It will be first demonstrated that the two new machines meet all the ISG requirements, especially extreme temperature stability, high starting torque and constant power combined with high efficiency at exceptionally wide speed range. After that, the more suitable machine for the application will be selected, based on technical parameters and other relevant criteria.  Finally, some other interesting applications of memory motors, such as propulsion machine, alternator in gensets etc. will be discussed.

Return to Titles

Navigation methods, techniques, sensors and applications in mobility

G. E. Smid, Ph.D, Prof. Ka C. Cheok,

128 SEB, SECS, Oakland University

tel. (248) 370-2082/2232, fax. (248) 370-4633

smid@oakland.edu

The 21^st century tactical employment of strategic and logistic vehicles is inherently related to tight mission planning and information processing technology. The methods and techniques, however, to exploit the available information and to enhance the usability of the available tools, are a significant and critical part of the implementation of a navigation system, and often determine the usability and success of a system in the field.

This workshop will outline the methods and techniques – both hardware and software – that are required to understand and architect an effective navigation system for military purposes. During the workshop there will be numerous design examples from commercial applications and simulation demonstrations in Matlab/Simulink.

The term Navigation implicates a number of tasks related to tracking, positioning, mobility, path-planning, and recognizance. The relative emphasis on each of these tasks is closely related to the application area of the system. Autonomous robotic vehicles require positioning and obstacle avoidance, but does not necessarily need to interface its navigation information with the operator. A human-assist system does not need the strict robustness of an autonomous vehicle, but requires a fluent Human-Machine Interface (HMI) for its operation to be effective.

The mathematical techniques for positioning and tracking vary in relation to the sensors and the type of information from the sensors that is to be used. Most designs or navigation systems will include multiple sensors, which requires additional mathematical techniques to incorporate the additional streams of information, and to resolve conflict. The methods to simulate and implement these techniques will be discussed in the workshop.

The workshop will provide the necessary background to understand the design considerations, technical issues on sensor hardware and mathematical techniques, as well as the differences in critical requirements for human-assist navigation systems and navigation systems for autonomous vehicles. This background will provide a solid basis of knowledge and information to evaluate proposals related to navigation technologies, to participate in design review meetings and to manage navigation related projects in a program with confidence.

Return to Titles

Tranceiver Adaptation in UWB Wireless Systems

Dr. M. Saquib

Department of Electrical and Computer Engineering

University of Texas at Dallas

Abstract

Ultra-wideband (UWB) technology promises to deliver large amounts of data with very low power spectral density.  The UWB radio concept is very attractive as is seeks to open large amounts of spectrum to a variety of uses and at the same time it claims little interference between users. Unlike conventional wireless communications systems that are carrier-based, UWB-based communications is baseband.  It uses a series of short pulses that spread the signal energy over the spectrum.  This approach produces a signal that is more covert, has higher immunity to interference effects, and has improved time of arrival resolution.  At the same time, it offers challenges to design of low complexity receivers which perform RAKE combining and synchronization efficiently.  In the first part of the workshop those issues will be addressed thoroughly.  Besides these, other issues such as ways of improving the system throughput, better understanding of propagation channel will be discussed.

In the second part of the workshop, UWB radio concept will be discussed in light of Direct Sequence Spread Spectrum technology.  It will be shown how one can minimize the total transmitted power of the system subject to signal-to-interference ratio (SIR) based on quality of service (QoS) requirements to maximize the system capacity.  Specific scenarios will include ‘one transmitter to many receivers’, ‘many transmitters to one receiver’, and ‘many transmitters to many receivers’. 

Soft packet length concept for UWB systems will be introduced to ameliorate the throughput (frames/second). The length of the frame will depend on the duration of its transmission time rather than the number of bits in it.

Return to Titles