Research Assistantship positions

Several research assistantships are available for M.S. and Ph.D. students in the general areas stated below:

1. Linear motor elevators
   1. Magnetic design of stator and rotor optimized for speed or lift force.
   2. Active position sensing for linear motors (Patent pending)
   3. Hybrid brake control and mechanical design of safety brake
   4. Multi stator and multi mover cooperative control of linear motors
2. Networked control systems
   1. Real-time systems
   2. Embedded systems, networked embedded systems
   3. Design of power electronics
3. Remote diagnostics for commercial vehicles

Research Hand positions

Drive and Safety Methods for New Generation Elevator Systems

Background

As modern building heights increase, transporting people efficiently to higher floors is becoming a bottleneck. To minimize the waiting time for passengers, several elevators must be installed. However, each elevator requires its own hoistway thus precious office space is lost. Therefore there is diminishing returns in higher buildings even though technology is available to construct them.

Another problem in transportation in high buildings approaching 1000m height is the increasing weight of the elevator cable, which after a certain height is not able to carry its own weight unless special materials are used.

To answer these questions, the idea of multi-car elevator systems has been considered for some time. Multi-car elevator systems have two or more elevator cars moving independently in the same hoistway. They promise extraordinary improvements in the space utilisation for large buildings. Since the classical rope and pulley structure is not viable in this scenario and because of the proposed building height, another method of moving the elevator cars must be used.

One idea is to use a linear motor located within the hoistway to move the cars. This way, each car in the hoistway can move independently from the others without interfering (except,obviously, that it cannot overtake the next car above and below), and irrespective of the building height. With today's available technology, producing such a linear motor capable of hoisting the required weight is feasible.

However, there are still many problems to be solved before multi-car elevators can be offered as a viable alternative for current elevator systems. The two most important open problems are how to prevent reliably and economically the collision of the independently moving elevator cars and to prevent them from falling freely, even in the case of failures in the power supply or the control system. Non vital problems such as group coordination of such systems have been studied extensively.

In our research, we have developed a preliminary safety system for multicar elevators driven by linear motors with the stator coils arranged along the hoistway and magnet assemblies on the elevator cars. We envision use of the same stator coils both for generating thrust, and for operating the brake system. The unique feature of this system is the use of the edge surface of the linear motor coil as the actuator of the brakes. Initial theoretical and experimental work has already been done.

This project is to be done in partnership with a Japanese vertical transportation systems company. Project will be made of several phases. The objective of the first phase is to develop and implement a safety device integrated with a linear motor, implement it (which requires a purpose built linear motor) and perform tests. The second phase is to implement the multi car elevator system in a much larger scale, capable of carrying one person. The project is planned to continue after that, with further improvements.

The project is multi disciplinary, involving the fields of electronics (digital and power), mechanics, embedded systems, control, electromagnetics and fault tolerance methods. There will be cooperation with Japanese universities and other overseas and Turkish companies.

Current Status

Work has begun in 2007 and the first project was succesfully completed. The company has now offered to fund research in areas shown as the first main item on the list above, in two projects. One project will start in April 2010, and the other will start in April 2011.

Student Profile and Openings

Students working in this project will design and build the aforementioned systems and perform tests on it. They will be integrating methodologies from the above disciplines which will give insight into many theoretical concepts from their education. Several visits to Japan for on site discussions and training are planned.

Interested students should have basic knowledge in the fields of electronics, automatic control and electrical machines. Knowledge in other areas such as embedded systems, programming and electromagnetics is a bonus.

There are openings for a total of four masters students in two projects, suppored by the foreign company. The application for the first two positions will be in December 2009, and the next two positions will be in December 2010.

Remote diagnostics and Fault Detection on Commercial Vehicles

Background

Down time ue to component faults and breakdowns are a big problem for commercial vehicles such as buses and trucks because the interrupt service and reduce profit and customer loyalty. This project, done in cooperation with a major bus manufacturer is aims to predict probable faults by using on board data acquisition and prognostic algorithms. In case a high probability fault is predicted, it informs the driver, operating company, a suitable service station by describing the nature of the fault. The same information is used also by the manufacturer research and deveopment department.

The project has two main components, on-board system responsible for data acquisition, analysis and prognostics, and transmission to the data processing center over cellular network (GPRS, EDGE or 3G). Analysis of the problem will be followed by implementation on a real bus, including monitoring of the communication bus. The student who is interested in working on this part of the work should have knowledge in embedded systems, data communication and networks. Some knowledge in fault detection is a plus.

The second component is the data processing center which is in constant contact with the buses in service logging incoming information. The information must be processed and stored, and at the same time madde available to qualified people over a secure web interface.

This project is interesting since it is directly proposed by a major company for doing research. The students will have a chance to get direct insight into how industrial projects are managed.

The two masters degree students needed for this project will be supported by Sabanci University(SU) and Ministry of Industry and Commerce (Sanayi ve Ticaret Bakanligi). Where SU will provide tuition vaiver in return for 10hours/week teaching assistantship, and Ministry will provide scholarship equivalent to TUBITAK's (1250YTL monthly in 2008).

Current Status

This project has started in March 2009, and will finish in 2011.

Student Profile and Openings

Currently there are no funded openings for this project. However, especially undergraduate students who wish to take part in research carried out in fault diagnosis are welcome to participate. it is possible to work in eithter theoretical or experimental area.

1. Theroretical work involves:
   1. Modeling of nonlinear systems and
   2. Implementaion of fault diagnosis methods
2. Experimental work involves
   1. Implementing fault diagnosis methods based on neural networks and fuzzy logic on embedded systems (ARM9 systems with GCC)
   2. Data gathering on a real system through the CANBUS interface
   3. Preparation of fault scenarios
   4. Designing electronic circuits for experiments
   5. Web based aplication design for systems connected over GPRS or Edge

Control over Lossy Communication Networks

Background

A digital control system is made up of a real-time computer equipped with several sensors and actuators. This simple system may suffer from difficulty of modification, vulnerability to electrical noise or difficulty in maintenance and upgrades.

Networked control systems (NCS), on the other hand, aim to overcome these problems by locating sensors and actuators on separate computer nodes connected to the control computer by a communication network. Other advantages of introducing a communication network, such as high system testability and resource utilization, as well as low weight, space, power and wiring requirements, motivate the research on the NCS

However, networks can introduce unreliable and time-dependent levels of service in terms of delays, jitter, or losses. Quality-of-Service (QoS) can ameliorate the real-time network behavior, but the network behavior is still subject to interference (especially in wireless media), to routing transients, and to aggressive flows. In turn, network vagaries can jeopardize the stability, safety, and performance of the units in a physical environment. Active research aims at devising integrated control and communication algorithms to compensate for the vagaries of network service.

Networked control systems have direct applications in the industry, and in the control of large scale systems. A successor of this project is intended for such an application. Similarly, there are applications in group control of robots, teleoperation etc. Further expansions are towards decentralized control systems. For this reason, many researchers are taking up research in this new topic.

Students who are interested in this area will be exposed to such disciplines as real-time embedded systems, automatic control, communication, simulation and mechanical modeling of plants and electronics. There is opportunity to implement a full working digital control system from scratch, which will give insight into many theoretical concepts from undergraduate education.

Interested students should have basic knowledge in the fields of automatic control, embedded systems, and programming. Knowledge in other areas such as Linux kernel and socket programming, electronics is a bonus.

Current Status, student Profile and Openings

The project was concluded oficially. However, I am still interested in the area and actively working with cooperation with other faculty members and graduate students. Students (especially undergraduate) who are interested are welcome to help. I am planning to apply for another project as the continuation of this. Therefore a student who wishes to work on the subject can later be supported as a graduate student through the next project.