Home / Information / Laboratories and Academic Staff / Composite Systems Theory

Composite Systems Theory

Composite systems theory is the body of theory concerned with "using systems as components for constructing larger systems". The objective of our lab is to foster the development of theory and technology for dealing with complex systems, particularly methods for automatic control and scheduling, and to apply these to medical and industrial applications. Currently, the course contains subjects such as "Medical Systems Engineering","Theory of Process Control", and "System Optimization".

Academic Staff

Shinji DOI

Shinji DOIProfessor (Graduate School of Engineering)

Research Interests

Contacts

Katsura Campus A1-420
TEL: +81-75-383-2200
E-mail: doi(at)kuee.kyoto-u.ac.jp

Eiko FURUTANI

Eiko FURUTANIAssociate Professor (Graduate School of Engineering)

Research Interests

  • physiological state control systems
  • estimation of physiological states of patients
  • time delay systems

Classes

  • System Optimization
  • Control Engineering
  • Applied Systems Theory

Contacts

Katsura Campus A1-417
TEL: +81-75-383-2202
FAX: +81-75-383-2203
E-mail: furutani(a)kuee.kyoto-u.ac.jp

Shunji TANAKA

Shunji TANAKAAssociate Professor (Institute for Liberal Arts and Sciences)

Research Interests

  • production scheduling
  • warehousing (AS/RSs)
  • elevator group control

Contacts

Katsura Campus A1-418
TEL: +81-75-383-2204
E-mail: tanaka(at)kuee.kyoto-u.ac.jp
http://turbine.kuee.kyoto-u.ac.jp/~tanaka/index-e.html

Masayuki KIMURA

Masayuki KIMURAAssistant Professor (Graduate School of Engineering)

Research Interests

  • Nonlinear oscillation
  • Energy Localization Phenomena

Contacts

Katsura Campus A1-418
TEL: +81-75-383-2204
E-mail: kimura.masayuki.8c(at)kyoto-u.ac.jp
http://turbine.kuee.kyoto-u.ac.jp/~kimura/

Introduction to R&D Topics

Medical Systems Engineering

In this lab we pursue the development of automatic control systems for maintaining the physiological state of patients at appropriate levels, with the aim of reducing the burden on patients as much as possible in surgery and medical treatment. In addition, we conduct research into state estimation methods based on brain waves for assessing the state of patients as precisely as possible. By conducting clinical applications, we verify the effectiveness of a blood pressure control system for reducing bleeding and an intravenous anesthesia hypnosis control system for reducing the total amount of anesthetics during surgery. We are currently pursuing the research outlined below.

Anesthesia control system for surgery
During surgery it is desirable to appropriately maintain three elements – hypnosis, analgesia and muscle relaxation – of patients by using the minimum possible dose of anesthetics. We are investigating suitable indicators for monitoring these elements and working to develop a control system that automatically maintains all three elements at respective appropriate levels.
Blood glucose control system for post-operative diabetes
It is desirable to avoid post-operative diabetes, whereby elderly people and diabetics temporarily experience hyperglycemia following surgery, since it can lead to infection or failure of sutures. We are thus studying a model for the regulatory system of blood glucose that can deal with temporary changes, and we are developing a control system for automatically maintaining blood glucose at suitable levels.
State estimation for mental disorders
We are conducting research into methods for estimating the changes of state that occur through treatment, with the aim of determining the optimal parameters for modified electroconvulsive therapy, which is sometimes used for people suffering mental disorders when pharmacotherapy is ineffective.

Theory of Process Control

In this lab we are conducting a theoretical investigation of process control for application to production facilities. Details of this research are given below.

Design methods of two-degree-of-freedom control system
Two characteristics of control systems that are important to understand are its reference tracking ability (i.e., its ability to determine how well a parameter can be regulated to a set point), and its disturbance rejection ability (i.e., to indicate the extent to which the system is affected by external disturbances). We are working on design methods for two-degree-of-freedom control systems — that is, systems in which these two abilities can be regulated independently.
Research on control methods for plants with time delays
Often the input/output channels of systems can include a delay due to the time taken for a substance to flow or for a signal to be measured. This delay is known as a "time delay." Effective regulation is impossible when there is any time delay in a system unless state predictions are made, as is the case in shower water temperature control. We are pursuing research into design and analysis methods for control systems with this kind of time delay, focusing on state-predictive control methods and model predictive control methods for plants.

System Optimization

In this lab, we treat a variety of real-world issues as optimization problems and conduct research and analysis with the aim of finding solutions for these problems.

Production scheduling and machine scheduling problems
The term "scheduling problems" refers to the problems to  create efficient plans for tasks, machines and people in a production environment. It is generally very difficult to find optimal solutions to these problems. We are working on methods for efficiently finding optimal or near-optimal solutions to these kinds of problems.
Input/output scheduling problems of Automated Storage and Retrieval System (AS/RS)
An Automated Storage and Retrieval System (AS/RS) is a system that automates storage and retrieval of items in distribution centers and factories. It consists of parallel storage racks and stacker cranes to load and unload the items on the storage racks. We treat the input/output scheduling problem of AS/RSs as an optimization problem and are working on exact and heuristic algorithms to solve it.
Elevator group control problems
We are studying elevator group control as an optimization problem to control several elevator cars cooperatively, whose aim is to improve the overall transportation capability of the elevator system. In particular, we are focusing on the group control of an elevator system with destination hall-call registration.