Naoki Kimura Laboratory
Naoki Kimura Laboratory

Chemical plants produce a variety of things that are familiar to us. For example, crude oil is used to produce gasoline and chemicals that are raw materials for other products. On the other hand, it is a facility that uses a large amount of energy and has the risk of accidents, such as handling toxic chemicals and operating at high temperature and pressure.
Therefore, in our laboratory, we conduct research on safety management, operation technology, and design methods to operate chemical plants safely, efficiently, and in consideration of the environment, called "process system engineering," and various data analysis techniques. We are working on research aiming to realize the design and operation of environmentally friendly and safe chemical processes using artificial intelligence and artificial intelligence technology.

Basic Information

Faculty name/Affiliation Naoki Kimura / Department of Information and Electronic Engineering, Faculty of Science and Engineering
Specialized Fields Process system engineering, Chemical engineering
Research theme
  • Anomaly detection and diagnosis system at chemical factory
  • Design support system at chemical factory
Research keywords Anomaly detection, anomaly diagnosis, design support
Faculty introduction URL https://www3.med.teikyo-u.ac.jp/profile/ja.dc37aa5b65510b3c.html

Our Research

Production Scheduling for Chemical Factories Considering Energy Conservation
Due to the recent world situation, the price rise of energy such as crude oil and raw materials is serious. This research theme develops a method that simultaneously considers energy saving using heat storage and waste heat in chemical plants and schedules for efficient production.

Abnormality detection and diagnosis in chemical plants
Chemical plants are equipped with sensors such as thermometers, pressure gauges, and densitometers. In this research, we will investigate methods for detecting abnormalities in chemical processes and identifying the causes of abnormalities (diagnosis) from the values of these sensors. We have already developed a technology to detect anomalies using a negative selection algorithm, which is one of the artificial immunization methods. I aim.

Conference presentation

Title Society name Laboratory Contents
Plant Fault Diagnosis System using Negative Selection Algorithm 14th International Symposium on Process Systems Engineering (PSE 2021+) Naoki Kimura Laboratory detail