Some Aspects of Intelligent Control in Vessel Systems

• Main Authors: Ker-Wei Yu, 俞克維
• Other Authors: Jer-Guang Hsieh
• Format: Others
• Language: zh-TW
• Published: 1999
• Online Access: http://ndltd.ncl.edu.tw/handle/73745636438179658491

博士 === 國立中山大學 === 電機工程學系 === 88 === In this dissertation, some aspects of intelligent control in vessel systems are considered. The problems of vessel systems under consideration are divided into two parts, namely the fault diagnostic problem of marine engineering systems and ship maneuvering problems. First, the problems of intelligent fault diagnostic tool for marine engineering systems are investigated. The knowledge of experts and the results of experiment will be used to construct the relational database for symptoms and fault reasons. In this dissertation, two packages of real-time intelligent PC-based fault diagnostic tool will be developed for the system of oil-fired power plant with turbine generator and marine diesel engine system by using a neural network. These packages can be used to automatically monitor the running situation of the marine engineering systems and to reveal the fault symptoms in case the systems are in trouble. They can also be used as handy training tools for the laboratory course in marine engineering systems. Second, ship-handling problems of maritime navigation are considered. The ship-handling problems can generally be classified into four types, namely course keeping, course changing, course tracking, and other special motions. The PID controllers have been commonly used for traditional ship maneuvering control. However, to face more versatile environments, higher levels of robustness and/or adaptive capability, have been proposed as possible successors to the PID. Fuzzy control is a modern control technique which is currently finding an increasing and diverse range of novel applications, both in its fixed rule base and intelligent forms. In this dissertation, a fuzzy gain scheduling PID controller by using back-propagation algorithm is developed. The proposed controller is self-adaptive that may accommodate the effect caused by disturbance such as wave or wind. Some computer simulations are provided to illustrate the use of our main ideas.