Research On Practical Technology Of Monitoring Of Main Marine Power System

Safety at sea is gaining more and more concern in shipping industry. With the continuous development of electronic technology, monitoring system in marine engine rooms has come into being. At the present stage, monitoring system in the engine room has certain shortcomings, including inaccurate monitoring, single alarm means and imperfect fault early warning, etc. Besides, there is a great gap between domestic and international technology. Therefore, it is of practical significance to carry out the study regarding such a topic.Based on the weak points of the existing monitoring system in the engine room, with main power system as the carrier, this paper designs and fabricates monitoring devices of main marine power system.Firstly, combined with the main marine power system, several forms of modern monitoring technology are introduced and then classifications and analyses are done towards parameters of the main power system. This paper selects security thermodynamic parameters as the monitoring subject, and then analyzes composition of the main power system from the parameter angle.Secondly, in accordance with actual demands, functional framework is constructed to perform the following three functions:status monitoring of the parameters, fault alarm and animated simulation, and fault diagnosis and early warning. Then selection of the hardware is completed and further the study focuses on research into the acquisition means of the parameters in this device so as to establish the parameter acquisition system which combines acquisition in the control room and on-spot collection.Thirdly, since fault diagnosis and early warning is the core function of this device, this paper illustrates on its technical background. The neural network is selected as mathematical algorithms to analyze the process of fault diagnosis and early warning. On the basis of specific procedure, the following problems are dealt with respectively, including failure mode and its symptom set, desired output, training sample set, test sample set and pretreatment of parameters. Due to the limited computing power of IPC configuration software—MCGS, establishment of the neural network can not be accomplished. In this paper,"cloud diagnosis" is adopted and second development of the IPC is done. With the aid of functional module of neural network—MATLAB, early warning is eventually realized.Finally, actual operation of the monitoring device is tested. Summary of this paper is made and expectations are laid onto the development of monitoring of main marine power system.