Fire Fighting System Research for the Offshore Platform
|School||Harbin Engineering University|
|Course||Mechanical and Electronic Engineering|
|Keywords||offshore platform fire fighting data fusion neural network|
With the demand of petroleum is increasingly augmented in the world, people begins to pay their attention to the deep-sea oil exploitation. Nowdays the offshore platform can keep 24-hour uninterrupted oil exploitation. Under this circumstance, the offshore platform’s fire fighting system that plays a role of ensuring the normal operation is particularly important. This paper aims at designing the fire fighting system which is used in the offshore platform by considering the characteristics of the offshore platform and the onshore platform, for pointing out the direction for the follow-up study and engineering application.First of all, this paper expounds on significance and the current situation of the offshore platform’s fire fighting system in the world. By considering the characteristics of the offshore platform, the general structure of the offshore fire fighting system is defined. After that, there is a discussion on the automatic fire alarm system and fire extinguishing system respectively, then designs them according to the characteristics of the offshore platform.Finding the fire situation as early as possible, can gain the sufficient time for the fire alarm and fire quench. In view of this consider, the temperature detector, smoke detector and carbon monoxide detector are designed for the automatic fire alarm system, emphasises on the early fire stage detection. The next part designs the fire auto alarm controller, including the sensor module and data processing module. When fire is found, the fire auto alarm controller links the fire hydrant and smoke control system to put out the fire by linkage control system.Second is using the fire detector to realize the real-time monitoring of the offshore platform, and structuring the structure model and function model of the sensor data fusion simultaneously. Then here creates the BP neural network for the fire data processing module, applies MATLAB neural network module, trains the BP neural network by using sensor data, compares the results and actual data, then judges whether there is a fire in the current environment according to the setting value.Finally, the paper calculate the supply quantity and pipe network’s flow for the carbon dioxide extinguishing system and heptafluoropropane extinguishing system, analyzes the merit and demerit of both systems, and then puts forward a new extinguishing method that is combines heptafluoropropane extinguishing system and water spray cooling system, achieves an ideal effect.