Once the marine accidents happen, it's very important to take prompt and effective relief measures to ensure the maritime life safety. But there are lots of difficulties due to weather, equipment, technology and other aspects of factors. Therefore, it is important to improve the success rate of salvage and take targeted measures by analyzing and evaluating of the difficulty of salvage.The paper finds out the current situation of the ability of the life salvage in our country from the aspects of salvage power and salvage methods. Firstly, it summarizes the current capability of the maritime life salvage from studying the rescuing power. Secondly, it determines the factors influencing the maritime life salvage from acquainting ourselves with the salvaging methods.It is to determine the index system of evaluating the difficulty of life rescue at sea taking consideration of environmental factors, human factors and rescue object factors.The paper tries to introduce the technology of Artificial Neural Network to difficulty evaluation of maritime life rescuing area. Three layers of BP neural network models is established by the neural network toolbox in MATLAB. The data of life salvage through collecting is calculated by the fuzzy comprehensive evaluation the results of which can be used for the network training as the training sample. The trained network will be applied difficulty evaluation of the actual life salvage. The result of network evaluation is more accurate and reliable based on the results of the evaluation if the quantity of the sample is big enough. It provides reference basis for the maritime life rescuing.It is theoretically meaning and practically valuable to apply the Artificial Neural Network to the maritime life rescuing area. This paper just makes a preliminary discussion. It needs deeper and more careful research in the future.
|