基于ANN的舰船功率预测及性能恶化分析

Ship power prediction and performance deterioration analysis based on artificial neural network

通过利用人工神经网络(ANN)方法,综合舰船状态监测数据和海洋环境数据预测舰船推进功率,根据“功率-速度”曲线识别船体污垢后的性能参数,根据“燃油消耗率-功率”曲线识别部件老化后的主推进柴油机性能参数。同时对比分析采用ANN方法、K-最近域(KNN)方法、MLR(多元线性回归)方法预测模型的误差。旨在为使用维护人员提供对船体清洁和主推进柴油机维护的时间和方法,及时保持舰船在自主航行期间的良好状态。

Good maneuverability and good performance of power equipment are the basis for a ship to realize autonomous navigation.A ship’s overall performance degrades over operation time due to the hull fouling,equipment deterioration and other factors,which affect the ability of the ship to navigate autonomously.How to improve ship maneuverability and reduce fuel consumption and operating cost is becoming more and more important to the ship operation department and autonomous navigation.In this study,the artificial neural network is used to predict the propulsion power of a ship by ship condition monitoring data and ocean environment data.The performance deterioration caused by hull fouling is identified according to the power-speed curve,and the deterioration of main propulsion diesel engine performance caused by equipment deterioration is identified according to the fuel consumption-power curve.At the same time,the errors of the prediction model by ANN,KNN and MLR are 1.12%,1.36%and 5.23%.The purpose is to provide a method for maintenance personnel to clean the hull and maintain the main propulsion diesel engine in time to keep the ship in good condition during autonomous navigation.