关于舰艇油耗标准预测方法研究

Research on Prediction Method of Ship Oil Consumption Standard

舰艇油耗标准是舰艇部队的一项重要基础数据,正确地对舰艇油耗标准进行预测,是确定舰艇油耗标准的主要依据。由于舰艇油耗数据具有较明显的小样本特征,当部分数据异常,离散程度大时,传统方法预测的可信度就大大降低。针对现有方法在舰艇油耗标准预测中的局限性,设计实现了一种基于数据距离和灰关联度的舰艇油耗标准预测方法。首先通过设置数据距离的多级异常参数,实现异常数据的按层级分离;然后通过设置最小距离修正系数和最大距离修正系数,改进灰色距离测度法,实现油耗标准的预测。结果表明,通过修正最小距离的灰色距离测度法获得的油耗标准预测值,能体现舰艇的实际油耗,具有较高的可信度和准确度。对比分析证明了上述方法的有效性和合理性。

Ship oil consumption standard is an important basic data of naval forces. Prediction is the main basis for the standard amendment. The ship oil consumption data are small, and the reliability of prediction will lose when the data are abnormal. A method of ship oil consumption standard prediction based on data distance and grey relation- ship is proposed. First, the outlier data are separated as different groups and classes by outlier data parameters. Then, the method of the grey data distance is optimized by the parameters of the minimum grey distance and the maxi- mum grey distance, and the prediction of oil consumption standard is achieved. The results show that the prediction value of oil consumption standard based on the minimum grey distance method can represent the actual consumption of ship, and the value has high reliability and accuracy. The analysis shows that the method based on grey distance by minimum distance parameter is available and reasonable.