支持向量机回归算法在船舶磁场推算中的研究

Research of support vector machine regression algorithm in shipment magnetic field calculation

船舶航行中会遇到复杂的外界环境,其中磁场环境的变化对航行状况具有重要的影响。而船舶内部大量的电力系统会产生复杂的磁场分布。在建立的船舶磁场分布模型中,传统的基于磁场检测后的逆向拟合方法精确度不高。因此可以根据实验室的测量数据建立基于支持向量机回归算法的船舶磁场模型。在该模型中,预测的磁场分布具有更高的精度和复杂度,并且可以对磁场的动态变化进行预测,建立船舶磁场的时变模型。

Nowadays,the shipment will be surrounded by complex outside environment,and the complicated magnetic environment will greatly influence the navigation conditions. And the power system inside the shipment contributes to the complexity of magnet distribution. In order to build the model for magnet distribution,the traditional approaches of reverse-fitting based on measurement cannot provide accurate model. Hence we can build distribution model via support vector machine regression algorithm based on data measured in laboratory. This model could provide higher accuracy and complexity,and predict the dynamic change of magnet,and therefore the time-varying model of shipment magnet.