基于支持向量机的飞机颠簸预测方法研究  

Research on Aircraft Turbulence Prediction Method Based on Support Vector Machine

作　　者：李朝潞 吴俊杰[1] 万连成 LI Zhao-lu;WU Jun-jie;WAN Lian-cheng(School of Air Traffic Management,Civil Aviation Flight University of China,Guanghan 618300,China)

机构地区：[1]中国民用航空飞行学院空中交通管理院,四川广汉618300

出　　处：《西安航空学院学报》2022年第3期24-28,57,共6页Journal of Xi’an Aeronautical Institute

摘　　要：为提高航空飞行安全,精确预测多变地形的飞机颠簸极为重要。以风向稳定的多变地形为例,选取影响飞机颠簸的参数并对其重新组合来分析多变地形对飞机颠簸的影响。将栅格化的5种地形参数代入四种不同核函数的支持向量机模型对飞机颠簸进行预测,并与实际结果进行对比,结果表明多项式核函数针对此训练集可达到最优。使用多项式核函数的支持向量机进一步分析,发现使用地形起伏度、最高高度、风的水平速度与垂直速度四种参数的预测结果最佳,准确率高达94.44%,证明这四种地形相关参数与飞机颠簸相关性最高。In order to improve aviation flight safety, it is very important to accurately predict aircraft turbulence over changeable terrain. Taking the variable terrain with stable wind direction as an example, the influence of variable terrain on aircraft turbulence is analyzed by selecting and recombining the parameters. The five rasterized terrain parameters are substituted into the support vector machine model with four different kernel functions to predict aircraft turbulence and are compared with the actual results. The results show that the polynomial kernel function is optimal for the training set. The polynomial kernel support vector machine is used for further analysis, and it is found that the prediction result with the four parameters, namely: terrain relief, maximum height, horizontal wind velocity and vertical wind velocity, is the most accurate one, and the accuracy is up to 94.44%. It proves that these four terrain parameters have the highest correlation with aircraft turbulence.

关 键 词：地形参数 支持向量机 飞机颠簸 栅格化 预测建模 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] V211.4[自动化与计算机技术—控制科学与工程]