LS-SVM在提高气压测高系统精度中的应用研究

在利用计算机软件补偿飞机测高系统误差方法中,传统的采用分段拟合方法精度不高,非线性误差大;神经网络存在过学习、网络拓扑结构不易确定以及泛化能力差等缺点。提出利用最小二乘支持向量机(LS-SVM)对某型飞机测高系统误差进行预测、补偿,提高其测高精度。在分析LS-SVM回归算法的基础上,利用LS-SVMlab1.5对某型飞机测高系统误差样本数据进行了仿真。结果表明,LS-SVM在数据回归预测方面精度高、泛化能力强、稳定性好,可以有效提高飞机气压测高系统精度,具有应用推广价值。

Application Research of LS-SVM for Improving Accuracy of Altitude Measurement System

Among methods which compensate the altitude measurement system error using computer software,the classical method is lower in accuracy and greater in nonlinear error,the neural network method has many disadvantages in the aspect of over learning,difficult confirming the network topology structure and poor generalization ability etc.In order to improve the accuracy of altitude measurement system,a method based on least squares support vector machine(LS-SVM) is presented.On the basis of analyzing the algorithmic of LS-SVM for regression,the sample data of a altitude measurement system by LS-SVMlab1.5 are simulated.The result shows that the method based on LS-SVM is high in accuracy,better generalization ability and stability in regression and forecast of data.It can improve the accuracy of altitude measurement system effectively and has a good prospect of application and extension.