【摘要】 针对无人直升机控制中的滞后问题,文章提出了一种解决方法:先提取控制信号的特征值,然后利用特征值对无人机的进行判断,控制系统可以根据判断结果提前纠正系统的控制偏差。由于直升机飞行失败的样本极少,在稳定判断中引入了一个新的模式识别方法—支持向量机。支持向量机基于结构风险最小化原则,解决了小样本数据分类和泛化问题。文中在对支持向量分类机的原理进行了简单的介绍后,利用支持向量机和神经网络对直升机的飞行数据进行了分类,试验结果表明支持向量机具有较好的分类效果。更多还原

【Abstract】 The paper presents a method to solve the lag problem of a model-scale helicopter,which firstly extracts the control signal’s features,and then judges the flight stability,at last incorrects the control error.A stability classification method based on support vector machine is proposed.SVM can solve small sample problems and has good generalization ability using the principles of structural risk minimization.After a simple introduction to the principles of the method,the classification test is explained in detail;what’s more,several neural network methods are tested.Test results show that SVM has good performance to solve this kind of problem.更多还原