飞机起飞着陆性能智能计算模型及应用

The Aircraft Take-off and Landing Performance Intelligent Computation Model and Its Applications

提出了一种飞机起飞着陆性能智能计算思路。详细分析了影响飞机高原起飞着陆性能的主要因素,提出了高原起飞着陆性能智能计算的一般模型,然后采用支持向量机(Support Vector Machines,SVM)对某型飞机高原起飞滑跑距离实测数据进行了建模和验算,同时为说明支持向量机模型适合工程使用、精度高、推广性高的优点,还与贝叶斯正则化BP神经网络(BRBP)、RBF神经网络(RBF)、自适应神经模糊推理系统(ANFIS)做了比较。计算结果表明,支持向量机具有很好的推广性能,得到的结果优于BRBP,RBF和ANFIS等智能计算方法,推广误差能够满足工程的实际需要。该模型对于发展和丰富飞行器起飞着陆性能计算理论具有一定的参考价值。

A novel aircraft take-off and landing performance intelligent computation model was proposed in this paper. First, main factors towards aircraft tableland take-off and landing performance were minutely analyzed. Second, the general the model for aircraft tableland take-off and landing performance computation was proposed. Final, support vector machines （SVM） were applied in real taxing distance data of some type of aircraft. Meantime, so as to show superiority, high-precision, and virtues of SVM, Bayesian regularized BP neural networks （BRBP）, radical basis function（RBF） neural networks, adaptive network based fuzzy inference systems （ANFIS） were also investigated. Results show that SVM is well suited for practical engineering use with satisfied generalized performance, as it is better than other intelligent computation methods in the case studied in this paper. The proposed intelligent computation model is especially commendable for developing and enriching the aircraft take-off and landing performance computation theory.