论文部分内容阅读
随着国内经济的发展,航空汽车等产业近几年出现了井喷式的发展。应用于其中HMI(人机交互接口)的功能设计越来越复杂。在对HMI的测试中,自动测试方法与传统的手工测试方法相比,能够按照预先设定的顺序依次输入测试用例,通过对捕捉到的HMI反应画面进行识别分类,能够验证测试结果的正确与否,从而能解放部分劳动力,提高工作效率。人工神经网络作为图形识别的重要方法之一,已经得到了广泛的应用。本文将运用变步长的BP人工神经网络对大型客机的EICAS(发动机指示和机组警告系统)中的图像进行分类,实验表明BP方法能有效对图像进行分类。
With the development of the domestic economy, aerospace vehicles and other industries have seen a spurt-style development in recent years. The functional design applied to the HMI (Human Machine Interface) is more and more complicated. In the HMI test, compared with the traditional manual test methods, the automatic test method can input test cases in sequence according to a preset order. By identifying and classifying the captured HMI reaction images, it is possible to verify that the test results are correct No, so that we can liberate some of the workforce and improve work efficiency. Artificial neural network, as one of the important methods of pattern recognition, has been widely used. In this paper, the BP neural network with variable step size is used to classify the images of EICAS (engine indication and crew warning system) for large passenger aircraft. Experiments show that the BP method can effectively classify the images.