论文部分内容阅读
挠联式惯导系统的可靠性是其应用中的一个关键性问题，在改善制造工艺的同时，对其故障能予以及时、正确的诊断亦是至关重要的。神经网络技术具备解决故障诊断问题的能力，且与其它方法相比具有相当的优势。为了能对系统故障进行在线实时检测，我们基于DSP设计了一种种经计算模型，在软、硬件方面都进行了具体设计，且模型具有一定的通用性。
The reliability of a Torsion-linked Inertial Navigation System is a key issue in its application. It is also crucial to improve the manufacturing process while diagnosing its failure timely and correctly. Neural network technology possesses the ability to solve the problem of fault diagnosis and has considerable advantages over other methods. In order to detect system faults online in real time, we design a kind of computational model based on DSP and carry out specific design in terms of both hardware and software, and the model has certain versatility.