Search papers, labs, and topics across Lattice.
This paper introduces LiteInception, a two-stage deep learning framework for fault diagnosis in general aviation designed for resource-constrained edge devices. It uses a cascaded architecture for fault detection and classification, coupled with a multi-method fusion strategy for sensor channel reduction and a 1+1 branch LiteInception architecture for model compression. Knowledge distillation is used to adapt the model to different scenarios, and a dual-layer interpretability framework provides traceable evidence chains.
Achieve 8x faster CPU inference for aviation fault diagnosis with less than 3% F1 loss by compressing InceptionTime parameters by 70% – all while retaining interpretability.
General aviation fault diagnosis and efficient maintenance are critical to flight safety; however, deploying deep learning models on resource-constrained edge devices poses dual challenges in computational capacity and interpretability. This paper proposes LiteInception--a lightweight interpretable fault diagnosis framework designed for edge deployment. The framework adopts a two-stage cascaded architecture aligned with standard maintenance workflows: Stage 1 performs high-recall fault detection, and Stage 2 conducts fine-grained fault classification on anomalous samples, thereby decoupling optimization objectives and enabling on-demand allocation of computational resources. For model compression, a multi-method fusion strategy based on mutual information, gradient analysis, and SE attention weights is proposed to reduce the input sensor channels from 23 to 15, and a 1+1 branch LiteInception architecture is introduced that compresses InceptionTime parameters by 70%, accelerates CPU inference by over 8x, with less than 3% F1 loss. Furthermore, knowledge distillation is introduced as a precision-recall regulation mechanism, enabling the same lightweight model to adapt to different scenarios--such as safety-critical and auxiliary diagnosis--by switching training strategies. Finally, a dual-layer interpretability framework integrating four attribution methods is constructed, providing traceable evidence chains of"which sensor x which time period."Experiments on the NGAFID dataset demonstrate a fault detection accuracy of 81.92% with 83.24% recall, and a fault identification accuracy of 77.00%, validating the framework's favorable balance among efficiency, accuracy, and interpretability.
