arXiv subjects

Guang Meng

Publications and source records attributed to Guang Meng.

4 recordsLinked to original sources

arXiv2025

SHapley Estimated Explanation (SHEP): A Fast Post-Hoc Attribution Method for Interpreting Intelligent Fault Diagnosis

Despite significant progress in intelligent fault diagnosis (IFD), the lack of interpretability remains a critical barrier to practical industrial applications, driving the growth of interpretability research in IFD. Post-hoc interpretability has gained popularity due to its ability to preserve network flexibility and scalability without modifying model structures. However, these methods often yield suboptimal time-domain explanations. Recently, combining domain transform with SHAP has improved interpretability by extending explanations to more informative domains. Nonetheless, the computational expense of SHAP, exacerbated by increased dimensions from domain transforms, remains a major challenge. To address this, we propose patch-wise attribution and SHapley Estimated Explanation (SHEP). Patch-wise attribution reduces feature dimensions at the cost of explanation granularity, while SHEP simplifies subset enumeration to approximate SHAP, reducing complexity from exponential to linear. Together, these methods significantly enhance SHAP's computational efficiency, providing feasibility for real-time interpretation in monitoring tasks. Extensive experiments confirm SHEP's efficiency, interpretability, and reliability in approximating SHAP. Additionally, with open-source code, SHEP has the potential to serve as a benchmark for post-hoc interpretability in IFD. The code is available on https://github.com/ChenQian0618/SHEP.

cs.LG↗

arXiv2025

CS-SHAP: Extending SHAP to Cyclic-Spectral Domain for Better Interpretability of Intelligent Fault Diagnosis

Neural networks (NNs), with their powerful nonlinear mapping and end-to-end capabilities, are widely applied in mechanical intelligent fault diagnosis (IFD). However, as typical black-box models, they pose challenges in understanding their decision basis and logic, limiting their deployment in high-reliability scenarios. Hence, various methods have been proposed to enhance the interpretability of IFD. Among these, post-hoc approaches can provide explanations without changing model architecture, preserving its flexibility and scalability. However, existing post-hoc methods often suffer from limitations in explanation forms. They either require preprocessing that disrupts the end-to-end nature or overlook fault mechanisms, leading to suboptimal explanations. To address these issues, we derived the cyclic-spectral (CS) transform and proposed the CS-SHAP by extending Shapley additive explanations (SHAP) to the CS domain. CS-SHAP can evaluate contributions from both carrier and modulation frequencies, aligning more closely with fault mechanisms and delivering clearer and more accurate explanations. Three datasets are utilized to validate the superior interpretability of CS-SHAP, ensuring its correctness, reproducibility, and practical performance. With open-source code and outstanding interpretability, CS-SHAP has the potential to be widely adopted and become the post-hoc interpretability benchmark in IFD, even in other classification tasks. The code is available on https://github.com/ChenQian0618/CS-SHAP.

cs.LG↗

arXiv2021

Adaptive nonreciprocal wave attenuation in linear piezoelectric metastructures shunted with one-way electrical transmission lines

In contrary to elastic media, it is easy to attain one-way coupling feature among electrical elements, which enables unidirectional transmission of electrical wave. In this paper, we explore and exploit the interaction of a piezoelectric beam with a one-way electrical transmission line to facilitate nonreciprocal transmission of elastic wave in the linear fashion. Theoretical dispersion analysis and numerical simulations are performed to reveal transmission behaviors of elastic wave in the presented piezoelectric metastructure. It is uncovered that, when the one-way electrical coupling feature is introduced, local-resonance bandgaps are maintained in opposite directions of the piezoelectric metastructure. However, the bandgap attenuation capability in one direction is increased compared to conventional local-resonance metastructures, in which no electrical coupling exists between cells, while that in the other direction is decreased. Therefore, nonreciprocal transmission of elastic wave emerges in this system due to the distinct bandgap attenuation capability. It is also revealed that this sort of nonreciprocal wave attenuation capability is adaptive with the one-way electrical coupling coefficient and the resistance of electrical cells. An experimental set-up of the piezoelectric metastructure is built and experimental results verify the nonreciprocity property and its adaptiveness. Overall, the presented piezoelectric metastructure provides a linear and concise approach to realize adaptive nonreciprocal transmission of elastic wave.

physics.app-ph↗

arXiv2017

Muon momentum measurement in ICARUS-T600 LAr-TPC via multiple scattering in few-GeV range

The measurement of muon momentum by Multiple Coulomb Scattering is a crucial ingredient to the reconstruction of νμ CC events in the ICARUS-T600 liquid argon TPC in absence of magnetic field, as in the search for sterile neutrinos at Fermilab where ICARUS will be exposed to ~1 GeV Booster neutrino beam. A sample of ~1000 stopping muons produced by charged current interactions of CNGS νμ in the surrounding rock at the INFN Gran Sasso underground Laboratory provides an ideal benchmark in the few-GeV range since their momentum can be directly and independently obtained by the calorimetric measurement. Stopping muon momentum in the 0.5- 4.5 GeV/c range has been reconstructed via Multiple Coulomb Scattering with resolution ranging from 10 to 25 % depending on muon energy, track length and uniformity of the electric field in the drift volume.

physics.ins-det↗