Yi Qiu

Stacked autoencoders have been widely used in soft sensing of industrial process data. However, most of them extract features by minimizing the reconstruction error of the input data at each layer, leading to accumulated loss. Moreover, the extraction of quality-related features is crucial for soft sensing. To address these issues, this paper firstly introduces a novel stacked dual-guided autoencoder (SDGAE) to better characterize intricate data patterns and learn quality-associated features. Compared to SAE, the idea of SDGAE lies in stacking a series of hierarchical dual-guided autoencoder (DGAE), aiming to allow each DGAE to accurately reconstruct the original input data. It also enables the model to simultaneously extract features that are highly correlated with the output variable by introducing a supervisor term. Furthermore, practical scenarios typically involve limited labeled data and a substantial volume of unlabeled data, which renders the network deficient in adequate generalization capabilities. To address this predicament, we further propose a deep semi-supervised stacked dual-guided autoencoder (SSDGAE), it is able to obtain useful feature representations and structural information from unlabelled data in the pre-training phase in addition to the labelled data as SDGAE is able to do. Finally, the proposed methods are verified on two real industrial case for its effectiveness and superiority. The results indicate that our approach can simultaneously utilize both labeled and unlabeled data to enhance the predictive performance of quality variables.

We report a numerical study of an incoherently pumped fiber optical parametric oscillator. Following the convection-induced phase-locking mechanism, we show that the oscillating sideband can be grown into a higher degree coherence.

The recent development of fiber mode-locked lasers with fiber parametric gain has opened up an intriguing field of ultrafast pulse generation. Thanks to the tunability of the parametric gain in optical fibers, the gain media of mode-locked lasers are no longer limited to the choice of rare-earth-doped fibers. In our previous study, optical time-stretch was introduced into the cavity of a fiber optical parametric oscillator as a temporal diffracting distance to overcome localization of the parametric gain. The implementation of the diffracting distance for pulse stretching and compression relied on two dispersive elements with opposite sign dispersion. In this context, we propose a new design of passively mode-locked fiber optical parametric oscillators with only a single mono-dispersive element. We utilize the parametric process as a time lens at which the spectral components of the two oscillating parametric sidebands are phase-conjugated in the temporal domain. This allows both mode-locked sidebands to oscillate and switch periodically inside the fiber cavity with only a single dispersive element. We numerically simulate the pulse generation of this switching mode-locked fiber parametric oscillator.

We propose a passively mode-locked fiber optical parametric oscillator assisted with optical time-stretch. Thanks to the lately developed optical time-stretch technique, the onset oscillating spectral components can be temporally dispersed across the pump envelope and further compete for the parametric gain with the other parts of onset oscillating sidebands within the pump envelope. By matching the amount of dispersion in optical time-stretch with the pulse width of the quasi-CW pump and oscillating one of the parametric sidebands inside the fiber cavity, we numerically show that the fiber parametric oscillator can be operated in a single pulse regime. By varying the amount of the intracavity dispersion, we further verify that the origin of this single pulse mode-locking regime is due to the optical pulse stretching and compression.