We proposed and demonstrated the ultra-compact 1 310/1 550 nm wavelength multiplexer/demultiplexer assisted by subwavelength grating (SWG) using particle swarm optimization (PSO) algorithm in silicon-on-insulator (SOI) platform. Through the self-imaging effect of multimode interference (MMI) coupler, the demultiplexing function for 1 310 nm and 1 550 nm wavelengths is implemented. After that, three parallel SWG-based slots are inserted into the MMI section so that the effective refractive index of the modes can be engineered and thus the beat length can be adjusted. Importantly, these three SWG slots significantly reduce the length of the device, which is much shorter than the length of traditional MMI-based wavelength demultiplexers. Ultimately, by using the PSO algorithm, the equivalent refractive index and width of the SWG in a certain range are optimized to achieve the best performance of the wavelength demultiplexer. It has been verified that the device footprint is only 2×30.68 µm², and 1 dB bandwidths of larger than 120 nm are acquired at 1 310 nm and 1 550 nm wavelengths. Meanwhile, the transmitted spectrum shows that the insertion loss (IL) values are below 0.47 dB at both wavelengths when the extinction ratio (ER) values are above 12.65 dB. This inverse design approach has been proved to be efficient in increasing bandwidth and reducing device length.

Quantum cascade lasers (QCLs) have broad application potentials in infrared countermeasure system, free-space optical communication and trace gas detection. Compared with traditional Fabry-Pérot (FP) cavity and external cavity, distributed feedback quantum cascade lasers (DFB-QCLs) can obtain narrower laser linewidth and higher integration. In this paper, the structure design, numerical simulation and optimization of the Bragg grating of DFB-QCLs are carried out to obtain the transmission spectrum with central wavelength at 4.6 µm. We analyze the relationship among the structure parameters, the central wavelength shift and transmission efficiency using coupled-wave theory and finite-difference time-domain (FDTD) method. It is shown that the increase in the number of grating periods enhances the capabilities of mode selectivity, while the grating length of a single period adjustment directly determines the Bragg wavelength. Additionally, variations in etching depth and duty cycle lead to blue and red shifts in the central wavelength, respectively. Based on the numerical simulation results, the optimized design parameters for the upper buffer layer and the upper cladding grating are proposed, which gives an optional scheme for component fabrication and performance improvement in the future.

A D-type photonic crystal fiber (PCF) sensor based on surface plasmon resonance (SPR) principle is designed. In order to excite the SPR effect, a gold film is plated on the open-loop channel of the sensor, the free electrons in a metal are resonated with photons. The structural parameters are fine-tuned and the sensing performance of the sensor is studied. The results show that the maximum spectral sensitivity reaches 18 000 nm/RIU in the refractive index range of 1.24–1.32, and the maximum resolution is 5.56×10⁻⁶ RIU. The novel structure with high sensitivity and low refractive index provides a new perspective for fluid density detection.

Due to their unique physical properties, nonlinear materials are gradually demonstrating significant potential in the field of optics. Gold nanoparticles supported on carbon black (Au/CB), possessing low loss and high nonlinear characteristics, serve as an excellent material for saturable absorber (SA) in ultrafast fiber lasers. In this study, we investigated the performance of Au/CB material and designed an ultrafast fiber laser based on Au/CB SA, successfully observing stable fundamental mode-locking and pulse bunch phenomena. Specifically, when the fiber laser operates in fundamental mode-locking state, the center wavelength of optical spectrum is 1 558.82 nm, with a 3 dB bandwidth of 2.26 nm. Additionally, to investigate the evolution of real-time spectra, the dispersive Fourier transform (DFT) technology is employed. On the other hand, the pulse bunch emitted by the laser is actually composed of numerous random sub-pulses, exhibiting high-energy characteristics. The number of sub-pulses increases with the increase of pump power. These findings contribute to further exploring the properties of Au/CB material and reveal its potential applications in ultrafast optics.

High peak-to-average power ratio (PAPR) is the main disadvantage of visible light communication-based orthogonal frequency division multiplexing (VLC-OFDM) systems. To address this problem, a novel precoding method is proposed in this paper. The complex-valued precoding matrix is constructed by a Vandermonde matrix. The researched results show the proposed precoding scheme has better PAPR performance when compared to the conventional real-valued precoding methods. Moreover, a general closed-form expression of bit error rate (BER) for Vandermonde precoded VLC-OFDM is derived for multipath fading channel. The obtained BER formula shows that Vandermonde precoding can improve the BER performance of VLC-OFDM system over multipath fading channel. This is verified by the simulation results. The researched results also show that different precoding schemes have the same BER performance but different PAPR performance.

In this paper, a two-stage light detection and ranging (LiDAR) three-dimensional (3D) object detection framework is presented, namely point-voxel dual transformer (PV-DT3D), which is a transformer-based method. In the proposed PV-DT3D, point-voxel fusion features are used for proposal refinement. Specifically, keypoints are sampled from entire point cloud scene and used to encode representative scene features via a proposal-aware voxel set abstraction module. Subsequently, following the generation of proposals by the region proposal networks (RPN), the internal encoded keypoints are fed into the dual transformer encoder-decoder architecture. In 3D object detection, the proposed PV-DT3D takes advantage of both point-wise transformer and channel-wise architecture to capture contextual information from the spatial and channel dimensions. Experiments conducted on the highly competitive KITTI 3D car detection leaderboard show that the PV-DT3D achieves superior detection accuracy among state-of-the-art point-voxel-based methods.

An improved cycle-consistent generative adversarial network (CycleGAN) method for defect data augmentation based on feature fusion and self attention residual module is proposed to address the insufficiency of defect sample data for light guide plate (LGP) in production, as well as the problem of minor defects. Two optimizations are made to the generator of CycleGAN: fusion of low resolution features obtained from partial up-sampling and down-sampling with high-resolution features, combination of self attention mechanism with residual network structure to replace the original residual module. Qualitative and quantitative experiments were conducted to compare different data augmentation methods, and the results show that the defect images of the LGP generated by the improved network were more realistic, and the accuracy of the you only look once version 5 (YOLOv5) detection network for the LGP was improved by 5.6%, proving the effectiveness and accuracy of the proposed method.

The detection of surface defects in concrete bridges using deep learning is of significant importance for reducing operational risks, saving maintenance costs, and driving the intelligent transformation of bridge defect detection. In contrast to the subjective and inefficient manual visual inspection, deep learning-based algorithms for concrete defect detection exhibit remarkable advantages, emerging as a focal point in recent research. This paper comprehensively analyzes the research progress of deep learning algorithms in the field of surface defect detection in concrete bridges in recent years. It introduces the early detection methods for surface defects in concrete bridges and the development of deep learning. Subsequently, it provides an overview of deep learning-based concrete bridge surface defect detection research from three aspects: image classification, object detection, and semantic segmentation. The paper summarizes the strengths and weaknesses of existing methods and the challenges they face. Additionally, it analyzes and prospects the development trends of surface defect detection in concrete bridges.