Onchip refractive index sensing has broad application prospects in the fields of safety inspection, environmental monitoring, health diagnosis, food supervision, etc. Currently, the microring resonator refractive index sensor based on silicononinsulator platform has become a research hot spot due to its capacities on miniaturization and integration. In this paper, its principles are introduced firstly, and then, the latest research progresses in and abroad are sorted out according to the principles of detection, and also the advantages and disadvantages are analyzed and summarized simultaneously. Finally, the existing problems and future development trends of refractive index sensing based on silicononinsulator microring resonators are discussed.

Photonic crystal fiber (PCF) can be designed flexibility. In general, the propagation properties of the PCF are adjusted by changing the sizes, shapes or arrangement of the air holes, and then the PCF with high birefringence, high nonlinearity, flattened dispersion and low confinement loss can be obtained. PCF has potential applications in different fields, such as distributed optical fiber sensor, femtosecond laser, gas sensor and so on. In this paper, firstly, the structure characteristics of PCF and the difference between the traditional singlemode fiber (SMF) and the PCF are introduced. Secondly, their typical structures and propagation properties are analyzed. Thirdly, the sensing mechanisms and performance of distributed optical fiber sensor, femtosecond laser and gas sensor based on the PCF are introduced in detail. The results show that compared with the traditional SMF, the PCF can achieve more excellent performance in applications. Finally, the developments and applications of PCF are summarized and prospected.

Allfibertype sideviewing opticalcoherencetomography (OCT) probe has attracted extensive interests in both academic and industry because of its ultrasmall size and the ability of fullcircumferential imaging. In this manuscript, the design of allfibertype sideviewing OCT probe is optimized for better focusing performance by numerical simulation. Furthermore, the robustness of the optimized allfibertype sideviewing OCT probe versus the size of probe tube and the position of reflector is numerically analyzed for the power fading resulting from the random transverse displacement between reflector and GRIN fiber.

To achieve highsensitivity refractive index measurement, an optical fiber refractive index sensor based on doublemode interference is designed. The sensor is a singlemodemultimodesinglemode fiber (SMS) structure, in which the singlemode fiber and the multimode fiber are of the same diameter and core refractive index. To strip the cladding of the multimode fiber and place it in the environment to be measured, interference will occur between the two modes excited out of the multimode fiber. The transmission dips of the sensor will shift when the ambient refractive index changes, so the simultaneous measurement of the refractive index can be achieved by monitoring the wavelength shifts. Using FDTD Solutions software for simulation, the interference spectra at different refractive index were obtained. The results show that the sensitivity of the dip at 1517.32nm to the refractive index is 1848.2nm/RIU, and the minimum resolution error is only 2.74×10-5RIU. Compared with the traditional optical fiber sensor, the sensor processes simple structure, high sensitivity and low resolution, making it has high potential applications.

Designed is an onboard balanced photoelectric detector with high sensitivity, which is composed of a balanced photodiode chip and a transimpedance amplifier. The balanced photodiode chip adopts the internal balance structure of dual InPInGaAs photodiode unit to reduce the noise of the chip itself and improve the detection sensitivity. The closedloop amplifier integrated with positive and negative bidirectional current input circuit, automatic gain control circuit and inverter input circuit was simulated by Cadence software, and the relationship of the equivalent noise power, bandwidth and gain was obtained. Then a tranresistance amplifier being adaptive to balanced photodiode chip was produced. A 1.55μm laser test system was set up to test the performance of the developed detector, and the results show that its 3dB bandwidth of 1.58GHz, equivalent noise power density of 5.96pW/Hz1/2, and the commonmode rejection ratio of 42.04dB (@DC~1.58GHz) are realized, and the receiving sensitivity in the coherent laser communication system reaches -61dBm.

In this work, organic small molecule 2,9Dimethyl4,7diphenyl1,10phenanthroline (bathocuproine, abbr. BCP) was selected as a buffer layer for interface engineering between active and electrontransporting layers (ETL) in a reversed perovskite solar cells. It was found that the device performance was significantly improved after THE deposition of a BCP layer between the ETL and the perovskite layer. It reveals that the spacing between the grains was effectively filled by the BCP by scanning electron microscope, which can be responsible to prevent the generation of defect state at perovskite interface. Further analysis of the charge accumulation in the device in combination with AC impedance measurement proved that the migration of ions and the photongenerated carrier recombination in the perovskite solar device were reduced after the deposition of BCP layer. As a result, the optoelectronic powerconversionefficiency is improved from 15.7% to 17.4%.

For the imaging requirements such as high frame rate, global shutter and adjustable output channel amount, a kind of CMOS image sensor for hyperspectral imaging is proposed. The schematic plan and working principles of 8T pixel structure and readout circuit are introduced in detail. Postsimulation and tapeout of this sensor were completed, and the experimental results show the design meets the expectation by achieving targeting features and realizing highspeed and lownoise analog signal output under the condition of large layout area and small amount of output channels. The pixel array size of 2048×256 and the pixel pitch of 24μm×24μm are achieved, the maximum signal output frequency for a single channel is 40×106pixel/s and the frame rate can reach up to 4000f/s, being able to meet the general requirement of a hyperspectral imaging system.

In order to explore the influence of local high radiation on short-circuit current of multijunction solar cell in concentrated photovoltaic module with Fresnel lens as the concentrator, the diameters of the high bright spot of Fresnel lens were measured so as to measure the shortcircuit current of the GaInP/GaInAs/Ge triplejunction solar cell with different sizes of lens under different local illumination areas. And the experimental results were analyzed by using circuit network model. The results show that the shortcircuit current is independent of the area of bright spot, and the current along the optical axis changes synchronously with radiation when the small size Fresnel lens is focused. When the size of lens increases to a certain extent, high radiation presents inthe center of solar cell, and the current is limited by peak tunneling current, which is reflected as the decrease of shortcircuit current at the focal plane. Putting the solar cell on both sides of the focal plane canalleviatethe local high radiation, and the shortcircuit current can be increased by 8.0%.

Lowcost and miniaturized semiconductor lasers play an important role in optical fibersensing system. In this paper, a design of semiconductor laser driving circuit for temperature tuning is experimentally presented. Using the ARM chip as the core controller, the design applies a thermistor to collect the temperature signal, and uses the thermoelectric cooler (TEC) to control the temperature, thus the wavelength scanning of the laser can be achieved by temperaturecontrolled tuning. Then the output power is collected by a photo diode (PD), and the optic power of the laser is controlled by electric current, so that the optic power can keep stable with the changing temperature. Experiments results show that the circuit can work reliably for a long time, the max temperature tuning range reaches 5nm, and the output optical power remains stable during the whole wavelength scanning process.

A tunable random fiber laser based on random grating and high reflection FBG is proposed. A 980nm pump light source is used to pump a 7m long erbiumdoped fiber for gain amplification, and random feedback is provided by a random grating. The random grating is 7cm long and has about 10,000 refractive index modification points. These points are written point by point by the femtosecond laser and randomly distributed along the fiber direction. The distance between the two points is less than 10μm. At the same time, a semiopen cavity structure was formed by using a highreflection FBG with a center wavelength of 1548nm to achieve random laser output. The experimentally measured pump threshold power is only 18mW, and the slope efficiency is as high as 13.2%. By changing the center wavelength of the FBG, the output laser wavelength can be tuned with a tuning range of 4.45nm (1548.04~1552.49nm). Thanks to the design of the semiopen laser cavity and the high gain of EDF, the entire system has the advantages of low threshold, high efficiency, and simple structure.

Black silicon with textured surface was prepared on mcSi by metal assisted chemical etching (MACE) technique using Ag element. The surface Ag residues, morphology and optical absorption of the prepared black silicon samples were characterized and analyzed by energy dispersive spectrometer, scanning electron microscope and reflectivity tester. On this basis, solar cells were fabricated with the black silicon samples with standard processes of diffusion, etching, PECVD coating, screen printing and sintering. And their electrical and electroluminecence (EL) properties were characterized and analyzed by IV and EL test. The results show that compared with the regular and uniform circular texture surface without Ag residues, the surface morphology of black silicon samples with Ag residues was disorderly in direction and different in depth. The efficiency of the related solar cell was 3.03% lower than that of the normal cell due to the sharp boundary around the pit and the irregular shape.

Endactuator is very important for realizing fast, reliable and efficient transmission in substrate detection of semiconductor packaging. Under the condition of satisfying material strength and structure stiffness, a 3D model of the endactuator was established with the goal to reduce the weight of the endactuator. The static calculation and modal analysis of the endactuator of the substrate transfer robot were performed with ANSYS software. Then the stress, strain characteristics and corresponding modes of the endactuator under the maximum load were obtained. Based on the optimized topology, a new endactuator structure was established. And the strength of the new structure was checked, verifying the validity of the design scheme. The results demonstrate that the first four order natural frequencies of the endactuator are much higher than the gyrofrequency of the servomotor and the quality of the new structure is reduced by 26.7%. It achieves the goal of lightweight and provides a new technical solution for the subsequent development of related products.

A fluid model is used to investigate the discharge kinetics of ArF excimer laser. By comparing the gas discharge at different initial preionization intensities, the effects of preionization on discharge excited excimer laser system are analyzed. The preionization effect under different gas parameters is also discussed. The results show that the initial preionization intensity has significant effects on the breakdown voltage, the formation of ArF excimer, the optical quality and gain. Under the premise of uniform electric field and effective discharge, it is difficult to reach the gas breakdown threshold with low initial preionization intensity, but higher laser power can be obtained. Increasing the initial preionization intensity can effectively reduce the breakdown voltage, but it’s not conducive to the absorption and conversion of energy. The preionization capability is affected by the gas pressure and fluorine ratio. The increasing of gas pressure or fluorine concentration can reduce the effectiveness of preionization.

The nearzero dielectric constant mode provides a new way to control the interaction of light and matter at the nanoscale. In this paper, the principle of attenuated total reflection was used to excite the ENZ modes of phonon polariton and gallium nitride films of aluminum oxide substrates respectively. By adjusting the film thickness and optimizing the gap between air layers, the coupling hybrid of the doublelayer structure mode was achieved. The transmission matrix method and finite time domain difference method were used to numerically calculate the electric field distribution and dispersion of the mixed mode, indicating that the coupling splitting occurs in the two modes due to the strong coupling conditions. In the spectrum simulated by FDTD, both the highfrequency and lowfrequency branches of the dispersion relationship show strong coupling at the anticrosspoints, and the ENZSPhP mixed mode can be observed from the electric field diagram. ENZSPhP hybrid mode has high propagation characteristics and subwavelength light limitation, which can be used as the basis for future infrared and terahertz nanophotonic integration and communication equipment.

A twodimensional geometry model of hydride vapor phase epitaxy (HVPE) growth chamber with the diameter of 15.24cm (6inch) was established to simulate the growth of Ga2O3 material. The key parameters were optimized, such as the inlet velocity of GaCl and O2 and the distance between the nozzle and the substrate. The uniformity of the thickness of Ga2O3 film on the substrate reached 7.02% at a relatively high growth rate. In addition, simulation experiments were carried out under different reactive activation energy parameters. It is found that although the activation energy parameters affect the average growth rate significantly, they have little influence on the optimal design of growth rate distribution and uniformity of the samples.

Polarized lightemitting films have potential applications in liquid crystal displays and can significantly reduce energy consumption. The key to the preparation of polarized lightemitting films lies in the ordered orientation of the lightemitting materials. In this paper, the anisotropic CsPbBr3 nanowires were used as the luminescent material, and the offcenter spincoating method was used to perpare the polarized luminescent film. The fluorescence spectrometer was used to test the degree of polarization of the film, and scanning electron microscope was used to observe the orientation of the nanowires. The orientation of the nanowire film was optimized by changing the substrates, nanowire dispersion concentration, and solvent composition, and a good fluorescence polarization degree was obtained.

The heat dissipation effect of the interleaved micropinfin heat sink was studied by numerical simulation, and the influence of the array spacing and height of the pin fins on the wall temperature and pressure drop of the heat sink was studied. Experimental results indicate that the press drop increases nonlinearly with the decreasing array spacing and height of the pin fins, while the wall temperature increases with the increase of the pin fin spacing, but it shows a fluctuation trend with the increase of the pin fin height. Based on the above research, it further optimizes the lateral spacing, longitudinal spacing and height of the microneedle fins with the method of response surface approximation. The pressure drop can be reduced by a maximum of 61.11% after optimization.

Quantum temporal ghost imaging utilizes the timefrequency correlation of photon pairs to realize the image transmission over optical fiber link of 50km, which shows the potential applications in protocols of quantum communications. But the present schemes usually present poor imaging quality poor and low transmission speed, limiting the future application to some extent. In this paper, the scheme is considerably modified and a programmable optical filter is used to digitally set the information of the imaging object. Based on the modification, the twodimensional image is transmitted automatically in realtime. Experimental results demonstrate that both the imaging quality and rate are notably improved with the modified scheme, pushing the applications of quantum correlation imaging in the field of quantum communications.

In order to maintain the fast performance of the direct method and the high precision and loop closure capability of the featurebased method, a RGBD simultaneous localization and mapping (SLAM) algorithm combining the direct method and the featurebased method is proposed. The proposed algorithm is composed of three parallel threads: tracking thread, local mapping thread and loop closing thread. In the tracking thread, the nonkey frames are tracked, the initial pose estimation and the corresponding relationship calculation of pixel points are carried out by minimizing the photometric image errors, and the camera pose is further optimized by minimizing reprojection errors of the local map points to achieve fast and accurate tracking and positioning. In the local mapping thread，the ORB features are extracted and matched features in the key frames, and the local BA (Bundle Adjustment method) is performed to optimize the position and posture of local key frames and the location of local map points, so as to improve the local consistency of SLAM. In the loop closing thread, the loop detection and the loop optimization for key frames are executed, to enhance the global consistency of SLAM. In addition, according to the RGBD image and camera pose information, a complete and accurate 3D dense environment map is constructed through Octomapbased mapping framework. Experiments on TUM datasets show that the proposed method achieves the same accuracy as featurebased method with less time.

For the degradation problem of microwave signal phase noise caused by the additional phase noise of analog microwave photonic link, the characteristics of the white noise and frequency noise of microwave photonic link (MPL) were fully analyzed. The addictive phase noise (APN) model of MPL was established to expound the APN behaviors, and the correctness of the model was also verified by experiments. As for the typical applications of local oscillator optical distribution system, the EDFAassisted MPL can be used to improve the APN. Experimental results indicate that the proposed phase noise model can be effectively applied to the microwave photonic link and can provide theoretical basis for the analysis and optimization of additional phase noise.

Compared with the external cavity semiconductor laser, distributed feedback laser has a higher temperature regulation rate. In order to achieve saturated absorption frequency stabilization, it is necessary to control the laser temperature precisely. In this paper, the temperature control requirements of the saturated absorption frequency stabilization system is analyzed, a precise temperature control system is designed based on MAX1978 chip.The linear optimization of the temperature bridge circuit is realized by using constant current source, and the parameters of the analog PID circuit are quickly adjusted by using genetic algorithm. Finally the temperature control stability of 0.2mK is achieved, which is 1~2 orders of magnitude higher than that of the similar design. It solves the problem of obvious shaking of saturated absorption spectrum and has a broad application prospect.

In view of the fact that the current monitoring technology of track occupancy detection is easy to be influenced by electromagnetic interference and external environment, and it is difficult to realize large capacity and high density measurement, a new method based on strain sensing of weak fiber grating array is proposed. Firstly, the weak fiber grating array with reflectivity of 0.2‰ and the hybrid multiplexing technology are used to solve the problem of large capacity and high density detection. Secondly, based on the finite element model and wheelrail coupling theory, this paper deduces and analyzes the force on the track and the sensitivity of the sensor. Finally, the packaged fiber grating array is installed on the test site, and the track load experiment is carried out. The experimental results show that the system can accurately identify the axle information, running direction and track occupation information. The system has good reliability and high sensitivity, which provides a new idea for track occupation inspection technology.

Aiming at the problems of low spatial resolution and narrow field of view of infrared images, which leads to low image registration rate and poor realtime performance, an infrared panoramic mosaic algorithm based on the region of interest (ROI) of image is proposed. The method is built to run as follows: firstly, the ROI of the two adjacent infrared images is calculated according to their position relationship. Then, the SIFT feature points are extracted in the ROI and used as moving targets, thus the feature points in the infrared image to be registered are determined by combining the KLT tracking algorithm. Secondly, the RANSAC algorithm is used to eliminate mismatches. Finally, the pixel level fusion method is used to eliminate the stitching traces and an infrared panoramic image with stable resolution and wide field of view is synthesized. The experimental results show that compared with the traditional SIFT algorithm, the registration rate and the running time of the proposed algorithm is improved by 3.491% and nearly 50%, respectively, which can realize the seamless splicing of multiframe infrared images accurately and effectively.

In order to reduce assembly errors and solve problems such as forced assembly and incomplete assembly during the automatic assembly process, the optical fiber sensing calibration system was designed. The system consists of a fiberoptic sensor array, a demodulation module, and a tuning control mechanism. A solution model for strain distribution was proposed, and the function relationship between wavelength and position deviation was deduced. The distribution characteristics of the strain sensitive area of the workpiece were analyzed by simulation, and the mapping relationship between strain and position error was quantified. In the experiment, adjustable force structure was used to calibrate the system strain monitoring data, and visual measurement was used to calibrate the system position deviation. The experimental results show that the position deviation has a linear mapping relationship with the wavelength change. It is feasible to invert the deviation of the assembly position by the wavelength. Compared with the standard value of visual measurement, the average relative error of the position deviation of the system is 8.6%. The strain field of the workpiece can be inverted by this system in real time.

Aiming at the shortcomings of current image feature point face recognition algorithms such as low matching accuracy, a face recognition algorithm based on convolutional neural network is proposed. The algorithm uses traditional algorithm operators to fuse convolutional neural networks for recognition. First, the idea of local receptive field is used to segment the overall image to obtain a local image set. Each local image pixel in the set is stored in the pixel matrix Ai. Then the convolution operation is performed on each local image to obtain the intrinsic feature relationship between the local images, and it is stored in the Bi matrix and pooled for feature mapping. Finally, the network weighting coefficients are trained and the recognition results are obtained. Experimental results show that compared with other algorithms, the proposed algorithm improves the problem of low matching accuracy of image feature points of the original algorithm, and verifies the effectiveness of the proposed algorithm.

Existing unmanned aerial vehicle (UAV) detection mainly applies radar scanning as the main detection method and other sensors as the supplement, but the cost of radar is relatively high. As the device with low cost and low power, the laser has received widespread attention. In this paper, using the laser as a carrier and combining with the acquisition, pointing, tracking (APT) system in wireless optical communication, a UAV scanning positioning system based on laser detection technology was designed and implemented. The positioning accuracy of the system was tested by using a balloon test method. The scanning test results show that when the average power of the laser transmitter is less than 40mW, the target can be captured and positioned with a high accuracy with an error of lower than 5%, and the maximum detectable range is 45m.

In the RF fingerprint identification scheme for the constellation diagram, with respect to the problem of the low identification precision rate in low signal noise ratio (SNR) environment, a twodimensional identification algorithm based on the Euclidean distance and amplitude distance is proposed. This solution optimizes the constellation diagram to extract RF fingerprints with better identification performance from the optimized constellation diagram, and then uses the twodimensional identification algorithm to improve the identification precision rate. Simulation results show that compared with the method based on the Euclidean distance, the identification precision rate of the proposed twodimensional identification algorithm can be improved by up to 8% and the identification precision rate is 77.8% when the equipment capacity is 50. The RF fingerprints characterized by the optimized constellation diagram present better uniqueness and robustness.

In order to achieve discharge point location and electrical equipment fault diagnosis, UV imagers need to register UV and visible light images from different sensors. Due to the large difference between the imaging principles of ultraviolet and visible light images and the difficulty in feature extraction and matching, a mutual imagebased UV and visible image registration algorithm is proposed and applied to the UV imager for realtime registration. Firstly, the mutual information in the ultraviolet and visible light images is calculated. The mutual information is used as the similarity measure. The image is transformed by the rigid body transformation model. The image transformation parameters are obtained by the 1+1 evolutionary algorithm to obtain the spatial transformation parameters of the optimal similarity measure, and the effectiveness of the proposed algorithm is verified by simulation and UV imager.