With the development of information technology，optical communication becomes the main way of data communication. Optical switches are the core components in optical communication systems. Micro-electro-mechanical system（MEMS）has been widely used in the optical field with the features of miniaturization and integration，etc. MEMS optical switches not only retain the advantages of low loss and low crosstalk of traditional mechanical optical switches，but also have the advantages of small size，light weight，and short switching time. This paper introduces the principle of MEMS optical switch，describes the research progress of MEMS optical switch，and analyzes the advantages of MEMS optical switch. MEMS optical switches have characteristics such as fast response speed，high integration，low insertion loss，and good scalability，which have obvious advantages over traditional optical switches and show great potential for application.

Optical fiber fuse effect is the main cause of optical fiber laser path damage，and also an important bottleneck factor limiting the output power improvement of optical fiber laser. The optical fiber fuse damage effect in narrow line-width lasers at 1 560 nm band is studied in this paper. The conditions and causes of optical fiber fuse effect from the perspective of the mechanism of optical fiber fuse effect are analyzed. And in the output power range of 4~9.9 W，for the single-mode polarization-maintaining fiber with a mode field diameter of 10.5 m，the verification experiment is carried out to determine the power density threshold corresponding to the occurrence of fiber fuses 11.44 MW/cm2. A comprehensive method of hardware current limiting and temperature early warning to optimize heat dissipation is proposed，which reduces the probability of fiber fuse damage，and is conducive to improving the long-term stability of narrow line-width lasers working at high power.

Lidar technology based on high-energy narrow linewidth pulse lasers is widely used in the field of atmospheric remote sensing. An in-band core pumping pulsed fiber laser based on MOPA is introduced in this paper. The laser adopts a split design，in which the seed module can output laser pulses with a pulse energy of 1.8 J，and switch the center wavelength between 1 572.018 nm and 1 572.480 nm at a repetition rate of 1~100 Hz. Within the temperature range of 10~40 ℃，the center wavelength drift of the seed module output is not more than 0.58 pm，and the output pulse width is adjustable between 200~400 ns. The amplification module amplifies the output pulses of the seed module by using in-band pump. By suppressing the stimulated Brillouin scattering effect of the main amplification stage，high-energy outputs of 76.49 J at a pulse width of 200 ns and 135.7 J at a pulse width of 400 ns are achieved，with near diffraction limit spot and excellent spectral signal-to-noise ratio. The laser can be used as a source for differential absorption lidar and has practical significance in the field of remote sensing measurement of CO2 gas concentration.

The Fabry-Perot cavity（F-P cavity）is a widely utilized optical device with a multitude of applications，including those in astronomical detection，precision measurement，and sensor technology. At present，the conventional gas-gap Fabry-Perot cavity continues to be the preferred option for atmospheric detection and filtering applications. However，this type of cavity is subject to significant disadvantages，including a large volume and susceptibility to mechanical vibrations，which can have a detrimental impact on its performance. In contrast，the solid-gap Fabry-Perot cavity offers the potential for integration advantages that can effectively mitigate these issues. To guarantee the optimal filtering performance of the solid-gap Fabry-Perot cavity，a methodology for the precise determination of its cavity length is presented in this paper. By proactively regulating intrinsic hardware parameters，precise control of the filtering performance can be attained. This not only reduces the hardware size but also minimizes tuning-induced vibrations，thereby supporting the utilization of Fabry-Perot filter devices in spaceborne applications. A transmission spectral detection system is developed based on the aforementioned methodology，and the experimental tests are conducted on substrates. The results demonstrate that the system displays a fluctuation of 20 nm when measuring processing amounts at discrete points on the substrate. This method has successfully demonstrated initial precision in detecting the cavity length of the solid-gap Fabry-Perot cavity while confirming the feasibility of controlling its inherent parameters. Furthermore，the use of light feedback within an enhanced stabilization system enables the successful detection of an 8 mm thick substrate with clear transmission spectral lines，thereby highlighting the scheme's advantage in offering a broader detection range. This research establishes a fundamental basis for achieving precise control over filtering performance at the hardware level，which has the potential to facilitate accurate calibration of polishing removal quantities within existing polishing processes.

Advanced bearings are widely used in extreme conditions such as high temperature，high speed and high pressure environments. As a key component of bearings，the surface quality of advanced bearing rollers directly determines the performance and reliability of advanced equipment. Hence it is very important to carry out early defect inspection on bearing rollers surface. The working surface of the bearing roller contains cylinder，cone，drum，etc. In addition，roller material is metal and its complex refractive index property combined with a certain degree of surface roughness characteristics will lead to extremely complex optical scattering. So，it is difficult to obtain good contrast in images collected by conventional light field. In order to solve these problems，a far-field electromagnetic scattering simulation model of surface microscopic defects by finite element method is built in this paper. The far-field distribution based on the solution results of the near-field distribution of finite element is calculated by Fresnel diffraction integral formula. The defect scattering imaging under different illumination angles，defect sizes，defect types and other parameters is simulated. Results show that the intensity distribution of the far-field scattered light obtained by the proposed method is basically consistent with that obtained by the full-space finite element method. In the illumination incident angle range 0°~15°，coaxial field illumination can be taken for defect inspection. In the illumination incident angle range 30°~37°，dark field illumination can be taken for defect inspection. The results can provide a theoretical basis for multi-field scattering imaging system design and also help to improve the localization ability of China advanced bearings.

With the widespread application of optical lenses in industrial automation inspection，it is of significant importance to evaluate the imaging quality of lenses quickly and accurately. To meet this demand，modulation transfer function（MTF）measurement instruments have been developed and put into use. The primary measurement methods include the vertical slit method and the slanted edge method. However，the vertical slit method requires strict alignment of the slit with the sensor，which is difficult to achieve in practice. On the other hand，the slanted edge method involves differentiation during the measurement process，leading to unavoidable noise amplification，which affects the measurement results. To address these challenges，an improved inclined slit method is proposed in this paper. This method overcomes the noise amplification problem associated with the slanted edge method and is suitable for slanted slits，thereby enabling precise MTF measurement. The proposed method uses multi-frame image pixel averaging to remove random noise and obtains the slit image. It extracts the pure background region of the image to compute the pixel mean as the background noise value，and then subtracts this value from the image to remove background noise. Additionally，a new centroid calculation method is introduced for the slit image. First，the highest pixel value in a line or column of the line spread function（LSF）in the image is used as the center，and pixels to the left and right are traversed to determine the effective slit region. Data from this region is then processed using a cubic spline interpolation algorithm to obtain a sub-pixel level centroid position. With the centroid position of the current row or column as the center，discrete sampling of pixel positions is performed at 0.25 times the camera pixel size sub-pixel intervals and categorized into bins. The mean pixel values along the slit direction within the same bin are calculated to obtain a fourfold oversampled sub-pixel level line spread function. The improved slanted slit method not only accurately calculates the slant angle of the slit but also precisely measures the lens MTF. A detailed analysis of factors affecting MTF measurement accuracy is conducted，and the results show that，compared with advanced foreign instruments，the measurement error of MTF values is less than 1% within the Nyquist frequency before oversampling. After oversampling，the measurement error is less than 1.3% within 2.6 times the Nyquist frequency. The experimental results validate the effectiveness and practicality of the proposed method，demonstrating that it meets the needs of industrial production. This method effectively addresses the issues of high cost and limited control associated with purchasing advanced MTF testing instruments from abroad.

Accurate monitoring of particle size，shape，distribution，and composition is crucial across many industrial and scientific domains. However，existing particle analysis methods are affected by the refractive index and morphology of particles，and have limited measurement accuracy，and are difficult to handle irregular particles or high-throughput application. To overcome these limitations，an innovative approach for real-time，high-throughput classification of micron-sized particles is presented in this paper. Particles of various types are introduced into a custom-designed microfluidic channel，and images are captured by a polarization camera equipped with an on-chip array of four polarization filters. These filters allow the simultaneous registration of scattered intensities polarized at 0°，45°，90°，and 135°. From these，Stokes parameters，as well as the degree and angle of polarization，are calculated，facilitating the initial differentiation of particles. By fusing the most relevant polarization components，classification accuracy is further enhanced.

Optical fiber sensors for humidity and temperature measurement are extensively utilized in medical and biological applications due to their ease of integration，immunity to electromagnetic interference，and resistance to corrosion. Polyvinyl alcohol（PVA）is frequently employed as a sensing material for humidity owing to its distinctive properties. Consequently，we designed and simulated a high-sensitivity optical fiber hybrid plasmonic waveguide sensor that integrates the humidity sensitivity of PVA with the principles of surface plasmon resonance（SPR）. This sensor comprises PVA optical fibers connected at both ends to micro-nanofibers，with a metallic grating substrate positioned beneath the PVA optical fiber to enhance sensing sensitivity via surface plasmon effects. The experimental results indicate that the sensor exhibits a sensitivity of -1.749 1 nm/%RH within the 20% to 90% relative humidity range，while its temperature sensitivity is measured at -0.518 93 nm/℃ across temperatures from 20 ℃ to 90 ℃. It demonstrates a strong linear correlation between wavelength shifts and external parameters such as temperature and humidity.

To effectively monitor the inclination angle of transmission towers and ensure the safety and reliability of the power grid operation，a high-precision transmission tower inclination angle monitoring system based on an optoelectronic oscillator（OEO）is proposed and experimentally verified. First，the inclination angle sensor model is theoretically analyzed and established based on the optical fiber stretch length，mapping the inclination angle information to the oscillation frequency change of the OEO by using the relationship between the OEO oscillation frequency and the cavity length. Next，the oscillation frequency is divided by the frequency divider，and then the Field Programmable Gate Array（FPGA）is used to perform real-time calculations of the inclination angle parameters，achieving real-time monitoring of the transmission tower inclination angle. Finally，the feasibility of the theory is verified through experiments，achieving measurements within the range of 10°，with a monitoring precision of 0.011′，a resolution of 0.003 6″，and a system real-time processing response time as low as 50 ms. The feasibility and effectiveness of this system for real-time monitoring of transmission tower inclination angles are thus verified.

Raman hyperspectral imaging is an effective method for acquiring molecular spectral-image data cubes of the target being studied. To enhance the detection of weak spontaneous Raman scattering signals，a high-throughput Sagnac Fourier-transform Raman spectroscopic imaging technique is explored. This system does not use the traditional slit design and has a light throughput similar to imaging cameras，enabling a more efficient capture of faint spontaneous Raman scattering signals. Moreover，the utilization of the Sagnac lateral shearing interferometer exhibits polarization insensitivity and minimal dispersion effects，resulting in superior performance in spectral reconstruction via Fourier transformation. An experimental setup is built to evaluate the properties of Er∶YAG and Nd∶YAG ceramic materials. It clearly identifies their Raman characteristic peaks in the scattered spectra and produced sharp imaging outcomes.

Accurate segmentation of three-dimensional（3D）point cloud on the ice cream board surface is the basis of defect identification and detection for ice cream board. A 3D point cloud segmentation algorithm for ice cream board with various deep defect types is proposed. Firstly，the cross-section data of ice cream board surface is extracted，and the ice cream board types about warp or near-plane are identified by line fitting and the standard deviation criterion. Then，for these two kinds of targets，random sampling consistency（RANSAC）algorithm or iterative nearest point（ICP）registration algorithm are used to achieve point cloud segmentation. The experimental results show that four types of typical targets can be reliably and efficiently segmented by the method presented in this paper. The segmentation accuracy is higher than 94.6%，and the integrity is higher than 91.5%，which lays a foundation for subsequent defect indentification and quantitative detection. The proposed algorithm can effectively improve the segmentation accuracy of wood plate plane and has practical engineering application value.

s The clock error prediction of atomic clock plays an important role in the time and frequency control of atomic clock，which is related to the accuracy of time scale calculation and the stability of the punctual system. In order to further improve the long-term stability of clock group time and improve the accuracy of clock error prediction of atomic clocks，a deep learning network prediction model suitable for clock error prediction is constructed. The influence of the network model’s hyperparameters on clock error data is analyzed. And the best parameter index is given. That is，the number of hidden layers is 3，and the number of hidden layer neurons is 256. The results show that in terms of long-term prediction，the prediction effect of the optimized deep network model is 86.93% higher than that of the least squares model，and the deep network model is more suitable for high-precision time and frequency control，and has application prospects in the processing of clock error data in punctual systems.

This paper systematically explores the application of computerized data in sports detection equipment and speed training. By summarizing the current sports detection technology and equipment，the principles and development status of related technologies are sorted out. For the data acquisition and processing technology，this paper elaborates the key algorithms applied in motion detection，and optimizes the method of speed training based on this. The computer-aided training program proposed in this paper，based on the analysis of motion detection data，aims to improve the effectiveness and accuracy of speed training. Comprehensive research results show that the application of computer data analysis significantly improves the performance of motion detection equipment and rationalizes the training means，providing a scientific basis for technological innovation and practical application in related fields.

Stray light can seriously affect the imaging quality of star sensor optical systems. Existing methods for suppressing stray light often use baffle，which have limited ability to suppress small angle stray light and are relatively large in size. A method based on fiber optic panels is proposed to suppress stray light inside and outside the field of view in response to the above issues. Three systems are analyzed by using simulation software：a system with only a baffle，a system with only a fiber optic panel in front of the detector，and a system combining the fiber optic panel in front of the detector and the baffle. The point source transmittance of the three systems is compared and studied at different field angles. It is verified that the suppression ability of fiber optic panels on stray light at small angles（10°~20°）is superior to that of baffles. Under high angle incident light，the addition of fiber optic panels can effectively suppress stray light. Under the condition of a field of view angle of 25°，the point source transmittance decreases to below 3.8×10-5. Innovatively proposing the use of fiber optic panels instead of traditional baffles to suppress stray light in the system has important engineering significance for promoting the miniaturization of star sensor technology.

Aiming at the problems of high thickness，high cost and difficult support of 2 m grade glass-ceramics mirror，ANSYS finite element analysis software is used to carry out static analysis on the parametric mirror model，and the position and number of support points and layers of Whiffletree 18-point support，27-point support and 36-point support are compared. The relation between the deflection angle of the optical axis and the variation of the surface profile of the mirror is analyzed. According to the results of static analysis，the deflection angle of the 2 m class mirror should not exceed 10°. The support scheme adopts the combination of axial 36-point Whiffletree passive support and lateral 12-point tangential rod A-Frame support. Using this support scheme，the theoretical thickness of the mirror is reduced from 285 mm to 125 mm under the premise that the RMS of the surface type is ≤/50（=632.8 nm），which greatly reduces the production cost of the mirror and the system weight，indicating that the support structure has good engineering application ability.

With the rapid development of laser cooling technology，ultra-narrow linewidth lasers，and femtosecond optical frequency combs，the uncertainty and stability of laboratory optical clocks have now reached a level of small coefficients of 10-18. Miniaturized，integrated，and portable optical clocks have unique advantages in next-generation“second”definitions，gravitational potential measurements，elevation difference measurements，and solid earth tide measurements. In the process of moving from the laboratory to practical applications，the control system needs to complete miniaturization and integration design without compromising performance indicators. During the operation of the optical clock，the dead time caused by the control system delay affects the system stability indicators of the transportable optical clock. To reduce the impact of dead time on the stability of the optical clock，a hardware platform is built using Xilinx's Zynq architecture chip as the processing core and embedded Linux system as the software core. An embedded control system for transportable 40Ca+ ion optical clock is developed. Starting from the real-time interrupt requirements of the embedded control system of the 40Ca+ ion optical clock，this paper analyzes the impact of the newly developed hardware platform on the timing delay during the operation of the optical clock. Through analysis，it is found that the impact of the control system on system stability is below 2×10-16，which can meet the real-time requirements of the control system for transportable 40Ca+ ion optical clock.

In atomic interference based on two-photon Raman transitions，the relative angular deviation between the magnetic field of the quantization axis and the direction of the Raman light wave vector，as well as its spatial uniformity and temporal stability，are the key factors affecting the atomic interferometer. In this paper，a triaxial orthogonal Helmholtz coil is used to compensate for the angular deviation between the magnetic quantization axis direction of the atomic gravimeter and the Raman light wave vector in a dynamic environment，and the driving current of each axis coil and the magnetic field gradient in the interference region are calculated based on COMSOL simulation. The control strategy of simulated PID+ disturbance suppression is used to control the constant current source to drive the triaxial coil to improve the long-term stability of the magnetic field of the quantized axis. The experiments verify the feasibility of the three-axis Helmholtz coil to compensate the magnetic quantization axis，and realize the direction control of the magnetic quantization axis in the interference region and the magnetic field gradient with the perturbation not exceeding 3 mGs，which lays a technical foundation for the development of dynamic atomic gravimeter and has certain reference significance for dynamic atomic interferometer based on two-photon Raman transition.