In order to improve the laser output quality of a dual-wavelength annular cavity laser, a saturable absorber of erbium-doped fiber was added to the 3 dB ring mirror to form a filter ring, and the filter ring and the output coupler were cascaded to realize the laser output filtering. Theoretical analysis and experimental verification were carried out. The results show that the laser output bandwidth of 1554 nm and 1562 nm wavelengths are 0.029 nm and 0.038 nm, the peak fluctuations of dual wavelengths are 0.0048 mW and 0.0087 mW, and the signal-to-noise ratio is 55 dB, respectively. The results show that the dual-wavelength laser based on the saturable absorber and 3 dB ring mirror filter structure has better linewidth narrowing characteristics and stability. It can be used in the field of high capacity communication and laser therapy.

Aiming at the joint analysis of wind field and depolarization ratio in atmospheric remote sensing, a dual-polarization coherent detection method was proposed. Based on traditional coherent wind light detection and ringing(LiDAR), a detection system that can achieve diversity reception of orthogonal polarization signals was designed, and the calculation principle of atmospheric depolarization ratio under this system was derived. A systematic experiment was conducted to compare and observe the radial wind speed information of parallel polarized light and vertically polarized light under different weather conditions, and to describe the atmospheric state in combination with real-time atmospheric depolarization ratio parameters in two dimensions. The results show that when the laser emission energy is 100 J, the pulse width is 400 ns, and the distance resolution is 30 m, the radial wind speed range in cloudy, foggy, rainy and sunny days can reach 6.0 km, 5.5 km, 3.0 km and 3.0 km, respectively. The range of atmospheric depolarization ratio is 2.0 km, 2.0 km, 1.0 km, 1.0 km; the consistency of the dual-channel radial wind speed in the range does not change with the weather. This study provides a reference for the integrated detection of multiple atmospheric elements.

In order to meet the demand for fast steering mirrors (FSM) with 15 mm aperture in constrained space, an in-depth study of magneto-resistive electromagnetically driven FSM was conducted. Theoretical analysis and modelling simulations of magneto-resistive electromagnetic actuators have been carried out. The influence of parameters such as coil, permanent magnet parameters, magnet conductor, air gap and core on driving torque in electromagnetic drives was analyzed. According to the simulation optimization results, a FSM prototype was designed and experimentally tested. The results show that the external dimension of the FSM is 24.0 mm×33.5 mm, the angular stroke is 2-D ±1°, the working bandwidth is more than 350 Hz, the weight of the main body is 79 g, calibre 15.0 mm and the average power consumption is 7.36 W. The FSM prototype is small, lightweight, and has a large cornering range, which allows it to be applied to beam pointing in constrained spaces. The procedure disclosed is helpful for small magneto-resistive FSM for applications in space, load weight constraints are high for applications.

In order to achieve rapid and high-precision measurement of dissolved oxygen mass concentration using optical methods, a dissolved oxygen measurement method based on sampling-average to detect phase difference has been proposed. The multi-sampling average method was used to eliminate single-point error and improve measurement accuracy. Theoretical analysis and experimental validation were conducted. The measurement accuracy, measurement speed, and the influence of sampling frequency on the measurement error were studied. Comparative assessments were made against the performance of various existing methods, and a dissolved oxygen measurement system was set up to validate measurement results. The results indicate a clear advantage ofusing the sampling-average method to detect phase difference, with high measurement accuracy (measurement error is 1.13°, measurement standard deviation is 1.24°) and fast measurement speed. The relative measurement error of the dissolved oxygen measurement system is 2.84%, and the relative measurement standard deviation is 4.51%, achieving high-precision measurement of dissolved oxygen mass concentration. This research provides a simpler and more accurate phase difference detection method for high-precision dissolved oxygen measurement using optical methods.

In order to suppress the image shift of the circumferential scanning photoelectric reconnaissance system and reduce the false alarm rate, flexure support vibrating mirror was used for high-frequency image shift compensation. The composition and basic principle of the image shift compensation system were analyzed, and a flexure support structure was used instead of a bearing support structure. The transfer function model of the compensating vibrating mirror was given. A test system for the image shift compensation effect of the flexure support vibrating mirror, including the target, azimuthal turntable, image shift compensation vibrating mirror and other functional parts, was constructed. The results show that: The image shift compensation vibrating mirror swing sweep frequency is greater than 105 Hz, the uniform time is greater than 5.5 ms, the speed stability is greater than 99.60%, the detector imaging is clear, obviously inhibit the image shift, the flexible support vibrating mirror can meet the circumferential scanning photoelectronic reconnaissance system of high-frequency image shift compensation. This study is informative in the field of image shift compensation for circumferential scanning photoelectric reconnaissance system for high-frequency imaging detection.

In order to solve the problem of speckle image quality degradation caused by non-uniform lighting in the scene, which affects the testing accuracy of digital image correlation, the multi-scale bilateral filtering Retinex algorithm was used to perform optical correction on speckle images, achieving high-precision optical measurement of displacement and stress fields in non-uniform lighting scenes. Experiments were conducted to simulate sub-pixel displacement of speckle images and compared them with existing image optical correction algorithms, the corresponding speckle image displacement mean error, standard deviation, and quality evaluation index data were respectively obtained. The results show that the proposed algorithm has a good correction effect on non-uniform illumination speckle images, and the problems of image boundary blur and local overcorrection can be solved. The average grayscale gradient of the corrected speckle image increased from 12.986 to 14.574, and the displacement standard deviation decreased from 0.0443 to 0.0335. This method can effectively enhance the quality of speckle, ensure the testing accuracy and stability of digital image correlation methods in scenarios with uneven lighting, improve the testing accuracy of digital image correlation methods in complex lighting scenarios, and provide reference for non-contact measurement of lighting mutations in actual environments.

In order to solve the phase change caused by factors such as carrier phase delay, initial phase difference, and low-frequency environmental noise in traditional phase generation carrier(PGC) demodulation technology, as well as waveform distortion caused by phase modulation depth, a new improved demodulation algorithm based on traditional arctangent demodulation was proposed. The phase changes and modulation depth dependence caused by different factors such as carrier phase delay could be theoretically eliminated by using this algorithm, without the need of additional calculations and compensation. By using simulation software and constructing a Mach-Zehnder interferometer demodulation device, the waveform and performance indicators of the proposed improved algorithm were compared with the other two algorithms. The results show that the calculated signal-to-noise ratio of the new algorithm is as high as 71.51 dB, and the total harmonic distortion rate is as low as 0.0004501%. The improved algorithm is insensitive to phase changes and modulation depth caused by factors such as carrier phase delay, initial phase difference, and low-frequency environmental noise. This study provides a reference for improving the accuracy of phase generation carrier demodulation technology and improving system performance.

In order to obtain a green laser with high conversion rate and narrow pulse width, the pumping source of the laser consists of two 80 W laser diodes with a center wavelength of 878 nm and a line width of 1 nm was adopted. Two Nd∶YVO4 were pumped end-face of the laser. The diaphragm mode selection and acousto-optic Q crystal modulate the YVO4 crystal, and the LiB3O5(LBO) crystal doubles the frequency before exporting the light with the 1064 nm fundamental frequency out of the cavity. A stable 532 nm green laser with a stable output was achieved when the flat convex unstable cavity regulates the crystal’s thermal lens effect. The results show that, under operating conditions between 80 kHz and 200 kHz, the green laser’s repetition frequency is 100 kHz, its pump power is both 70 W, its output power is 36.7 W, its pulse width is 7.3 ns, its light conversion efficiency is approximately 26.2%, its spot roundness and focus spot roundness are both greater than 90% and 97%, and its pulse stability is less than 2%, respectively. It has been verified that the parameter green laser has a better performance than other lasers of the same type in glass drilling, with the edge collapse range controlled within 100 m. This study is significantly helpful in optimizing the process effect of nanosecond laser products and expanding the application scope of products.

There exist the problems of short range measurement, noise interference, and influence of nonlinear frequency modulation by using frequency modulation continuous wave (FMCW) laser ranging technique based on beat frequency algorithm. In order to solve these problems under the same institutional condition, the ranging method based on matched filtering was proposed in this paper. Novel method modifies The optical path was modified, such as adding signal reversal unit and signal conjugate unit, replacing the mixer with a convolver. At same time, the system response and transmission waveform matching was satisfied, and the simulation experiment was carried out in this paper. The result of simulation show that ranging by matched filtering can get twice detection distance and 9.5 dB higher signal-to-noise ratio(SNR) than ranging by beat frequency algorithm, and spectral resolution from matched filtering were higher than that from beat frequency algorithm in the condition of nonlinear frequency modulation error were 19.8 MHz, 198 MHz and 396 MHz, reveal the ability of reduce the influence of nonlinear frequency modulation. This study can provide technical guidance and new ideas for researchers in this field.

In order to study the scattering characteristics of the incident beam by an infinite cylinder with arbitrary regular cross section, the semi-analytical method was proposed. Appropriate cylindrical vector wave function was used to expand the scattered field and internal field, and the expansion coefficient was determined by electromagnetic field boundary conditions and projection method. The normalized near-field intensity distribution was numerically simulated by the example of an infinite cylinder with circular, elliptical and rectangular cross-sections irradiated by a fundamental mode Gaussian beam and a radial annular beam. The results show that, there is obvious interference phenomenon after the beam propagates through the cylinder and elliptic cylinder, and the rectangular cylinder has a certain convergence effect on the beam. This study provides a convenient semi-analytical solution for solving the scattering of arbitrary beams by an infinite cylinder with arbitrary regular cross section.

In order to address the potential deviation in rank function approximation caused by the kernel function used in the latent low-rank representation(LatLRR) method, an approach based on Schatten-p norm and latent low-rank decomposition was proposed. Theoretical analysis and experimental validation were conducted using this method. The images were first denoised using a median filtering method. The images were decomposed into low-rank and salient parts using the Schatten-p norm-based latent low-rank decomposition method combined with LatLRR. Then, an arithmetic mean strategy was employed to fuse the low-rank parts of the infrared and visible light images, while a summation strategy was used to fuse their salient parts. Finally, a summation strategy was applied to fuse the already merged low-rank and salient parts, resulting in fused infrared and visible light images with clear texture information and prominent thermal fault information. Through qualitative and quantitative experimental analysis, a p-value of 0.6 was determined to achieve the optimal fusion effect, and the proposed method outperformed seven other algorithms in fusion performance comparison. Through this approach, rich structural information at both global and local levels in infrared and visible light source images of power systems can be effectively captured.

In order to simplify the process, and to improve the modulation bandwidth and data transmission rate of surface emitting lasers, a vertical cavity surface emitting laser (VCSEL) with an emission wavelength of 850 nm was prepared using methods such as multi quantum well active region design, medium planarization process preparation, and oxidation constraint limitation. Low dielectric constant polyimide (PI) was used for planarization in the process, instead of the commonly used benzocyclobutene on high-speed VCSEL chips. The parasitic response of high-speed devices based on PI glue planarization process was studied. The experimental results indicate that, at low bias current, the VCSEL with 3 m oxide aperture shows a maximum modulation bandwidth of 25.2 GHz, resonant frequency of 24.3 GHz, parasitic cutoff frequency of 13.2 GHz, modulation current efficiency factor of 22.726 GHz/mA1/2, D factor of 19.670 GHz/mA1/2, threshold current and differential resistance of 0.27 mA and 215.965 , respectively, and maximum optical output power of 0.9 mW. The PI scheme can be used to modulate products with a bandwidth of 25 GHz and below. This result has certain guiding significance for simplifying the process and improving the data transmission rate of surface emitting lasers.

In order to solve the problem of low illumination and signal-to-noise ratio of hyperspectral (multispectral) images under short exposure conditions, a hyperspectral (multispectral) image enhancement model was proposed based on the improved residual U-Net (Res-UNet). The Res-UNet was taken as the backbone network, atrous spatial pyramid pooling and coordinate attention were used to enhance the feature aggregation ability of the model, and the Z-score loss function was introduced to improve the reconstruction ability of the model on spectral curves. The performance of the proposed model was evaluated using image enhancement quality metrics and a downstream task (drought leaf segmentation accuracy). The peak signal-to-noise ratio, structural similarity and spectral angular mapping of the improved model reach 0.9852, 39.71 and 3.120, which are better than the comparative algorithms. And its segmentation accuracy for arid leaves is higher than that of various mainstream algorithms. The result shows the effectiveness of this model for plant low-light multispectral image enhancement, which can provide information support for various spectral image downstream tasks.

In order to achieve high repeatability and resolution of a 2-D ultra large stroke fast steering mirror(FSM) that integrates high-speed scanning and precise pointing functions, the 2-D segmental calibration was studied. The angle calibration system of 2-D ultra large strok FSM was constructed. The polynomial fitting method was optimized to obtain the segmental nonlinear fitting method, the calibration model was theoretically analyzed and experimentally tested. The results show that with the range of ±5°, the calibration error of the traditional polynomial fitting method is merely 1.7 mrad, while the fitting error of the optimized segmental nonlinear fitting method is less than 8.2 rad, and the resolution of x and y axes can reach 4.6 rad and 5.6 rad respectively, and the repeatability of x and y axes is 19.4 rad and 9.6 rad respectively. This result plays an important role in the accurate control of 2-D ultra large stroke FSM.

In order to realize a new and simple magneto-optical trap(MOT) atomic capture, a combination of large aperture single beam and multiple mirrors was used to simplify the magneto-optical trap structure, and permanent magnets were used to provide the required gradient magnetic field to increase the portability of cold atom capture devices. The polarized light in the chamber and the composition of the experimental light path were analyzed and verified by experiments. The number of atoms in different dispenser currents and the change of the group lifetime after the current is turned off were measured by the methods of collection fluorescence, release and recapture. The results show that the size of the trapped 87Rb group is about 1.5 mm, the number of cold atoms is about 1.7×107, and the temperature of the group is about 177 K, respectively. The structure of the magneto-optical trap system can achieve ideal atomic trapping on the basis of miniaturization, and the research results provide a reference for the field of new quantum sensors.

In order to improve the maximum gain saturation energy of similariton laser, an optimized laser system was designed by replacing single-mode fibers(SMF) with dispersion decreasing fiber (DDF) in similariton laser ac-cording to the advantage of the strong resistance to optical wave breaking while similariton generated in dispersion decreasing fibers. Theoretical analysis and simulation verification were both carried out. The result shows that the maximum gain saturation energy that the modified laser can tolerate is increased by 33.3%. Replacing SMF with DDF for pulse evolution can effectively improve the maximum gain saturation energy of the laser. With the larger gain saturation energy, the output pulse energy of the laser is increased by 109.2%, the pulse peak power increased by 20.6%, the pulse width shortened by 23.1%, and the self-similar evaluation factor decreases from 0.077 to 0.065, resulting in achieving a high-quality similariton with a larger pulse energy, higher peak power, narrower pulse width. This study has certain reference value for obtaining high-power self-similar pulses.

Light detection and ringing (LiDAR) point clouds exhibit significant discreteness, with information density much lower than that of images, resulting in significant challenges for pedestrian detection in this background and exhibiting low detection accuracy. To address the above issues, a LiDAR pedestrian detection method was proposed based on improved complex you only look once (complex-YOLO). Firstly, a new network feature extraction structure was designed to enhance the feature learning ability of the backbone network, in order to more fully mine information in situations of data scarcity. In addition, the designed cascaded feature pyramid network and its multi branch linear fusion method can fuse features of different scales and depths, improve the generalization performance of the network, and cope with data feature distortion. During the training process, enhanced intersection over union (EIoU) was adopted to accelerate the convergence speed of the network. Through experimental verification on the dataset, the detection accuracy of the improved algorithm has been improved by 3.03%. This study is helpful for improving pedestrian detection accuracy in sparse data situations.

In order to realize the integral forming of tungsten carbide WC-12Co metal powder products, the selective laser melting (SLM) technology was used for the theoretical analysis and experimental verification of the single-pass melting pool forming of WC-12Co alloy powder. The results show that: Different combinations of laser scanning rate and power affect the continuity and morphology of the single pass molten pool. When the laser power is 340 W, the scanning rate is 600 mm/s, the scanning interval is 60 m/pass, and the powder thickness is 40 m/layer, the single pass molten pool morphology reaches the best state, presenting a stable and continuous “fishscale” morphology. The single layer cracks of the forming object are the least. In the process of single pass forming, W element aggregation occurs in the section of single pass melting pool, which is an important reason for the anisotropy of formed WC-Co block. This research result plays a decisive role in the forming of cemented carbide WC-12Co metal powder products.

In order to solve the problems of long time and low automation for the coupling assemble of optical fiber collimator in production of wavelength division multiplexer, an optical fiber collimator automatic coupling device with five-axis based on Hough circle detection was designed. The five-axis adjusting platform has X, Y and Z moving three degrees of freedom and R and J two rotating degrees of freedom, which was used to drive the U-shaped vacuum suction nozzle for stable suction and pose adjustment of collimator. A vision system was constructed by using semi-transparent infrared light card and camera module. The center coordinates of the spot on the optical card were obtained through Hough circle detection. Furthermore, the collimator was moved to the same axis of the center of the light spot through five-axis adjustment platform to achieve coarse alignment of collimator pose. The collimator was accurately tuned in R and J degrees of freedom to obtain the optimal coupling pose according to the coupled optical power measured by optical power meter for achieving precise coupling of the fiber collimator. The prototype test demonstrate that the efficiency of the automatic coupling device is 6 times higher than that of manual assemble, and the coupling efficiency of optical fiber collimator is larger 75%, which meets the quality requirements of the product. The research provides a reference for efficient automatic coupling of optical fiber collimator.

In order to effectively improve the accuracy of fiber optic signal recognition, a signal event type discrimination method based on multi classification Gaussian support vector machine (SVM) was adopted. Firstly, the Hanning window convolution method and the principle of 95% energy were used to identify the information of the beginning and end stages of event occurrence. Then, from the perspectives of time domain, frequency domain, and scale domain, the mean and discreteness of various normalized feature parameters were analyzed, and appropriate main feature parameters were selected. Finally, the multi classification Gaussian SVM algorithm was used to classify and recognize three different event types. Theoretical analysis and experimental verification were conducted, and data on different types of fiber optic event signals were obtained. The results showed that pattern recognition of event types in 30 sets of experimental data achieved an accuracy rate of over 96%. This method process meets the high-precision identification requirements of event signals in fiber optic sensing and provides important reference value for the application of fiber optic sensors.

In order to analyze the effect of removal function on machining accuracy and the anti-intermediate frequency error performance of removal function in magnetorheological finishing, the error suppression ability of magnetorheological finishing was evaluated by using the smoothing spectrum method. The smooth spectrum method and the traditional power spectrum method were theoretically analyzed, simulated, and verified by the aspherical optical element. The suppression ability of the removal function on the surface error and the mid-spatial error of the aspherical optical element were analyzed by the smooth spectrum method and the traditional power spectrum method. The results show that error suppression ability of smoothing spectrum method is 0.22 mm-1. The smoothing spectrum function can intuitively judge the suppression ability of the surface error and mid-spatial error of the removing function, and the variation rule of mid-spatial error can be clearly quantified and judged. This method provides a solution to the problem of error suppression evaluation in magnetorheological finishing.

In order to investigate the processing quality of carbon fiber reinforced plastic(CFRP) using a CO2 pulse laser, a single-factor experimental method was employed for theoretical analysis and experimental verification. The influence of scanning speed and laser power on processing quality was determined, and circular pattern carving experiments were conducted on CFRP. The results show that the material undergoes penetration at a laser power of 50 W and a scanning speed of 60 mm/s, and as the scanning speed decreases, the material absorbs more energy, resulting in complete pattern penetration. When the scanning speed is set to 100 mm/s and the power varies within the range of 10 W~5 W, white stripes gradually appear on the surface of the material. The surface roughness of the carved material was measured using both horizontal and vertical methods. Both methods indicated that the roughness values showed a pattern of first decreasing and then increasing. When the laser power was 7 W, the roughness reached its minimum value of about 7.030 m; After ultrasonic cleaning, when the laser power is 9 W and the scanning speed is 20 mm/s, the roughness of the carved material reaches its maximum value and there is less surface residue. At this time, the carving effect is better. This research result can provide certain guidance for the process of pulse laser processing of carbon fiber composite materials.

To ensure the safe takeoff and landing of aircraft and accurately monitor the visibility distribution of airports on a large scale, the error analysis methodwas adopted, the light detection and ranging(LiDAR) observation data was compared with the manual prevailing visibility data and the weighted visibility data calculated by the self-viewing system of Shenzhen Baoan International Airport from 2022-01 to 2023-06. The accuracy and reliability of prevailing visibility measured by visibility LiDAR were preliminarily verified. The results show that the average relative deviations between LiDAR data and manual data and self-viewing system observation data are -5.30% and -5.45%, respectively. LiDAR is more sensitive to early changes in visibility over a wide range. When rainfall and the prevailing visibility is less than 1500 m, the error of the observation results of LiDAR is large, and manual observation is required. When the prevailing visibility is greater than 1500 m, LiDAR can replace the manual output of the prevailing visibility data. This study provides a basis and reference for airport business operation based on the prevailing visibility measurement of visibility LiDAR.