The basic principles and system structure of typical DOFS technology were compared and summarized， and the development status of existing transmission line icing monitoring technology was introduced. A multi-dimensional parameter distributed optical fiber sensing technology was proposed that could realize the distributed real-time monitoring of state parameters， such as strain， temperature， vibration， and attenuation in the lines. In comparison with the existing single-parameter and dual-parameter sensing monitoring schemes， the accuracy of monitoring could be effectively improved， and the rate of misjudgment and false alarm could be reduced by using a multi-dimensional parameter scheme for fusion sensing of lines， providing important protection for the safe and stable operation of transmission lines.

A lightweight infrared target detection algorithm MEGI-YOLOv5 was proposed. The algorithm was based on the YOLOv5 model. Firstly， the backbone network was replaced with the lightweight Mobilenet-v3 network， and part of the CBL structure in the neck network was replaced with the deep separable convolution of the reciprocal residual structure. The C3 module was replaced by the combination of ordinary convolution and GhostConv to reduce the model parameters and calculation amount. Secondly， the Efficient Channel Attention （ECA） module was embedded in the neck network to improve the model's attention to the channel， so as to improve the model's feature extraction ability. The experimental results showed that compared with the YOLOv5 model， the number of parameters of the model was reduced by 22%， the detection speed was increased by 37%， and the detection accuracy of the model could reach 96.42%， which could meet the accuracy and real-time requirements of the identification of substation equipment categories and hotspots， and provide conditions for the subsequent timely detection of substation equipment faults.

A deep learning-based method was proposed， for defect classification and localization of photovoltaic panels to quickly and accurately determine the location and type of defects. To overcome the perspective limitations of traditional single-image defect detection methods， algorithms were adopted， such as image registration and stitching to generate high-resolution panoramic images of the photovoltaic panels. Deep learning techniques were then used to classify the infrared images of the photovoltaic panels and effectively identify the types of defects by comparing them with visible light images. The accuracy， precision， recall， and F1 score of the photovoltaic panel defect classification could reach 93.71%， 93.13%， 93.20%， and 93.11%， respectively. Compared with traditional methods， this approach had advantages， such as non-contact， high efficiency， and fast speed， making it suitable for detecting and locating defects in large-scale photovoltaic panels. It could provide accurate and comprehensive information about photovoltaic panel defects in a short time.

A new subpixel arrangement scheme called Checkerboard RB-G was proposed， which could further reduce the number of subpixels compared with most mainstream subpixel rendering schemes currently. Additionally， a subpixel rendering algorithm was proposed based on image binarization detection， which could solve the color deviation problem caused by Checkerboard RB-G. The test results showed that displays using the proposed scheme could achieve the same display resolution and similar display effects as those using traditional RGB stripe arrangement while reducing the number of sub pixels by 50%， significantly improving PPI and display quality.

In order to solve the problem of line selection in the analysis of hydrogen laser leakage detection spectral signal， a simulation model of direct absorption concentration inversion process of hydrogen laser absorption spectrum was established， and the reconstruction accuracy of spectral lines of three different linear functions was studied by combining Hitran database， and the influence of pressure change on the absorption coefficients of hydrogen absorption spectra under three linear functions was studied. The results showed that the reconstruction accuracy of the Gauss linear function was the highest among the three linear functions， and its reconstruction error was less than 0.02%. The maximum reconstruction error of Voigt linear function and Lorentz linear function were 0.37% and 2.58%， respectively， which could verify the accuracy and reliability of the simulation algorithm. The simulation results showed that the peak absorption coefficients of the three types of lines increase with the increased of pressure. In the range of 0.1~5.0 atm， the peak absorption coefficient of Lorentz linear was the largest， and the simulated peak absorption coefficient of Lorentz linear was the minimum 1.38 times and the maximum 17.77 times of Gauss linear. The minimum value was 2.07 times that of Voigt and the maximum value was 18.41 times that of Voigt. This study has laid a scientific basis for further research on hydrogen laser absorption spectrum detection， thus providing theoretical guidance and reference for practical application.

A blob detection algorithm based on multi-scale second-order anisotropic Gaussian directional derivative filter was proposed to address the shortcomings of current blob detection algorithms in locating and describing blob shapes. The second-order anisotropic Gaussian directional derivative （SOAGDD） could effectively extract image information from different directions and cope with noise interference. In this paper， a multi-scale second-order anisotropic Gaussian directional derivative filter was used to smooth the input image and calculate the regional extremum of the processed image. The experiments have demonstrated the superiority of our proposed method in the localization of interest point detection， shape description of detected interest points， and image matching.

Aiming at the practical problem of lacking quantitative guidance data on the performance of electro-optical system in complex environment， a method for evaluating the performance of electro-optical system based on the calculation of target and background optical characteristics was proposed. Considering the important influence of target and background optical characteristics， the calculation of target characteristics was introduced into the performance evaluation of electro-optical system. Based on the NvThermIP model that adopted the target task performance criterion， the calculation method of apparent contrast in the original model was modified. Using the radiation characteristic modeling technology， the calculation formula of apparent contrast based on radiance calculation was given， which could extend the applicability of apparent contrast calculation from infrared band to visible band. Thus， the NvThermIP model could be applied to electro-optical systems from infrared to visible wavelengths， and the accuracy of performance evaluation could be improved. Based on the measured data of the electro-optical system performance， the effectiveness of the method described was verified. The new method could achieve the goal of improving the accuracy of performance evaluation.

A new method for preparing high-quality perovskite films using dispersed nano-silica in chlorobenzene as the anti-solvent was introduced in this article. By dispersing nano-silica in the anti-solvent chlorobenzene， nucleation and crystal growth were controlled， ultimately enhancing the overall device performance of perovskite solar cells. The optimized perovskite films had uniform and flat surfaces with high densities， as well as the significantly extended carrier lifetime of 2 386.5 ns. The device had an increased built-in electric field of 1 143.2 mV， an optimum device photoelectric conversion efficiency of 20.6%， and an increased fill factor of 76.6%.

Copper indium gallium selenide （CIGS） solar cells have been extensively researched for their outstanding performance. Solution-based preparation techniques offer significant potential for large-scale industrialization. CuIn0.7Ga0.3S2 nanoparticles were synthesized at 200 ℃ via a hot injection method. The method maintained the quantity of Cu， In， and Ga of raw materials while precisely controlling the composition of Cu， In， Ga， and S elements by adjusting the volumes of surfactant oleylamine， solvent dibenzyl ether， and coordinating agent 1-octadecanethiol to 80 ml， 4 ml， and 15 ml， respectively. The atomic ratio of the synthesized nanoparticles was 1∶0.688∶0.299∶2.03， corresponding to CuIn0.688Ga0.299S2.03， with the nanoparticles' size ranging from 5 to 20 nm. The nanoparticles were purified and cleaned using organic solvents with different polarity， resulting in stable dispersion of the nanoparticles in toluene， as well as other more environmentally friendly solvents， such as o-xylene and cyclohexane. With these nanoparticle-based inks with different solvents， the copper indium gallium selenide sulfide （CIGSSe） absorber layers and their thin film solar cells were prepared.

By studying the signals and noise generated by the photoelectric detector and operational amplifier when the target passed through， the theoretical relationship between the signal-to-noise ratio and the laser output power was derived， as well as the retroreflection efficiency of the original reflection screen. Multiple sets of experiments were conducted to verify the influence of laser power and reflection efficiency on signal-to-noise ratio. The results indicated that， by appropriately increasing laser power and using a reflective screen with high reverse reflection efficiency， the signal-to-noise ratio of high-speed small target passing signals could be effectively improved in the designed laser screen system.

Uniformly dispersed SiO2 microspheres were prepared by the improved Stöber method. The effects of different environment temperatures and the dosage of TEOS on the particle size of the microspheres were discussed. The morphology and optical properties of the powder and film were characterized by scanning electron microscope， spectrophotometer， X-ray diffractometer and ultraviolet-near infrared absorption emission spectrometer. The results showed that， with the increase of temperature， the particle size of microspheres decreased； the particle size of microspheres increased with the increase of TEOS concentration； the self-assembly ambient temperature had a certain influence on the quality of photonic crystals， and the quality of silicon dioxide photonic crystal film was the best with the self-assembly temperature of 55 ℃. Under the excitation of 980 nm， the luminescence test was carried out on the up-conversion film，spin-coated on the glass substrate and photonic crystal， and the emission intensity of red light and green light was enhanced by 1.72 times and 1.93 times， respectively. This was because the photonic crystal surface with a band gap in the excitation light band had a high photon density of states， which could enhance the electromagnetic field strength around the luminescent particles， increase the probability of electronic transition， and thus enhance the fluorescence intensity.

As a system engineering project，the designs of the electromagnetic pulse protection for rugged LCDs were proposed at three levels， the component level，the module level and the equipment level. With all the designs of the electromagnetic pulse protection， one rugged LCD was verified through the electromagnetic pulse test. The designs of the electromagnetic pulse protection for the rugged LCD were proved effective by the conclusion of the test.The viability of the rugged LCD was improved in the electromagnetic pulse environment.

To address the challenges associated with the detection of low， slow， and small （LSS） targets， a low， slow， and small target detection system based on Field Programmable Gate Array （FPGA） has been developed. This system employed a COMS camera for image acquisition， DDR3 SDRAM for video frame buffering， and utilized FPGA to process algorithms combining inter-frame differencing and optical flow. The results of target detection were presented using the VGA standard. Test outcomes verified the system's capability to detect and real-time display low， slow， and small unmanned aerial vehicle （UAV） targets in hovering and low-speed flight postures.

The glue filling process has been widely used in oversize optical-bonding， but the lack of theoretical research on its process parameters makes the verification more complicated. In this paper， under the premise of reasonable assumptions， a theoretical calculation model was proposed based on the basic equations of fluid dynamics， and it was simulated by numerical analysis software. Finally， the optimal parameters of the process were obtained. The results showed that the designed theoretical calculation model was basically consistent with the actual flow pattern， and the glue filling time had an error of 10 s， which could meet the actual and practical accuracy requirements. The model could be used to explore the variation law of the glue filling process parameters， so as to quickly determine the best solution to process parameters.