Bandwidth of adaptive optics (AO) system based on stochastic parallel gradient descent (SPGD) algorithm is analyzed theoretically and experimentally. The relation between convergence speed and control channel number of AO system based on SPGD algorithm is researched. Then the bandwidth of AO system based on SPGD is analyzed theoretically by comparson with the traditional AO system with the Hartmann-Shack (H-S) wavefront sensor. A dynamical wavefront distortion correction experiment is performed, and the experimental results verify the theoretical analysis results well. This research can provide a basis for the design and use of AO system based on SPGD algorithm.

According to the high bandwidth, low packet-loss ratio, high channel utility requirements of satellite optical switching network, an on-board assembly algorithm based on mixed threshold and round-robin scheme is proposed combining the ground optical assembly algorithm. The algorithm caches the IP packet according to the parity, and assembles the packet according to the mixed threshold and round-robin scheme. Through the scheme, the padding ratio of burst is lowered. From the results of analysis and emulation, it is found that the complexity is the same as that of the round-robin method; the padding ratio of burst data packer is less than 10-4 when the load is 0.65, and the packet loss rate is 10-6 when the load is 0.5. This algorithm meets the requirements of optical satellite communications.

The influence of various types of index profiles on the characteristics of power transmission, including the effective mode area (Aeff) and the maximum power density (Dmax) in the cross section, is investigated. A unique expression for the index profile, which is appropriate to the most practical index distributions of optical fibers, is employed in the calculation of Aeff and Dmax with the comparison of the step index profile. The simulation result indicates that under the condition of equal transmission power and the difference of the mode effective index (Δneff) larger than 10-4, the index profile with a center dip will effectively lead to the reduction of Dmax and the increase of Aeff. This research provides theoretical basis for the design and fabrication of passive large core fiber for large power transmission.

An all-optical XOR logic gate for phase-modulated (PM) or amplitude-modulated (AM) signals is proposed. It achieves XOR operation by direct interference of light. Through simulation, this scheme is proved to be effective for XOR operation of PM or AM signals at 10 Gb/s. When the inputs are AM signals of which the duty cycle is less than 0.4 and the peak powers are equal, the Q factor of XOR output is 103. When the inputs are PM signals with π phase modulation depth, the Q factor of XOR output is 272.

Based on controlling the polarizations of the pump and Stokes beams in an optical fiber, a simple scheme for enhancing the stability of stimulated-Brillouin-scattering (SBS) slow light is proposed. By fusing a polarization-maintaining circulator (its fast axis is blocked) to each terminal of a polarization-maintaining fiber (PMF), we construct a slow-light element which ensures the two beams always having identical polarizations (along the slow axis of the PMF). Meanwhile, an experimental setup is established to study the slow-light performance. A tunable time delay of 0.87 ns/dB is demonstrated for a Gaussian pulse with a pulse width of 38 ns. The gain factor of the Brillouin gain in this element enhances three times as compared with that in the single-mode fiber. The experiment indicates that such a polarization-managed scheme significantly improves the stability of SBS interaction and thus that of the slow-light time delay.

The quality of the reference wavefront is constrained by the diameter of pinhole, which is the critical element that determines the precision of phase-shifting point diffraction interferometer (PS/PDI). The effects of film thickness, diameter of the pinhole, machining error and single and multiple aberrations of incident light on the diffraction wavefront errors are analyzed based on the vector diffraction theory. The analysis shows that in order to obtain a reference wavefront with a numerical aperture of 0.2 and a root-mean-square (RMS) wavefront error below 1.4×10-3λ, the best choice for the pinhole is the material chromium with a thickness of 200 nm and diameter of 1.5 μm, and the simulation error requirements should be satisfied in machining. The test window size is optimized based on the spatial frequency domain analysis. The results reveal that the optimal window size is approximately 60 μm. Experiments are carried out for the designed mask to verify the feasibility.

561 nm laser output is achieved from LBO intra-cavity frequency doubling of a laser in which NdYAG is end-pumped by LD. Using the 2 mm×2 mm×10 mm LBO crystal with type-Ⅰ phase matching cutting, the 561 nm laser output power is 123 mW when the pump power is 5 W. The conversion efficiency of 2.46% is obtained. The frequency-doubled light waves of 556 nm and 558 nm easily appear in experiment. From the influence of nonlinear conversion efficiency on fundamental-frequency light oscillation, the reason of 1112 nm and 1116 nm line oscillation is analyzed. As a contrast, we carry out the same experiment using a 2 mm×2 mm×8 mm KTP crystal with type-Ⅱ phase matching cutting. Results show that the laser is easy to realized 561 nm single spectral-line laser. The experimental results and theoretical analysis are consistent with each other.

According to the Fraunhofer diffraction theory, a theoretical model of far-field intensity distribution of two-dimensional M×N laser coherent array based on phase modulation is established. Numerical simulation on far-field intensity distribution of a 5×5 laser coherent array aiming at the application practice is made and the influence of different phases on intensity distribution is also analyzed. The results show that the positions of the main maximum or minimum and the value of the intensity distribution change with the phase modulation, and the intensity distribution varies with different array structures. More symmetric main maxima and secondary maxima appear in the simulation results when the phase difference loaded in each column and each row is π. All of these provide a helpful reference for the control of far-field intensity distribution by phase modulation.

A method based on thermo-elastic excitation of ultrasonic waves generated by laser at the surface of a cylinder is presented. The propagation rules of head waves at the aluminum cylinders with different radii or different thicknesses are researched. The propagation characteristics of head waves are analyzed. The numerical results show that head waves are generated at the certain range (critical angle) in the body, and then spread along the straight line at a speed slightly larger than that of the Rayleigh waves. In the propagation process, the head waves move away from the Rayleigh waves and approximate longitudinal waves. The sine values of the critical angle (at the maximum amplitude) are almost equal to the wave-velocity ratios of the Rayleigh wave and longitudinal-wave in the cylinder. At the interface, a part of the head wave escapes from the body and spreads along the circumference with a rapid attenuation, while the other part of the head wave has a reflection at the interface with a small attenuation.

A new method for calculating the beam radius distribution of the input Gaussian beam and the laser mode volume in the laser rod of a solid-state laser amplifier is reported. By this method, we investigate how the strong thermal lens effects in the rod of amplifier and the waist radius and mode of the laser beam inputting to the rod of amplifier influence the input laser beam distributions and input beam mode volume in the rod of amplifier. It is shows that the expanded and collimated laser beam inputting to the rod of amplifier will be strongly reflected to the axis of the rod by the strong thermal lens effects in the rod, because the small far-field divergence angle of the collimated input laser beam results in the signigicant decrease of the mode volume of input laser beam and the overlap efficiency of amplifier drops. It can obviously increase the mode volume of input laser beam in the rod of the amplifier to input the laser beam with larger angle, which can counteract the reflecting effects on the input laser beam. Moreover, the beam waist radius of the multi-mode laser beam inputting to the amplifier should be larger than that of the fundamental-mode input laser beam, since its far-field divergence angle is larger than that of the fundamental-mode laser beam with the same beam waist radius.

In order to fabricate an upper or lower chord blank of the C919 plane, different scanning paths are set with different laser solid forming (LSF) process parameters based on style-database of LSF. A test region is calculated using offset algorithm of the scanning path. “Similar scanning path” is automatically found by using parameters of the above test region and “hit probability” and a same style number is set for the next path. A numerical control operation file is automatically generated in terms of the style of every scanning path. What is concerned is how to set different process parameters for different scanning paths and transfer the style of the scanning path. Using this method, the operational complexity of process designer is reduced, and human factors are excluded during the manufacturing process. Processing design and manufacturing operation are separated and airworthiness certification requirements are satisfied.

For the laser-induced breakdown spectroscopy (LIBS) for steel liquid online quantitative analysis, due to the influence of matrix effect, the precision is seriously affected, therefore in the quantitative analysis we should try to reduce the influence of the matrix effect on the elements under tests. Internal standard method is the most common method of LIBS quantitative analysis. This method can partially eliminate the matrix effect in measurement process, but it can only be applied to samples with a simple elemental composition, and for complex samples it cannot effectively correct the matrix effect. By adopting the general matrix correction method, real-time online quantitative analysis of molten steel cannot be realized and the following data processing will be difficult, so this method is not suitable for field tests. We use the improved linear regression equation in the matrix correction method in combination with the Mathematica brief calculation program. It can effectively work for the matrix correction of LIBS steel liquid online quantitative analysis. Experiment verifies the feasibility of this method.

Electroluminescence (EL) defects and different process influencing factors of mono-crystalline silicon solar cells are analyzed by experimental comparison method. The open circuit voltage, short circuit current, fill factor and photoelectric conversion efficiency of the solar cells were tested, and the EL defect proportion of the solar cells is collected. It is found that the paste, sintering process, and the silicon wafer are the main causes of the EL defects. The quality of paste, the parameters of screen-printing, the stability of sintering process and the anti-reflection and passivation effect of coating process have some effects on the photoelectric conversion efficiency of solar cells. It is suggested that improving the sintering process stability, using better performance of paste and silicon wafer and optimizing the parameters of screen-printing and the flow ratios of ammonia and silane can improve the EL defect and promote the photoelectric conversion efficiency and qualified rate of solar cells.

About 5000 pieces of low-efficiency defective crystalline silicon (C-Si) solar cells are collected, inspected and analyzed, and a defect inspection and classification evaluation system for C-Si solar cells is established. The system which includes current-voltage (I-V) test, thermal imaging test, electroluminescence imaging test etc., can identify 16 kinds of defects. The system can give out detailed description and evaluation of these defects following the steps of "defect definition, testing feature, influience on performance, original mechanism, preventive action, value of repairing". The result will be helpful in improving C-Si solar cell production and repairing defective cells.

A finite element model of optical glass aspherical lens is constructed with the nonlinear finite element software MSC.Marc. The general Maxwell model is chosen as the viscoelasticity model and the boundary conditions are set, the heating and pressing processes are simulated, and their temperature and stress are observed and analyzed. The equivalent stress at the touching point is observed, and the result is that it increases when the curvature increases, and decreases when the curvature decreases. Pressing processes at 4 different temperatures are simulated, and their stress and strain effects are compared. The result shows that the equivalent stress decreases when the temperature increases, and the equivalent total strain increases when the temperature increases. 580 ℃ is the optimized temperature. Stresses at 3 different molding velocities are simulated, and the result shows that the equivalent stress increases when the velocity increases. But high temperature or low velocity causes unnecessary waste. 0.1 mm/s is the optimized velocity.

Plastic rectangular phase grating wavelength division multiplexing (WDM) device can effectively use plastic optical fiber bandwidth resources and improve the effective bandwidth of metro access network. The formula of the negative first-order diffraction efficiency is derived by using rigorous coupled-wave analysis in the optimum incident angle and the grating effective modal index varying with grating period is numerically simulated. Simulation results indicate that when the rectangular phase grating with refractive index of 1.58 and duty cycle of 0.5 is illuminated by 1.5 μm light, the negative first-order diffraction efficiencies for both TE and TM polarized waves can reach the maximum simultaneously under the conditions of the optimum grating period of 0.82 μm and the minimum groove depth of 1.47 μm. The normalized diffraction optical intensity is measured for different incident angles. The experiment result shows that, with the optimum incidence, the polarization dependence loss of the plastic rectangular phase grating can be lower than 0.5 dB.

According to the light distribution of LED, we figure out the formula of a dot illuminance on the plane of the edge-lighting backlight light guide plate (LGP), and an arrangement formula of scattering dots in the LGP as well. The main purpose of this paper is, by studying the shape of the scattering dots in the LGP, to find out the scattering dots shape which is the most suitable for the LGP. Based on the hemispherical dots shape, according to principle that the light coupled into the LGP in two directions is reflected by the hemispherical scattering dots and just outputs from the LGP, it is obtained that the optimum shape of scattering netted dots is the cone whose apex angle between 90° and 132.2°. Under the condition of a certain uniformity of output light flux, by means of TracePro software simulation, the total output light flux is compared for the LGP with different apex angles of cone scattering dots. The experimental and theoretical results match well with each other.

In this paper based on generalizing of two-mode squeezed operator, we construct a new three-mode squeezed operator. Further, the three-mode squeezed partical-number state is proposed by the technique of integration with an ordered product of operators. Its squeezing and antibunching effects are analyzed. The results indicate that its squeezing is weakened with the increase of photon number, and the antibunching effect of a1-mode field is also weakened with the increase of photon number; on the other hand, its a2-mode field and a3-mode field both exhibit antibunching effects.

Compared with the traditional uniformly polarized light, the radially polarized laser performs perfect spatial axial symmetry and its light intensity configuration is always zero along the central axis. Based on reviewing recent work, its optical properties, generation methods and newest applications are summarized. In accordance with this methods, we summarize that using the diffraction grating mirror to generate radially polarized laser is the most valued method of generating high-quality and high-power radially polarized laser.

This paper first provides an overview about the past 10 years of crystalline silicon solar-cell market development in detail and clarifies that the crystalline silicon solar cell plays a very important role in photovoltaic power generation field. Then various factors affecting the cost and efficiency of crystalline silicon cell module are considered, and the various factors influencing the crystalline silicon solar-cell power generation cost are analyzed in detail. For the crystalline silicon solar-cell module production process, although the package cost accounts for the largest proportion, the reduction in cost of module mainly depends on the improvement of the technical level, such as silicon ingot growth, slicing and cell manufacture. For the production process of the basic elements, reducing the raw materials cost for module manufacture and improving the processing technology continuously will be the basic approach to reduce crystalline silicon-cell cost. Finally, combined with the new technical reports about crystalline silicon solar cell, from the perspective of production process and efficiency-enhancing techniques, the development trends of crystalline silicon-cell module manufacturing technology in the future are prospected.

Chaotic photonic reservoir computing is a new information processing technique, which employs chaotic laser as the reservoir. The advantages of the new technique are of high processing speed, big computing capacity and simple physical implementation. This new kind of computing can be applied in the future photonic computer, intelligent information processing and other fields. This paper introduces the concept, principles, processes and implementation of chaotic photonic reservoir computing in detail. The advantages and disadvantages of different implementation schemes are compared. Some issues to be resolved for the chaotic photonic reservoir computing are listed. The trends of chaotic photonic reservoir computing are also demonstrated.

The main role of the pinhole mask of point diffraction interferometer is to produce a nearly ideal spherical wave for interferometry through diffraction. The quality of the reference wavefront depends on the pinhole diameter, roundness and three-dimensional surface profile. The structure and principle of pinholes are introduced, and the classification and comparison of the pinhole diffraction electromagnetic field simulation technology are made. The pinhole mask processing technologies are summarized, and the machining mechanism, precision and technical features of focused-ion-beam etching and electron-beam lithography are expounded. The influence of mask alignment accuracy on measurement repeatability is pointed out. The different testing approaches and the main technical problems are analyzed and the trend of the pinhole three-dimensional topographical measurement technology is described.

The optics contamination and the contamination control strategies used in extreme ultra-violet lithography (EUVL) directly affect the performance of optics system. We summarize the main achievements in the research of optics contamination, such as carbon deposition and oxidation of optical surface, and contamination control strategies which include intellegent gas blending, protective capping layer and contamination cleaning. The challenges and key technologies are also presented.

Wide-field high-resolution imaging is widely used in aerial surveillance, topographic mapping and safety monitoring. A summary is made for wide-field high-resolution imaging mainly in terms of optical design, and a scaling law that determines how resolution increases as a function of lens size is stated. The generalized scaling law shows that, by using computation to correct aberrations, a greater resolution can be achieved with a smaller camera size. Monocentric multiscale design, which is a good way to achieve wide-field high-resolution photographs, is emphatically described. Finally, an introduction to wide-field cameras based on monocentric multiscale theory is given, a comparison is made on their performances, and existing issues of monocentric multiscale camera are discussed.

Integration, high-precision, and low-cost optical gyroscope embodies the developing tendency of gyroscope, in which fast-light enhanced resonant optical gyroscope shows great potential. The development of the fast-light enhanced resonant optical gyroscope is overviewed through analyzing fast-light arising mechanism and considering experimental reports of dielectric dispersion and structural dispersion. By comparing the relative advantages and disadvantages in fast light generation mechanism, it is more suitable to integrate optical gyroscope system through implementing structural dispersion．Moreover, a novel structure of integrated fast-light enhanced resonant gyroscope is proposed, which may have great potential applications in low-cost civilian areas. Finally, the prospect for the development of fast-light enhanced resonant optical gyroscope in inertial navigation system is discussed.

Two-dimensional photonic crystal waveguides have special photonic band structure, controllable optical dispersion, small physical dimension and high nonlinearity, which draw wide attention in areas of nonlinear optics, optical logic gates, all-optical buffer, optical-power splitter, photonic-crystal lasers and high-sensitivity sensor. The fabrication technique of two-dimensional photonic crystal waveguides is introduced, the propagation loss, self-imaging effect of multi-mode coupling and slow light effect are summarized. In particular, two-dimensional photonic crystal waveguides based on chalcogenide glasses are mentioned. At last, the development prospect of two-dimensional photonic crystal waveguides is discussed.

A method is established based on laser tweezer Raman spectroscopy (LTRS) for rapid quantitation of lipid and carotenoids in Rhodosporidium toruloides cells, and the effect of different C/N, C/P, C/S ratios in a culture medium on synthesis capability of lipid and carotenoids is evaluated at the single-cell level. The Raman spectra of plant oil in different concentrations are determined and the high correlation between the peak intensity at 1751 cm-1 and lipid concentration is observed. The correlation coefficient of standard curve for lipid is 0.97349. The data demonstrate that LTRS is efficient in the quantitative analysis of lipid. Lipid and carotenoids within the Rhodosporidium toruloides cells cultured in different media are acquired quantificationally according to the Raman peak heights at 1751 cm-1 and 1520 cm-1, respectively. The result shows that the lipid concentration in Rhodosporidium toruloides cells increases as the ratios of C/N, C/P, C/S increase, while the carotenoids concentration significantly decreases. The above results indicate that LTRS is an efficient approach to the quantitative analysis of lipid and carotenoids in microbial cells and the optimization of fermentation medium.

We study the feasibility of pork freshness grading using the spectra detected at different time with near infrared spectroscopy and establish the analysis model by using near infrared OPUS software. In order to reflect the pork freshness level better, we use SOM neural network clustering method to differentiate afresh the volatile base nitrogen national standard (Total Volatile Basic Nitrogen, TVBN) level, from 3 levels to 5 levels. In order to improve the prediction accuracy, based on the pretreatment by selecting a derivative+vector normalization method (13 smooth points), principal component analysis is used for dimension reduction before clustering analysis. It can reduce the prediction deviation, further improve the prediction accuracy of sample, reduce the prediction level deviation, and improve the prediction ability.

According to variable incidence angle demands of anti-reflective film in infrared detectors, the influence of incidence angle on coating polarization effect is analyzed. Many effective methods were usually taken to reduce the influence of polarization effect, but non-polarizing can only be realized for a single incidence angle or a very narrow spectral range. 3～5 μm infrared anti-reflective film is designed with gradient and needle optimization methods on the designed initial film stack formula by using TFCalc formula design software. The p and s polarized light transmittance curves of designed formula with incidence angle from 0° to 60° are analyzed, and the effect of changing the film thickness by 2% on the transmittance curve of the designed formula is simulated. In this way, the infrared non-polarizing anti-reflective film is successfully designed with wide range and wide incidence angle.