In the Ramsey scheme, if one modulates two microwave pulses at the same frequency but with ±π/2 phase differences, the Ramsey fringes with zero amplitude at the central frequency can be obtained. By monitoring the variance of probe light absorption by atoms at the central frequency, one can get the error signals which feedbacks to the local oscillator. Theoretically, this method doubles the absorption signal intensity, compared with the case of Ramsey fringes with the same phase. Experimentally, the phase modulated Ramsey pattern in the integrating sphere cold atom clock system is the same with that from theoretical calculation. Moreover, the Ramsey fringe oscillation amplitude here is 1.6 times of that of Ramsey fringes with the same phase.

A scheme of manipulating duplex hybrid entanglement based on linear optics is presented. In the scheme, a variable beam splitter (VBS) is used to generate duplex hybrid entangled states, the hybrid entanglement manipulation is realized by adjusting the splitting ratio of VBS, and the influence of the cat state size on fidelity is analyzed. This scheme provides theoretical reference for the generation of multiplex hybrid entanglement and its application in hybrid quantum communications.

Based on the extended Huygens-Fresnel principle and the method of Stokes parameters, the model of degree of polarization (DoP) for partially coherent flat-topped electromagnetic beam propagation in slant turbulent atmosphere is studied. Not only the DoP in the same direction and the DoP in the cross-direction are considered, but also the effects of the beam order, zenith angle, frequency, receiving height and other factors on the DoP are discussed. The numerical results show that the effect of zenith angle on the DoP can be reduced by increasing the receiving height when laser transmits in slant turbulent atmosphere. The DoP decreases with the increase of beam order and the increase of frequency near the center angular frequency. Unlike the DoP only considering the same polarization，the change of DoP is unbounded with the increase of propagation distance when the cross polarization is considered.

A distorted wavefront to be corrected is built up based on the random phase screen. Then, the stress load spectrum of deformable mirror (DM) in the correction process is obtained by the stress analysis model with finite element analysis method. The fatigue life prediction model is proposed based on stress-cycle (S-N) curve and Miner cumulative damage theory. On this basis, the fatigue damage characteristics of DM for wavefront correction process are analyzed, and the influence of the different driving ways and different structural parameters on the lifetime of DM is discussed in detail. The results indicate that, in the wavefront correcting process, the fatigue damage of rear surface of DM substrate is more serious than that of front surface, and the joints between substrate and poles are most likely to be damaged. For a given shape of the distorted wavefront to be corrected, the fatigue life of DM decreases gradually with the increasing of the peak valley (PV) value of the wavefront. On the other hand, when the PV value of the wavefront is determined, the fatigue life also decreases due to more obviously concentrated stress caused by the increasing high frequency components in the distorted wavefront. Furthermore, the fatigue life can be affected by the structural parameters of DM as well. The life of DM decreases gradually with the increasing of the thickness of substrate, the decreasing of the length of poles and the increasing of the diameter of poles, among which the effects caused by the variation of the diameter of poles are more significant.

Ten different morphological distributions of melanin were proposed according to their formation and migration characteristics, and the two-scale skin model was applied to investigate the influence of melanin distribution, pulse duration, diameter and depth of pathological vessels on the radiant exposure thresholds of epidermis and blood vessels during the laser therapy of Port Wine Stains. The results show that distribution of melanin has a greater influence on the epidermal radiant exposure threshold, but has no obvious effect on that of blood vessels. It is conducive to cure vascular skin diseases with more uniform distribution of melanin, smaller and shallower blood vessels. Spray duration should be prolonged when melanin only distributes in the basal layer of epidermis. With optimized curative effect for vascular skin diseases, long laser pulse (15-20 ms) will be a good choice for more uniform distribution of melanin and larger blood vessels, which is close to the thermal relaxation time of pathological vessels. Shoter pulse (<5 ms) should be chosen for smaller and deeper vessels.

Based on the diffraction wavelength selectivity of reflection holography and the colorimetric principle, the expression of reflection hologram colorimetric system is given, and the conversion relationship between reflection hologram colorimetric system and CIE XYZ system is studied. The factors influenced color of reflection hologram are analyzed and discussed, and the color evaluation method between real object and reconstructed image of reflection hologram is proposed. For various color samples, the chromatism of reflection holograms under different trichromatic systems is calculated by the conversion relations of trichromatic system and chromatic evaluation，and the color difference analysis is carried on.

Currently, the digital holographic microscopy mainly uses a laser with good coherence as the light source, it needs high equipment and environmental condition, and a large amount of speckle noise and parasitic interference fringes can be introduced in the hologram. The digital holography based on incoherent light is a good way to solve those problems. The incoherent light is used as light source, and the spatial light modulator as well as the CCD composes the incoherent interference for recording holograms. The recording and reconstructed principles of the module are analyzed from the angle of wave optics and the point spread function, the axial magnification and the reconstruction distance of the system are given. Combined with the microscope objective, the system of in-line incoherent digital holographic microscopy in reflection configuration is established. Using this system to experiment on resolution imaging, a high spatial resolution as good as the traditional wide-field optical microscopy is obtained. The holograms of one diamond whose size is between 300 μm and 500 μm are captured and digital focused in different planes. The results show that the proposed system can obtain three-dimensional spatial information of small objects simply and rapidly. It has limitless applications in medical tests, material analysis and other fields.

One method to fabricate 45° micromirror based on the excimer laser, namely, the laser stepped ablation method, is proposed and its fabrication process is introduced. Via the parameter optimization, micromirror samples are obtained, and the influences of sample parameters on the reflecting performance of micromirrors are analyzed in detail. A vertical coupling experiment is conducted with these micromirror samples. The main factors that affect the system loss are discussed deeply. The experimental results indicate that the loss induced by micromirrors is approximated to 3.5 dB. The fabrication technique proposed is expected to be widely applied in the fabrication of coupling components of large size optical waveguide interconnection backplane.

The experimental study of the double-pass grazing incidence Nd:YVO4 slab amplifier is carried out based on a saturable absorber and passively Q-switched microchip seed laser. The microchip laser with pulse duration of 90 ps, repetition rate of 100 kHz and beam quality factor M2=1.16 acts as seed laser with varying output power from 100 μW to 10 mW. The new liquid pure metal grease and the traditional thin indium foil are used to conduct as the thermal contact material of the bounce amplifier, respectively. Experimental results show that the liquid pure metal grease conducted as the thermal contact material can significantly reduce the temperature of the slab pump face. An output power of 13 W is obtained with the liquid pure metal grease conducted as the thermal contact material for 10 mW seed power with pump power of 55 W at an optical-optical efficiency of 23%, achieving pulse peak power of 1.2 MW and pulse energy of 130 μJ. The beam quality factor is M2x=1.30,M2y=1.28.

A LD side-pumped, high-energy, high-beam-quality, Q-switched all-solid-state Nd:YAG laser without water cooling is developed. The laser is integrally cooled by using a thermoelectric cooler (TEC), which is conducive to the miniaturization and portability of laser. The dimension of Nd:YAG crystal rod used in experiment is φ7 mm×100 mm and the Nd atom number fraction is 1.1%. The maximum peak power of LD pumping is 15 kW. The laser output at 1064 nm with the maximum pulse energy of 350 mJ, pulse width of 9.7 ns, optical-electrical conversion efficiency of 6.7% and energy stability less than 5% is obtained at 10 Hz repeat frequency. The beam qualities M2 of horizontal and vertical directions are 7.7 and 12.3, respectively.

A new Yb-doped silicate Yb3+∶Sc2SiO5 (Yb∶SSO) has gained considerable attention in recent years because of its advantageous laser properties. Thermal lens focal length of Yb∶SSO in the case of different pump powers is calculated theoretically, and the influence of the material negative refractive index on the laser is analyzed. Quasi-continuous wave (QCW) lasing performance of the Yb∶SSO end-pumped by a 976 nm laser diode (LD) in a plane-plane cavity is studied. Laser output with peak power of 27.6 W and optical-to-optical conversion efficiency of 21.4% is obtained when pump laser is with repetition rate of 50 Hz, pulse width of 500 μs and peak pump power of 128.8 W. Meanwhile, beam qualities (M2 factors) in x and y directions are 1.24 and 1.20, respectively. Finally, the law of center-wavelength movement caused by the change of pump power is observed, and the cause of dual-wavelength asynchronous oscillation is theoretically analyzed.

A L-band high power sub-picosecond erbium-doped fiber laser is reported. Two polarization controllers (PC) and a polarization dependent optical isolator (ISO) are fused in all fiber ring cavity, and the mode-locked laser pulse output of all fiber structure is realized based on nonlinear polarization rotation mode locking principle. The center wavelength of the laser is 1603 nm, the pulse repetition rate is 37.8 MHz, the single pulse energy is 4 nJ, and the average output laser power is 152 mW. Reasonable dispersion management is carried out on the all fiber mode-locked laser, and the mode-locked laser output with pulse width of 370 fs can be obtained. The erbium-doped fiber with high doping concentration is adopted in experiment, and its length is reduced effectively. The pump conversion efficiency is improved. The L-band sub-picosecond fiber laser with simple and compact structure, stable and reliable performance is realized.

A full range-gated imaging system is designed based on the underwater high-repetition-rate range-gated imaging system. In the spatial domain, the whole detecting range can be divided into multiple improved gating slices which are superposed mutually, and the laser pulses during the integral time of one image frame are distributed in these improved gating slices according to certain rules. In the time domain, the echo isotime movement of different laser pulses after the wide gate is achieved according to certain rules, the target information of all distances can be integrated into one frame, and the full range-gated images can be obtained in real-time. The imaging quality of the continuous-wave imaging system, the traditional range-gated imaging system and the full range-gated imaging system is compared. The experimental results show that the full range-gated imaging system has advantages of large detection range possessed by the continuous-wave imaging system and long detection distance possessed by the range-gated imaging system. The detecting distance exceeds 4.5H (H is the decay length) and the frame rate of the output video is 20 Hz.

Highly effective frequency doubling conversion of continuous wave, single frequency 1064 nm fundamental laser is realized based on the Pound-Drever-Hall extra-cavity frequency doubling technique. When the length of the ring frequency doubling cavity is accurately locked, the green laser output with the highest power of 8.73 W is obtained with the use of the type I noncritical phase matching lithium triborate crystal and the frequency conversion efficiency reaches 68.9%. Based on this, the green laser wavelength locking and tuning characteristics are investigated, and when the root-mean-square value of wavelength locking within 30 min is less than 3 fm, the wavelength is continuously tunable between 532.15 nm and 532.50 nm. The spectroscopic property of the single frequency green laser is measured by the self-heterodyne method, the measured line-width is 18.7 kHz, and the beam quality factor is 1.25 which indicates a superior beam quality.

A quaternary eutectic alloy composed of Ti70.38Fe26.39Sn2.93Y0.3 is prepared on a pure titanium substrate with the laser rapid prototyping technique. The microstructure, hardness, elastic modulus, and corrosion resistance of the alloy are tested and investigated, and the comparison with those of the Ti70.58Fe29.42binary eutectic alloy is made as well. The analysis results show that, under the non-equilibrium solidification condition of laser rapid prototyping, the solidification structure of Ti70.38Fe26.39Sn2.93Y0.3 alloy is a hypereutectic one which composed of the primary TiFe and β-Ti+TiFe eutectic. Compared with those of Ti70.58Fe29.42binary eutectic alloy, the hardness of Ti70.38Fe26.39Sn2.93Y0.3 alloy is increased by 59.6%, the elastic modulus is decreased by 30%, and the Ti70.38Fe26.39Sn2.93Y0.3 alloy possesses an excellent property of corrosion resistance in Hank′s solution.

The WC/SS316L composite coatings are successfully deposited on the surface of 316L stainless steel by both the supersonic laser deposition (SLD) and the traditional laser cladding (LC) technologies. The macroscopic morphology, WC distribution, microstructure, phase composition and wear-resistant property of the as-deposited coatings are comparatively studied. The results show that the multi-pass overlapped coating of LC has obvious macroscopic cracks, while the surface of SLD coating keeps smooth and compact without macroscopic defects. In the LC coating, the ceramic WC particles unevenly distribute, while in the SLD coating they uniformly scatter. In the LC coating, the microstructure distribution is non-uniform and is along with the generation of harmful phase, while in the SLD coating, the deposited powder maintains its original microstructure and performance and the severe plastic deformation can also be observed. The friction coefficient of the SLD coating is 28% lower than that of the LC coating, and thus it exhibits a better wear-resistant property.

To investigate the formation mechanism of residual stress holes on 7050 aluminum alloy sheet sample under laser shock, the 7050 aluminum alloy samples are shocked by laser with the power densities of 1.98 GW/cm2 and 2.77 GW/cm2, respectively. The finite element analysis software ANSYS/LS-DYNA is applied to simulating the sheet samples shocked by the laser with the power density of 1.98 GW/cm2. The distributions of residual stress on the thin plate and thick plate samples are analyzed by the X-ray stress analyzer, the dynamic strain of the samples shocked by laser is measured with the piezoelectric thin film sensor, and the surface microstructures on the shocked areas are observed through a three-dimensional microscopy system. The experimental results show that the residual stress holes on the 7050 aluminum alloy sheet samples can be caused by the laser shock with the power densities of 1.98 GW/cm2 and 2.77 GW/cm2. Under reflecting boundary conditions, the simulation results agree well with the experimental data, which indicates that the gathering of the rarefaction waves to the center of the light spot is the major reason for occurrence of the residual stress holes. Through analyzing the distribution of residual stress and dynamic strain, the influence of the shock waves reflected back and forth in the samples on the residual stress holes cannot be ignored. The thicknesses of the central areas of the thin and thick plate samples are 10.800 μm and 8.150 μm thicker than those of their surrounding areas respectively after the samples are shocked by laser with power density of 2.77 GW/cm2. The residual stress holes on sample surfaces are caused by the joint effect of rarefaction waves and laser shock waves.

We set up the three-dimensional transient model of laser deep welding considering the mechanical factors such as evaporative recoil back pressure, the surface tension, the thermal buoyancy and the thermal parameters such as the convection and radiation in and out of pool. In the model, the rotary Gauss body heat source attenuation along depth direction was applied to solving the laser absorption of molten pool, the volume of fluid method was adopted to realize the tracking of gas-liquid interface, and the liquid volume fraction method and enthalpy-porosity technique were used to compute the latent heat of molten metal solidification and the momentum loss in the liquid-solid mush zone. Then the transient temperature and flow field of pool and keyhole were obtained in the laser deep penetration welding of stainless steel with this numerical model. The results show that there are three stages for the maximum temperature in the welding, which are the linear growth, the tending to stable and the small oscillation, and there are two kinds of keyhole orientations, forward and backward, and there is periodic oscillation for keyhole in the welding. The simulation results of final weld formation agree with experimental data of weld cross section well, and the oscillatory behavior of keyhole is also verified by relevant references.

Ultrasonic vibration is introduced to the process of laser metal forming IN718, the surface roughness, residual stress, and mechanical property at room temperature of the forming parts are tested, and the microstructure of the forming parts is analyzed. The results show that, after the introduction of ultrasonic vibration, the surface roughness degree and residual stress of the forming parts are remarkably improved, the microstructural refinement is obtained, and the tensile strength and yield strength are both increased. When the ultrasonic power is 44 W and the ultrasonic frequency is 17 kHz, compared with those of the parts without experiencing ultrasonic vibration, the residual stresses along x direction and y direction respectively decrease by 47.8% and 61.6%, the yield strength, tensile strength and area reduction increase by 6.1%, 2.7% and 10.6%, and the elongation slightly decreases to 29.2%.

The dynamic finite element software ANSYS/LS-DYNA is used to numerically simulate the deformation velocity of 316L stainless steel target after the influence of shock waves induced by intense lasers with pulse duration of nanosecond scale, and a finite element model suitable under the high pressure and high strain rate conditions is established. The velocity curves obtained from the numerical simulation are compared with the previous experimental results and the validity of the above model is confirmed. In addition, the deformation velocity increases with the increase of the shock wave pressure and decreases with the increase of the target plate thickness. These simulation results provide a reference for the study of those processes, such as laser shock forming, spallation and penetration.

To investigate the mechanism of laser coloring and fabrication of micro- and nano-structures fabrication on stainless steel, the influence of such laser parameters as defocusing distance, pulse energy, scanning interval, scanning speed, and repetition rate is studied. The oxide film, grating-like structure, concave and columnar protrusion are produced. The four structures lead to thin-film interference, grating diffraction effect and light trapping effect. A BP (back propagation) neural network with one hidden layer between process parameters and color parameters is established via Matlab. The training root-mean-square error of this BP neural network is 0.0078. The relative errors of hue, saturation and brightness are 23%, 10.4%, 5.6%, respectively. To a certain extent, this neural network reveals the mapping relationship between process parameters and color. The laser coloring effect can be predicted effectively with the neural network model.

Laser shock processing tests of ЭП866 heat-resistant martensitic stainless steel for aero-engine were carried out with a Nd glass laser. The results show that the microstructure of stainless steel is composed of martensite laths and carbide precipitation. High density dislocation entanglement occurs in the surface layer during multi-impacts. After tempering treatment, the tangled microstructures still exist and more nano-sized precipitation occurs. Higher-amplitude compressive residual stress is produced in the surface layer by multi-impacts, whose amplitude can be reached to -715.4 MPa, and residual compressive stress increases with the increase of the lap times. After tempering treatment at the temperature from 500 ℃ to 650 ℃, the residual compressive stress decreases by 50%, which shows good thermal stability and weak effect of the impact times.

To study the transmission mechanism of laser shock wave inside the 690 high-strength steel sheet, the dynamic strain induced by laser shock loaded on 690 high-strength steel sheet was simulated on the platforms of Hyperworks and LSDYNA. The experimental results were measured by a polyoinglidene fluoride piezo electric sensor. A comparison between the simulated and the experimental results was carried out. The dynamic strain model of 690 high-strength steel surface under high strain rate of pulsed laser shock and a loading model under the laser shock waves on 690 high-strength steel sheet were established. The results show that under the loading power density of 12.7 GW/cm2 through changing the location of measuring position and thickness of the samples, the measured velocity of Rayleigh wave and laser shock is 3.08×103 m/s and 3.09×103 m/s respectively. The simulated value of the velocity of the surface Rayleigh wave was 3.24×103 m/s, displaying significant consistency with the experimental results. Shear wave and Rayleigh wave can be separated through adjusting the power density of laser shock. The experimental results demonstrate the accuracy and reliability of the dynamic strain model of 690 high-strength steel surface loaded by the pulsed laser shock wave. The loading model of the laser shock waves can be used to describe the transmission mechanism of the laser shock wave inside the 690 high-strength steel sheet.

By utilizing a fiber beam splitter and N single-photon sensors, a method for single pulse photon counting in the case of weak light signal detection is proposed. Combining the photon counting method and ghost imaging, a ghost imaging platform based on photon counting is established. The effect of dynamic range of the photon counting on the ghost imaging is investigated by experiments, and some problems to be further studied about this method are also discussed.

Theoretical analysis and numerical simulation on the error accumulation in sub-aperture stitching for flat optics show that the slope and piston difference of the surface figure of the reference flat in the overlap region lead to error accumulation. In order to improve the measurement accuracy of the stitching interferometer for large aperture flat optics, a simple and effective method is proposed to reduce the accumulated error of the sub-aperture stitching measurements. The 4th and 6th Zernike aberration are used to synthesize the virtual reference flat. Sub-apertures with the virtual reference surface figure removed are stitched together to obtain a full aperture surface figure. A flat mirror with aperture of 450 mm×60 mm is tested by 8 sub-apertures. The deviation between the stitching result and the test result of a Zygo 24 inch (600 mm) interferometer peak-valley (PV) value is reduced from λ/7 to λ/100 after removing the virtual reference surface figure. The flatness of the virtual reference surface is 0.02 λ (PV value), which is comparable to the real figure of the reference flat. The influence of removing of virtual reference surface on the measurement results of each sub-aperture can also be ignored. Experimental results show that the accumulated error is controlled effectively and the measurement accuracy is evidently improved by the proposed method.

In order to meet the surface shape detection requirements during the grinding stage of large aspheric mirror, the sources of angle error are analyzed, and one method for correcting angle measurement errors based on the S polynomial fitting is proposed. This method can effectively overcome the impact caused by the sampling location and random errors. As for the angle errors at different frequency bands, the generatrix and composite sampling methods are proposed, and the generatrix sampling method can suppress the random error within 10% in the calibration experiment. The above correction method is proved to be possible, which provides a theoretical basis for its application in the future experiments.

Inverse algorithm is one of the key issues in particle measurement technology of light scattering. Single parameter regularization algorithm, especially Tikhonov regularization algorithm is widely used for back calculation of particle size distribution (PSD) in laser particle analyzer. One of the disadvantages of such algorithm is that the inversion solution is usually oscillatory and contains negatives values. To improve the situation, a multi-parameter regularization algorithm is proposed. By constructing a band-pass filter function that is controlled by multiple parameters, oscillations and the height of the regularization solution can be controlled, respectively. Moreover, the regularization solution is nonnegative constrained. Simulated results and experimental evidences show that, with the optimized parameters, the algorithm can eliminate the oscillations and negative values brought by regularization algorithm. Meanwhile, the proposed algorithm has better ability to recognize the multimodal system, and exhibits high ability to effectively re-construct the PSD.

The directional excitation of surface plasmons play an important role in optical communication, biosensing, integrated circuit and nanolithography, etc. An asymmetric L-type single-slit structure which can excite surface plasmons directionally is proposed. Without changing structural parameters of the slit, both the unidirectional and the bidirectional excitations of surface plasmons can be realized by changing the incidence angle. The structure can also play a role in all-optical switch modulation. The machining process of the L-type single-slit is elaborated in detail. The slit structure is simulated by the COMSOL Multiphysics software. The results show that the asymmetric single-slit structure can realize the directional excitation of surface plasmons with a continuously adjustable extinction ratio from -16 dB to 15 dB. This single-slit structure occupies small area and is convenient for manipulation and machining, and it has guiding significance for the research on micro-nano plasmon optical devices.

A new method for multicarrier optical source generator based on cascaded electro-absorption modulator (EAM) and phase modulator (PM) is proposed. The multicarrier optical source based on single EAM cascaded PM and the multicarrier optical source based on two EAM cascaded PM are studied and analyzed respectively. Multicarrier optical source with frequency interval of 15 GHz, and 19 subcarriers in the range of 0.1 dB flatness using multicarrier optical source generator based on two EAM cascaded PM are achieved. Meanwhile, the influence of the parameters such as modulation index and chirped factor of EAM, modulation signal amplitude of PM, and phase difference of modulators on the multicarrier optical source are also analyzed.

Geostationary orbit (GEO) satellite-borne laser communication antenna suffers from solar intrusion near midnight every day. In order to analyze the influence of midnight solar intrusion on the temperature of the optical antenna, a thermal analysis model of the laser communication terminals on the geostationary communication satellite is built, and temperature variations of the optical antenna are calculated．According to the analysis results, when the solar declination is ±8.8°, the midnight solar intrusion has more influence on the temperature of the main mirror and the secondary mirror than that of 0° solar declination. Effects of the midnight solar intrusion on the temperature of the optical antenna can be effectively reduced by using the optical antenna sunlight avoidance strategy. When the maneuver angle is 40°, by choosing a time combination of start time being 2.0 h before midnight and stop time being 2.5 h after midnight, the maximum continuous working time can be obtained.

有源光子晶体光纤的芯径较大，主要用于实现高峰值功率（高能量）的脉冲放大输出。目前只有NKT Photonics公司可提供商品化的掺镱（Yb3+）有源光子晶体光纤，其最大芯径约为85 μm。光子晶体光纤的制作主要受光纤预制棒中纤芯尺寸的限制。为实现百微米芯径的光子晶体光纤，预制棒中纤芯材料的直径须达到5 mm，目前较难实现。中国科学院上海光学精密机械研究所立足于自身在材料制备方面的优势，利用溶胶-凝胶方法制作了直径大于5 mm的掺Yb3+石英棒，并拉制出纤芯直径为110 μm的有源光子晶体光纤。对拉制的掺Yb3+光子晶体光纤进行性能测试。当波长为915 nm时，该光纤的吸收系数为8 dB/m，此时采用较短的光纤（约为1 m）即可充分吸收抽运光，降低了脉冲放大过程中的非线性效应。对光子晶体光纤进行了皮秒级脉冲放大，种子光的波长为1030 nm，输出功率为10 W，脉宽为21 ps，重复频率为10 MHz。在抽运功率为590 W（抽运波长为976 nm）条件下，实现了功率为309 W的脉冲放大输出，峰值功率高达1.47 MW，放大效率为52%。图1为光子晶体光纤截面及光纤的功率放大曲线。自制光子晶体光纤的放大效率低于NKT Photonics公司光子晶体光纤的放大效率（60%）的主要原因为自制光子晶体光纤对抽运光的耦合效率偏低。但NKT Photonics公司提供的光子晶体光纤仅应用于功率小于100 W的脉冲放大输出。当脉冲输出功率为309 W时，检测光谱性能，未发现明显的非线性效应。图2为检测到的放大脉冲序列及脉宽。本课题组成功制作了百微米芯径的掺Yb3+光子晶体光纤，实现了功率为309 W、重复频率为10 MHz的皮秒级脉冲放大输出。

高功率光纤激光器在工业加工、材料处理等领域有着诸多应用，得到国内外研究机构的广泛关注。目前，高功率光纤激光器主要有两种结构，一种是直接振荡器结构，一种是主振荡功率放大结构。采用振荡器结构的光纤激光器具有结构简单、稳定性好、成本低廉等优点，是目前中低功率激光器市场使用较多的一类方案。2013年，国防科学技术大学基于单端抽运结构实现了输出功率1 kW的全光纤振荡器；2014年，又将该方案的输出功率提高到1.5 kW。2014年，芬兰CoreLase公司推出了2 kW的全光纤振荡器产品，美国相干公司基于空间结构实现了3 kW的光纤振荡器。但是，由于热效应、非线性效应的限制，尚未出现相关输出功率大于2 kW的全光纤振荡器报道。双端抽运方式既能够在一定程度上抑制非线性效应，又能使热分布更加均匀从而降低热效应的影响，因此具有较好的高功率应用潜力。基于双端抽运方式，在全光纤振荡器中成功实现了2.5 kW激光功率输出，其原理结构如图1所示。激光谐振腔由高反射光栅（HRFBG）和低反射光栅（OCFBG）构成。前向抽运源利用一个7×1的抽运合束器（PC）将6个300 W的975 nm 激光二极管（LD）合为一束，并注入到激光谐振腔中；后向抽运合束器为（6+1）×1的抽运/信号合束器（PSC），同样将6个300 W的LD合束并注入到激光谐振腔中。激光器增益介质采用纤芯、内包层直径分别为21，400 μm的大模场掺镱光纤（LMA YDF）。激光谐振腔输出的激光经过包层光滤除（CLS）后，熔接商用的光纤端帽输出。在前后向最大抽运功率为1.94 kW和1.52 kW、总抽运功率为3.46 kW时，振荡器输出功率为2.5 kW，光光转换效率为72%，如图2(a)所示。测量了不同功率光谱，如图2(b)所示，结果表明输出激光中无抽运光和放大自发辐射，与信号相比受激拉曼散射在20 dB以下。实现最高输出功率时，测得的光束质量因子M2约为1.3，并未观察到模式不稳定现象，典型的远场光斑形态如图2(c)所示。在该结构中，如果能够有效抑制受激拉曼散射，适当增加抽运功率，有望获得更高的激光功率输出。

Single-photon laser altimetry represents the future space-based lidar′s development trends. It has many advantages, such as high sensitivity, high repetition rate, light weight and small volume. For normal Gaussian echo signal, this work establishes the detection probablity model based on the lidar equation and the statistical property of single-photon detector. Then the quantitative relationship of ranging error is obtained. The model is simulated by Monte Carlo method. Results show that narrower pulse width can result in better accuracy and precision. Higher echo signal intensity can also lead to better precision, but it will bring about worse accuracy. Take the case of a Gaussian echo whose root-mean-square width is about 1.5 ns, when the mean number of signal photons is 1, the range accuracy is about 6 cm and the range precision is approximately 22 cm with a single shoot regardless of the noise.

In biomedical photonics, it′s extraordinarily significant to diagnose lesion by quantificational optical parameters. The high resolution absorption and refractive index spectra of samples with different mass fractions of anhydrous glucose are acquired by terahertz time-domain spectroscopy system in nitrogen atmosphere at room temperature. The characteristic absorption is analyzed based on the density functional theory. Gaussian 09 is used to simulate the optimized structure and infrared spectroscopy of single, double, triple and quadruple glucose molecules. Meanwhile, the B3LYP method and 6-31G basis set are adopted. In addition, Materials Studio 7.0 is utilized to calculate the frequency generated by the glucose crystal structure, which is completed with the help of CASTEP module and the generalized gradient approximation method. The experimental results show that the absorption peaks of anhydrous glucose appear at 0.94, 1.30, 1.44, 1.67, 1.88, 2.08, 2.31, 2.55, 2.70, 2.84, 2.96 THz, and the intensities of all peaks decrease linearly with the increase of glucose mass fraction in samples. The results of Gaussian 09 are in agreement with the test results, which is verified by the results of CASTEP. The absorption peaks of anhydrous glucose are identified accurately in the range of terahertz waveband. The reuslts provide experimental and theoretical reference for the application of terahertz time-domain spectroscopy in biomedicine.

Tunable diode laser absorption spectroscopy (TDLAS) combining with computer tomography (CT) algorithm can realize 2D reconstruction measurement of temperature of flow field, concentration and so on. Two different typical algorithms including algebraic reconstruction technique (ART) algorithm and simulated annealing (SA) algorithm are implemented to study the influence of tomography algorithm on the 2D reconstruction quality of temperature field. Reconstruction simulations of unimodal temperature field and bimodal temperature field are done under different beam arrangements and absorption lines with two algorithms, and the reconstruction results of two algorithms are compared and analyzed. The simulation results indicate that the reconstruction quality of ART algorithms is mainly affected by the beam arrangement while the SA algorithm is more sensitive to the number of absorption spectral lines. More specifically, for unimodal temperature field, the maximum deviation of the reconstructed results of ART algorithm is 5.6%, which is slightly better than 6.2% of SA algorithm using 6 absorption lines. For bimodal temperature field, the maximum deviation of the reconstructed results of SA algorithm is 5.5%, whiles that of ART algorithm is up to 22%.