The severe backscattering of the shallow seawater hinders the application of laser detection in the offshore. The backscattering of the echo signal restricts the depth, resolution and contrast of underwater target detection. Two novel ocean lidar with high scattering suppression radio, chaotic lidar and coherent dual frequency lidar, are studied. The two kinds of lidar signal owns the inherent intensity modulation characteristics, and then the backscattering owns the low frequency characteristics. Therefore, the backscattering of seawater can be removed by the band-pass or high pass filtering when the intensity of the object signal and the backscattering light are the approximate same magnitude, so as to improve the signal-to-noise ratio of the ocean lidar system.

For accurately reflecting the practical features of turbulence, the modified atmospheric spectrum model should be applyed in the simulation of optical wave propagation. To do this, the method of generating high precision turbulent phase screen is proposed for the modified atmospheric spectrum, which is on the bases of optimized phase screen model. By extending the low frequency region and changing the sampling Settings of the model, the maximum relative error in low frequency region was reduced to 1%. As a comparison, the maximum relative error in low frequency region was 6.75% for Optimization-based method before improvement, 22.99% for original FFT method and 16.81% for subharmonic method. By using this method, the simulation of Gaussian beam propagating in turbulence has been done and the second-order statistical properties include beam spread and beam wander have been estimated. The results show that, in the case of weak fluctuation level, the degree of compliance with the analytical approximations is good. However, under the condition of strong fluctuation level, the deviation between the simulation results and the theoretical results is increasing with distance. The deviation of beam spread is up to 6 cm, while the deviation of beam wander is up to 1 cm, which can be explained by the fact that the theoretical model cannot predict the beam wander saturation. In comparison with the simulation results of the Von-Karman spectrum, the beam spread estimated by modified atmospheric spectrum is slightly larger than that estimated by Von-Karman spectrum, and beam wander predicted by modified atmospheric spectrum shows a faster rate to reach saturation than that predicted by the Von Karman spectrum, it is precisely induced by the "bump" of the modified atmosphere spectrum. So we conclude that the phase screen generated by the method presented here can characterize the refractive index perturbation of the actual atmosphere effectively.

A testing method for X-ray detector is proposed through the study of the characteristics of pulsar radiation signals and the analysis of space observation requirements. Firstly, based on the photon radiation model, the probability formula of X-ray photon under detection is deduced, and the influences of different source flux and different detector time resolution on photon detection ability are analyzed. The relationship between pulse arrival time and the similarity of pulse profile is established by numerical simulations. And the observations of the crab pulsar of hard X-ray modulation telescope is processed, the pulse profile characteristics of the Crab pulsar at different energy are studied. Secondly, the testing and processing methods of X-ray detector for pulsar navigation are studied systematically, and the testing work of a self-developed focused X-ray detector is completed by using the ground testing system. The test results show that the background noise of the focused detector is 3.63×10-5ph/(cm2·s-1), the working energy range is 0.2~22.7 keV, the time resolution is 4.17 μs, and the spatial response is about 5'. The energy linearity of the detector is good, the integral nonlinearity is 0.52%. The energy resolution of the detector is better than 200 eV at five characteristic energy spectra, and the best detection efficiency is 39.18%@4.51 keV. Under the condition of weak pulse signal and strong background noise, the detector can accurately restore the pulse profile of Crab pulsar. The signal-to-noise ratio and similarity of the pulse profile increase with the increase of pulse flow and the decrease of background noise. The detector can detect the pulse signal whose radiation flux is 10 times less than the background noise in 2 400 s. The results show that the focused detector has excellent performance and can meet the space observation requirements of navigation pulsars(eg. PSR B1509), so the feasibility of the testing method is also verified.

Based on the gate controlled characteristics of GaN High Electron Mobility Transistors(HEMTs) and on the photovoltaic effect in lead zirconate-titanate(PZT) ferroelectric thin films, a new type of photosensitive layer/HEMT detector structure is proposed. For this purpose, a PZT ferroelectric thin film is deposited on the gate of a HEMT device to prepare films with optimized photovoltaic performance, the surface morphology and ferroelectric properties of PZT films prepared by different sputtering powers and annealing temperatures are analyzed. It is found that the best conditions for the grain growth on the surface of the film are at 200 W sputtering power and 700℃ annealing temperature, and the residual polarization intensity is 38.0 μC·cm-2.The output characteristics of the fabricated photosensitive gate PZT/GaN-based HEMT devices are compared to those of pristine HEMTs under both dark condition and 365 nm ultraviolet light. The results show that the source drain saturation voltage of the HEMT with the ferroelectric thin film decreases by 3.55 V and the saturation current increases by 5.84 mA, compared with those without light, clearly indicating a significant UV photodetection capability. To achieve the purpose of optimizing the structure of the new type of detector, detectors with different grid lengths such as 1 μm, 2 μm, and 3 μm are tested. The results show that under ultraviolet light, the drain saturation currents of the three detectors are 23 mA, 20 mA, and 17 mA, respectively. Therefore, the longer the gate length, the smaller the saturation current of the device, and the worse the detection performance.

A photonic microwave phase-shifting system with continuously tunable phase shift and the frequency multiplication factor is proposed. Without optical filters, the scheme is mainly consists of two integrated dual-polarization dual-parallel Mach-Zehnder modulators. By adjusting voltages of the radio frequency driving signal and direct current bias signal on the dual-parallel Mach-Zehnder modulators and phase modulator, frequency-doubling or frequency-tripling, ..., frequency-sextupling microwave signal can be generated, with 0° to 360℃ontinuously tunable phase shift. The simulation results show that, when the radio frequency signal frequency is 10 GHz, the output microwave signals with the frequency 20, 30, 40, 50, 60 GHz can be obtained respectively. The ratio of direct current bias voltage and half-wave voltage of the phase modulator is set to vary from 0 to 1, corresponding to phase shift of microwave signal vary from -180° to 180°. In addition, the effects of the extinction ratio of the modulators on the optical sideband suppression ratio and electrical spurious suppression ratio of the output microwave signal, as well as the effects of the phase balance of the 90° hybrid coupler on the phase drift and amplitude variation of the microwave signal are analyzed.

The influences of the geometric deformation of the fiber such as ellipse, misalignment and diametrical nonuniformity on the performance of Orbital Angular Momentum (OAM) modes propagated in the Hollow Ring-core Polymer Optical Fiber (HRC-POF) are studied by full vector finite element method. In addition, the maximum deformation that the fiber can withstand under the condition that maintains the stable transmission of the OAM mode is also studied. The results show that the ellipse and misalignment will cause the mode walk-off upon propagation, leading to the decrease of the purity of synthesized OAM modes and the increase of the crosstalk. Numerical results show that the purity of synthesized OAM modes is more than 99.02% and the crosstalk is less than -20.08 dB when the ellipticity or misalignment is within 1.0%. The diametrical nonuniformity of the fiber will only affect the number of OAM modes supported in the HRC-POF. The larger the core radius is, the more OAM modes can be transmitted in the fiber. Besides, the original 26 OAM modes can be supported in HRC-POF when the diametrical nonuniformity is -3% to 10%.

This research integrates learning mechanisms of weights calibration and multiple receptive fields in SENet into UNet, working out USENet to achieve super-resolution in digital holographic phase cell images. Structured with symmetrical topology and filtered with multi-scale, the model is trained to improve the image rebuilding accuracy and the capability of generalization. So as to enhance the estimation confidence in the region of interest, the weights calibration blocks are introduced to differentiate the importance of feature map channels. With better visual effects and details, the experimental results confirm that the average numerical score of structural similarity index on validation set has been verified to improve from 0.770 2 to 0.942 7. According to the performances between experimental group of USENet and reference group of network without calibration layer, the region of interests in global images have demonstrated a further improvement from 0.965 5 to 0.970 3 by the block of calibration.

A computational ghost imaging method based on convolutional neural network is proposed to solve the imaging quality and the speed of reconstructed images under low sampling condition. Firstly, a convolutional neural network is trained by using a set of training images which are reconstructed by the correlation calculation method and corresponding lossless images. Then, the test set images reconstructed by the correlation calculation are used as the input layer of the convolutional neural network to learn the sensing model and predict the corresponding images. Finally, the images reconstructed by the convolutional neural network are compared with the images reconstructed by computational ghost imaging and compressed sensing algorithm, respectively. The experimental results show that the proposed method can restore the measured object with high quality when the sampling rate is 0.08, and the image quality is higher than other methods. Meanwhile, it takes about 0.06 s without sacrificing image quality when the method is used to reconstruct the single image, which greatly improves the speed of image reconstruction. The effectiveness of our method is also verified by numerical simulation and optical experiments, which is of great significance for engineering applications.

Aiming at the problems of poor contrast and low signal-to-noise ratio of traditional ghost imaging, a fusion ghost imaging method based on frequency domain decomposition is proposed. This method transforms the speckle pattern obtained in the reference path to the frequency domain, selected an appropriate threshold to decompose it into high-frequency speckle and low-frequency speckle, and performs correlation calculation between the high-frequency speckle and the value obtained by the bucket detector to get high and low ghost image, and finally the inverse non-subsampled shearlet transtransform is used to reconstruct the final ghost image. By using PSNR and contrast as the evaluation index, the effectiveness of the fusion ghost imaging method was verified by 4 sets of experimental simulations. The simulation experiment results show that the PSNR/contrast of the fusion ghost imaging is improved by 41%/173% and 27%/135% on average compared with the computational ghost imaging and differential ghost imaging methods.

The geostationary satellite-borne differential optical absorption spectrometer detects the atmosphere by whiskbroom imaging. Aiming at the requirements of signal-to-noise ratio bigger than 1 000 and the spectrometer's detection period of less than 10 min in high-speed detection mode and less than 1 h in high-resolution mode, the CCD imaging system was designed. Selecting CCD47-20 as the detector, the imaging circuit was designed to realize the sample and upload of image signal. The effects of frame co-added and binning on the time and spatial resolution were analyzed. Combined with the characteristics of the frame transfer CCD, the imaging method of swinging the mirror when reading the last frame of each position was designed, and the number of frame co-added and binning were set reasonably to achieve the purpose of optimizing the imaging cycle. Under 1s exposure time, the detection period is 515 s in high-speed mode and 3 315 s in high-resolution mode, and the signal-to-noise ratio is greater than 1 000 in both two modes. In the pollutant observation experiment, there is no missing frames or duplication occurred. The CCD imaging system meets the detection requirements of the geostationary spaceborne differential optical absorption spectrometer, and provides a technical reference for the design of a spectrometer imaging system eventually applied to a geostationary satellite.

In order to improve the ability of the star sensor to detect limited magnitude, a combination of the improved Cassegrain system, aperture-corrected spherical lens group and field-of-view corrected spherical lens group is adopted to design an optical system of a large aperture catadioptric star sensor capable of correcting astigmatism, field curvature and distortion, with the spectral range of 450~950 nm, the semifield of 1.4°, an entrance pupil diameter of 250 mm and the focal length of 425 mm. According to the calculation of the initial structure parameters of the system based on the aberration theory and the optimization design of Ray tracing in Zemax software, the blocking ratio of the secondary mirror of the optical system reaches 0.43, the energy concentration of the imaging point is 80% within 30 μm, and the maximum distortion is 0.081%. The modulation transfer function is greater than 0.75 at a Nyquist frequency of 34 lp/mm, and the maximum magnification chromatic aberration is 1.138 μm, which meets the imaging requirements of star sensor pairs. Through the tolerance analysis of optical system, in the 20 Monte Carlo analysis results, the best structure is the 13th structure with the performance function of 4.975 16 μm, the worst structure is the 20th structure with the performance function of 7.799 57 μm. Through the performance function analysis of 20 times Monte Carlo structure, the selected tolerance value can well meet the basic requirements of optical system performance, and provide the basis for the errors in the process of processing and installation.

In order to reduce the error of phase calculation caused by the hysteresis nonlinear of piezoelectric ceramic actuator in phase-shifting interferometer, a control system for a piezoelectric ceramic actuator is designed. The high precision resistance strain sensor and signal conditioning circuit based on the principle of phase-locked amplification are used to detect the displacement of the piezoelectric ceramic actuator. A polynomial mathematical model is established to describe the hysteresis nonlinearity. And then, a feed-forward open-loop control method is proposed to compensate for the hysteresis nonlinearity. Finally, based on the proposed scheme, a tracking control experiment of the desired trajectory of the piezoelectric ceramic actuator is performed. At the same time, a compensation control system and an interferometer are used to detect the surface morphology of the optical element. The experimental results show that after compensation, the tracking error of the piezoelectric ceramic actuator is between -0.156 μm and +0.078 μm, and the hysteresis nonlinearity is reduced from 10.4% to 2.4%, and the surface shape undulated height Root Mean Square (RMS) and Peak Valley (PV) of the optical element measured by the interferometer are changed by 0.795 nm and 3.937 nm respectively. It shows that this system is of great significance for the high-precision shape detection of optical components.

In order to meet the needs of multi-parameter variable evaluation of composite materials under complex loads in nondestructive testing, a dual-function digital speckle pattern interferometry system based on optical multiplexing is proposed to realize the measurement functions of both digital speckle pattern interferometry and digital shearography. By controlling one of the combinations of a mirror and quarter-wave plate, the optical setup of digital speckle pattern interferometry is formed when the combination is in the first position, allowing the out-of-plane displacement to be measured. The optical setup of digital shearography is formed to realize measurement of slope when the combination is in the second position. During the process of measurement, the out-of-plane displacement and slope of the object surface due to a single load can be simultaneously measured by simply switching the position of the combination. The optical setup of the dual-function digital speckle pattern interferometry enjoys the advantages of simple structure and high switching efficiency. High quality measurement results of displacement and slope can be obtained using the dual-function digital speckle pattern interferometry. It has been proved experimentally that the dual-function digital speckle pattern interferometry is suitable for field use in nondestructive testing of composite materials due to its superior performance in anti-interference and high-sensitivity testing.

A carbon dots with blue light was synthesized by the N and S co-doped hydrothermal method. After a series of optical and micro-structural characterizations, N and S elements could be sufficiently doped by means of the heteroatom and functional groups on the surface of the carbon dots, which decided as-prepared carbon dots to emit blue color fluorescence with the highest quantum yield of 54.27%. Owing to the strong fluorescence, the as-prepared carbon dots can be used as a sensing probe for the detection of Ag+ and Fe3+ with high sensitivity and selectivity. Both ions possessed different quenching mechanism and limit of detection with the variation of the ionic concentration through the fitting by Stern-Volmer equation, providing a new approach for efficient detection of Ag+ and Fe3+ for the application.

In order to explore the potential of 2D materials in the application field of on-chip tunable active optical devices, a strain-tunable light source with monolayer MoSe2 was produced. High quality monolayer MoSe2 was obtained by micromechanical exfoliation method. The all-dry transfer method was used to transfer monolayer MoSe2 to biaxial piezoelectric ceramics coated with 150 nm thick Polymethylmethacrylate. The electric field was applied to the biaxial piezoelectric ceramic, which converts the electrical signal into the strain signal, to controll the optical properties of MoSe2. The variation of intrinsic exciton and charge exciton peaks of the MoSe2 with strain tuning were observed from photoluminescence spectra at low temperature of~5 K. The results indicate that blueshifts of ~3.8 meV and ~3.7 meV appear in the intrinsic exciton and charge exciton peaks, respectively, when the strain is tuned from tension to compression. And increasing compressive strain or tensile strain will decrease the intensity of intrinsic exciton and charge exciton peak linearly. At the same time, the circular polarization degree of the emission related to the circular polarization of the pump laser also shows regular change with the variation of strain. This research confirms the close correlation between the stain tuning and the optical properties of monolayer MoSe2. It provides support for the development of various on-chip tunable active optical devices based on 2D materials.

In order to study the excitation and evolution of rogue waves in the periodic background, the excitation and evolution of rogue waves in the case of two kinds of periodic potentials (real potential and complex potential) are discussed by using the split-step fast Fourier transform, according to the self-focusing nonlinear Schrodinger equation model with periodic external potential. The results show that rogue waves can be excited by plane waves with a little perturbation in both kinds of periodic potential backgrounds. In the case of only real periodic potential, different optical phenomena can be generated by changing the amplitude and period of the period potential. In the parity-time symmetry periodic potential, a very large rogue wave can be excited, and the real part amplitude has energy dissipation and deceleration effect on the rogue wave, while the imaginary part amplitude has energy gain and acceleration effect on the rogue wave. Therefore, choosing appropriate real part amplitude and imaginary part amplitude can generate large rogue wave with different intensity. The study can help to capture and adjust the strange wave in the optical lattice, and get high power waves.

In this paper, the Zernike polynomials are used to describe deformed optical surfaces. Using the adjoint method, the sensitivities of Zernike polynomials to design variables can be derived. This procedure effectively overcomes the bottleneck of computational cost in sensitivity analysis when using the finite difference method. Therefore, the topology optimization models, which usually have thousand or tens of thousands of design variables, can be implemented by using the objectives and design constraints directly based on Zernike coefficients. Meanwhile, within the frame of numerical finite element discretization, adaptive finite element basis functions and element numerical integrals can be implemented to solve structural deformation and Zernike coefficients accurately and efficiently. This algorithm is flexible to be applied to general structural topology optimization models with objectives or constraints being a reasonable linear combination of Zernike coefficients. The numerical examples illustrate that the algorithm can optimize Zernike coefficients effectively.

In order to realize the refractive index detection of substances such as aerogels and sevoflurane (an important component of anesthetic agents) in mid-infrared spectrum, and expand the detection range of the refractive index, a surface plasmon resonance sensor based on a dual-core photonic crystal fiber for low refractive index detection is proposed. The photonic crystal fiber is composed of two kinds of air holes with different sizes around the central air hole. The outermost large air hole directly contacts the substance to be measured by polishing the fiber, for real-time dectection of refractive index based on metal external coating can be achieved. The theoretical model was analyzed by the full-vector finite element method. The results demonstrate that in the refractive index range of 1.12~1.37, the operation wavelengths of the sensor are in mid-infrared spectrum between 2 505 nm and 3 181 nm. The maximum sensitivity is 12 000 nm/RIU, and the resolution is 8.33×10-6. The proposed sensor can adopt the mid-infrared band to achieve low refractive index detection, which has obtained a ultra wide detection range and a high sensitivity. It has a board application prospect in chemical detection, biomedical sensing, water environment monitoring and other fields.

Using the main absorption peak of methane molecule at the wavelength of 3.3 μm, an infrared methane sensor based on non-dispersed infrared spectroscopy was developed. The optical part of the sensor consists of a measuring LED with a peak wavelength of 3.3 μm, a reference LED with a peak wavelength of 2.7 μm, a photodiode with a cutoff wavelength of 3.6 μm and a spherical emitting surface. The circuit part mainly includes LED driving circuit, photosensitive signal processing circuit, temperature measuring circuit and microprocessor. The short pulse power supply control logic mode is adopted to reduce the power on time of the infrared light source, and the power consumption of the optical measuring device is reduced to 16 mW. The influence of temperature change on the measurement results of methane was studied experimentally, through data analysis and linear fitting, the temperature compensation algorithm formula was obtained. The experimental results of the sensor with compensation and detection system platform show that the average power consumption of the sensor is 23.56 mW, the influence of temperature change in the temperature range of -20~50℃ on the measured value is not more than 3% of the true value, humidity influence detection value is less than 4%, the response time is less than 25 s, the working stability time is more than 60 days, and the performance indexes meet or are better than the relevant requirements of AQ6211-2008 coal mine non-dispersive infrared methane transducer. Compared with the methane sensor based on the principle of thermal radiation infrared light source or laser detection, the power consumption of the infrared methane sensor based on double narrow-band LED is reduced by more than 70%, which can meet the technical requirements of low power consumption in portable and wireless applications.

The spectral distribution and time correlation of spontaneous parametric down-conversion were measured, and a radiometer based on spontaneous parametric down-conversion calibration was established. Based on the detection efficiency of spontaneous parametric down-conversion calibration, a radiance measurement scheme in terms of photon count was proposed, the radiometer can correct the degradation of its own response and observe the radiance of the target in real time. By using the experimental technique of inserting optical attenuator and changing detectors, the radiometer can correct its own degradation if observed radiance measurement results under the condition of optical degradation and electronics degradation. The results show that the consistency of observed radiance can meet 0.4%.The measurement results provide experimental basis for the application of spontaneous parametric down-conversion calibration to space platforms.

The traditional multivariate analysis method is the main method for quantitative modeling of LIBS spectral datasets, but the input dimension of the spectrum is relatively high. Reducing the dimension of the spectrum and extracting the characteristic spectral line in advance is needed by many algorithms, which results in partial loss of information and affects the accuracy. Aiming at this issue, a quantitative modeling method based on deep convolutional neural network inception is introduced, and the conventional 2D convolutional network is transformed into 1D form to realize full spectrum input and feature extraction of spectral information. Not only there is no need to reduce the dimension of the original spectrum in this method, but also it omits other preprocessing operations such as filtering. Through many experiments, when the number of training is 2 000, it has a good prediction result with no obvious overfitting phenomenon. Its average coefficient of determination (R2) is 0.957 9, and its root mean square error is reduced to 61.69% of those by Partial Least Squares Regression (PLSR). Compared with PLSR and the AlexNet deep learning method the proposed method both gets better results.

As the seed-injected lasers widely used in Brillouin Lidar area are too expensive and huge, the algorithm for a broadband Brillouin signal should be developed, which could accelerate the application of compact and economic diode lasers. A variation of Thomae's function was found during the signal processing of broadband Brillouin lidar, and the property of this function-reaching minimum value at the Brillouin frequency shift was used to recover the original signal spectrum and the corresponding frequency shift from a 1:1 superposition of pump light and Brillouin light spectra. Experiments using test data show nearly 100% accuracies for ideal cases; but for non-ideal cases, only when the noise level is less than -30 dB and the ratio of pump light to Brillouin light is less than 1.05 can this algorithm obtain accurate result.

The high dimension and huge data volume of hyperspectral remote sensing images and the complexity of surface feature lead to difficulty in distinguishing the anomaly pixel from the background. To solve these problems, an unsupervised nearest regularized subspace anomaly detection algorithm based on spectral space reconstruction is proposed. Firstly, in the process of band selection based on structure tensor, noise pixels are removed to obtain more effective bands. Then, the spectral space reconstruction is utilized to increase the absolute spectral distance between the background and the anomaly. Finally, to take full advantage of the spatial similarity information between background dictionaries, the spatial distance weight is introduced into the unsupervised nearest regularized subspace algorithm to improve the accuracy of linear representation.To validate the effectiveness of the proposed algorithm, experiments on four sets of real hyperspectral data are conducted, and the infulence of different parameters on the detection results is studied. Experimental results demonstrate that the proposed algorithm has a better detective performance than other anomaly detection algorithms.