The feedback control algorithm is the key integral part of an adaptive polarization mode dispersion (PMD) compensator. The particle swarm optimization (PSO) algorithm was introduced into the adaptive PMD compensation system, which showed the merits of rapid convergence to the global optimum, not being trapped in local sub-optima, robust to noise and capability of multi-degree of freedom control. The performance of two kinds of PSO algorithm, global version of PSO (GPSO) and local version of PSO (LPSO), was analyzed. The three kinds of topologic structures with which the PSO algorithms had 100% success probability were given. The experimental performances of GPSO and LPSO in compensating the first-order and second-order PMD showed that GPSO and LPSO were both qualitied when they are used in compensating the first-order PMD, but LPSO behaved better than GPSO when they are used in compensating the second-order PMD.

Using a beacon in the free space laser communication system is an important method of finishing the acquisition course.The far-field distribution of a beacon affects seriously the acquisition time and acquisition probability. In order to satisfy practical performance requirements, it is necessary to design the far-field distribution of a beacon. First, the impact of a far-field Gaussian distribution beacon on the acquisition course of laser communication is analysed and the far-field uniform beacon has some advantage compared with the far field Gaussian beacon in the case of transmitting beam jitter. One kind of far-field uniform beacon model is developed on the base of combination of semiconductor lasers. The structure of the model is described, the operating principle of model is analysed and the far-field distribution expression of the beam is derived. Experiment's data indicate that transmitting power of the model is more than 3.8 W, the degree of non-uniformity is less than 11% and the coupling efficiency is 54%, so the feasibility of the model is proved.

How to get the corresponding control signal according to the feedback information and respond timely are the key problems for the polarization mode dispersion (PMD) adaptive compensation technique. An adaptive PMD compensation experiment for 10 Gb/s fiber optical system is completed by using a multi-degree of freedom compensator with the power of specific frequency component in the signal spectrum as the feedback signal, with a novel adaptive tracking algorithm. The novel algorithm successfully solves the problem of trapping into local optima, efficiently eliminates the hysteresis of polarization controller (PC) and is adopted in the tracking course with no visible worsening of the feedback signal in the adaptive tracking compensation process. The long term testing with multi-times disturbances shows that the response time for the compensator to recover from a sudden disturbance is at the level of milliseconds.

An approach of fiber grating strain demodulation based on the principle of polarized-light interference is introduced, which can solve the so called cross sensitivity in the fiber Bragg grating sensor. To the yttrium vanadate (YVO4) crystal polarized-light interferometer, by choosing two different fiber Bragg gratings with a π rad phase difference, one as a strain sensor and the other for temperature compensating. The cross sensitivity can be conquered. The analysis of the strain demodulation resolution shows that with the decreasing of the crystal thickness the resolution increases. The experiment shows that when the thickness of the YVO4 crystal is 0.5mm the effect of cross sensitivity is reduced to 0.13 με/℃ which is 1.6% of the effect with only one fiber Bragg grating for strain measure and that when the thickness is 0.1 mm the cross sensitivity is reduced to 0.0067 με/℃ which is less than 0.08%. Reduced thickness of crystal benefits the resolution of strain measure but the machining of the small thickness crystal is difficult, and to solve this problem a new scheme of polarized-light interferometer with the structure of double layers crystal is analysed in the end.

Based on the full-vector finite element method, the birefringence in microstructure fiber with squeezed hexagonal lattice is numerically analyzed. The correlation between the birefringence and the structural parameters, incidence wavelength is obtained. The numerical results show that the birefringence is sensitive to wavelength, and can be achieved by a magnitude of 10-3 with a proper design and the fast and slow axes can swap as incidence wavelength is changed. To simulate the practical situation, each air hole diameter is assumed with a Gaussian distribution. The birefringence is found to be largely dependent on the variation of air hole diameter. In the statistical analyses, the correlation between the variation of air hole diameters and the birefringence in such microstructure fiber is obtained. One important result is that the non-intentional birefringence depends on the mean air hole diameter, and a increase of the size of the fiber structure the birefringence.

Encoders and decoders using 3-chip superstructured fiber Bragg gratings(SSFBGs) are designed, in which every encoder/decoder grating consists of three subgratings. The grating has an uniform amplitude refractive index level along its length, in which discrete (0,π/2,π,3π/2) jumps in phase are written into the grating at the boundaries of adjacent spatial chips. Quaternary encoding and uncoding can provide codes with desirable auto and cross-correlation characteristics with which the three-chip, 240-Gchip/s all optical encoding and decoding is achieved. As much as we know, this is the highest single channel rate obtained from fiber Bragg grating (FBG) based OCDM system. The agreement between experimental and calculated results indicates the feasibility of multi-phase all optical encoding and decoding using SSFBG. And such all optical encoding and decoding is required in optical code division multiplexing (OCDM) systems and packet-switched networks.

Three-dimensional imaging techniques are very attractive for many applications. A non-coherent visible three-dimensional imaging method is proposed based on the principle of coded aperture imaging. It is a two-step imaging method. In the first step, the object is recorded by a cameras array which the distribution of every camera in the array is spatially coded. In the second step the object images by the cameras array are synthesized into an image according to the position, encoded images are decoded by the optical reverse projection decoding method, the stratified surface images with various depth are reconstructed. Experiments have been done to test and verify the performance of the proposed method. A camera was used to captare the images of the object successively at each coded points. The code pattern is a NRA consisting 9 coded points. Object model is composed of only two stratified planes, and is placed at the position 1.5 m away from the cameras array. High SNR reconstructed stratified decoded images of the object are obtained using computational decoding program. Experimental results prove the proposed method is feasible and viable.

In order to simply obtain the phase of a surface in phase measuring profilometry (PMP), a fully automatic phase unwrapping method based on projected double spatial frequency fringes has been deduced. In this method, a series of formulas are deduced by analysing the relationship between the orders of and phase of the two fringes. Since the phase unwrapping is done on each point separately, the error propagation is avoided. Experiment has proved this method feasible, and a good measurement result is obtained.

Absolute phase measurement and phase reconstruction to object surfaces with complex geometric and topologic shapes still remain difficult for phase mapping based on three-dimensional (3D) sensing. Recently, an approach based on temporal phase unwrapping has been proposed to handle this task. However, it is found from the experiment that temporal phase unwrapping imposed severe restriction to the illumination system of structured light. If the required condition is not satisfied, the temporal phase unwrapping will cause severe noise in the reconstructed phase map. To overcome this problem, a new approach based on spatial and temporal phase unwrapping is presented, which is constructed with piece-wise functions by taking both temporal and spatial factors into account during the process of phase reconstruction. This leads to the effective phase reconstruction even under the situation that the number of fringes is not exact the integer exponential of power 2, which is an essential restriction imposed on previous temporal phase unwrapping algorithm. The approach eliminates the problem of phase ambiguity at the boundaries of phase jump corresponding to the wrapped phase map and therefore provides a more efficient solution to the phase reconstruction of 3D object surfaces with complex geometric and topologic shapes or to a scene with isolated objects. Experimental results have demonstrated the proposed approach.

The polarization-based holography of Fresnel zone plate (FZP) convolution without motion is proposed. Its principle is the same as that of the FZP scanning holography. But its configuration is modified significantly. First, the uniformly distributed source is used to form the superposition of the FZP projects on the object so as to realize the convolution operation. Thus the time-consuming problem and the unstability of the system due to the mechanical scanning are resolved. Furthermore, the imaging system is more stable than before. Secondly, the CCD camera is used to replace the photomultiplier tube to collect the whole hologram in real time. The degree of polarization is investigated experimentally when the linearly and circularly polarized light propagate respectively through the scattering media of intralipid. The results demonstrate that the polarization of the circularly polarized light is maintained better than that of the linearly polarized light. By using of the imaging system of the polarization-based holography of FZP convolution without motion, the experiments on a metal wire with a diameter of 0.4 mm embedded in the intralipid solution with concentration of 1% (volume fraction) and depth of 2 cm are implemented. The experimental results testify the principle of polarization-based holography of FZP convolution without motion is feasible.

Based on the power-series' and Zernike circle polynomials' decomposition theory of wavefront aberration, the relationship between Zernike circle polynomials and Seidel polynomials is illuminated. The correlation Zernike circle polynomials on the annular domain is analyzed theoratically, and the effect of central obscuration interferograms fitted with Zernike circle polynomials on calculating Seidel aberrations is emphasized. The experimental results are consistent with the theortical analysis,and a simple and intuitionistic enactment method of error tolerance is provided. With the increase of obscuration ratio, it is found that the error of Seidel coefficients is increased, and its changing rule dependes on the aberration type and magnitude of the optical element under test. When the obscuration ratio reachs a certain threshold value, the error curves would greatly change. For achieving the accurate Seidel coefficients, the order number of the Zernike circle polynomial fitting should be selected apporiately, and the Zernike annular polynomials should be selected in a large obscuration-ratio case. Additionaly, the effect of the obscuration ratio on the tilt and astigmatism along Seidel aberrations is less than that on the coma, focus and spherical terms.

Studying three-dimensional dynamic parameters of an insect flight requires reconstructing from 2D images, and two or more cameras are needed to capture images from different angles at the same time in the conventional method. A new method has been introduced. By adding four plane mirrors onto a high-speed camera, two different images are projected on the same frame, which is equivalent to that obtained using two cameras. Through calibrating the two optical paths, the external and intrinsic parameters can be computed, and then the three-dimensional information of fly's wings by using the mechanism of machine vision can be constructed. In contrast with conventional methods it is simple, inexpensive for three-dimensional reconstruction; moreover, there is no assumptions during reconstruction.

Tunable diode laser absorption spectroscopy (TDLAS) is a new method to detect trace gas qualitively or quantitavely, based on the tunable characteristic of the diode laser, by obtaing the absorption spectroscopy in the characteristic absorption region. For the significant advantage in sensitivity， selectivity and rapidity of response, it has been widely used in atmospheric trace gas measurements and industrial control. In air trace gas measurements, TDLAS should be combined with the long absorption path technique due to the low concentration. Based on the combination of TDLAS and 27 m absorption path after 108 times reflection a portable absorption spectrometer is developed for monitoring the ambient methane, and the detection limit is below 1×10-7 by second harmonic detection which is usually used to detect low signals. The system is tested by the gas with different concentration， and the result is very well.

To obtain the size of the pinhole in the spatial filter of the CPA system, Fourier transform and laser multimode theories are applied. And by means of reasonably mathematical approximation and numerical integration, the radial spatial frequency spectrum and the spatial close frequencies of the basic transverse mode and main low-order transverse modes are calculated. The feasibility of applying the close frequency of the basic transverse mode to determining the close frequency of the spatial filter is also discussed, thus the diameter of the pinhole is determined and a widely significant result is achieved. In conclusion, the pinhole, with the optimal diameter of about 6.93 times the diffraction limit length, can fairly filter the high frequencies of most transverse modes except for the TEM00 mode, so that the beam quality is ameliorated and this result is well verified in experiments. Besides, a beam quality-improving factor called “B” based on M2 factor to help describe the performance of the pinhole-lens spatial filter is introduced.

Passively Q-switched mode-locking Nd∶Gd0.42Y0.58VO4 laser is successfully demonstrated by using a piece of GaAs crystal grown at low temperature (LT-GaAs) as the passively saturated absorber as well as the output coupler. The fundamental properties of Nd∶Gd0.42Y0.58VO4 laser are investigated. At transmission of 10% and cavity length of 40 mm, the maximum average output power of 3.78 W is obtained when the incident laser pumping power is 8.6 W, which corresponds to an optical-optical conversion efficiency of 43.9%. The output performance is then tested for the passively Q-switched Nd∶Gd0.42Y0.58VO4 mixed crystal laser. The threshold power for Q-switching and Q-switching mode-locked (QML) are about 2 W and 3.7 W respectively. At the incident laser pumping power of 8.6 W, Q-switching mode-locking pulse with modulation depth more than 70% is available. The Q-switched envelope train with repetition rate of 670 kHz is obtained with pulse duration of 180ns. The average output power and the optical-optical conversion efficiency are 1.35 W and 15.7%, respectively.

A table-top Ti∶sapphire laser system with two-stage amplifier was built based on TSA-25 regeneration amplifier. Pumped by 0.6 J and 1.6 J, 532 nm, 10 Hz of Nd∶YAG lasers, the output energy reached 500 mJ and a pulse with duration of less than 41 fs was generated by using a single grating compressor, the energy convevsion efficiency of the compressor was 63% and the peak power reached 7.6 TW. By the experiments on the different absorption of crystals for 532 nm, pump-to-IR efficiency with pump fluence and passes of amplification, the system was optimized with an extraction efficiency of power amplifier ～32%, compacted less than 10 m2.

The cat's eye resonator is designed with a cat's eye reflector as the resonator mirror. The configuration of the cat's eye reflector is described. The relation between the focal length of the cat's eye reflector and the distance between the convex lens and the concave mirror of the cat's eye reflector is indicated. Through controlling that distance in the cat's eye reflector, laser transverse mode is selected to ensure fundamental transverse mode output. In Newton-Cotes method based on the diffractive integral equation of the caser, laser transverse mode of the cat's eye cavity laser under variable parameters is analyzed. The eigenvalue ηmn and power loss which every transverse mode corresponds with are calculated. The parameter assuring fundamental transverse mode output obtained is well accordant to the experimental results. Therefore, a new convenient on-line means for the control and selection of the laser transverse mode is discovered.

Highly intense backward first-Stokes (BS1) in CH4 is generated by using a single longitudinal mode second-harmonic Nd∶YAG laser [532 nm, linewidth Δνp <100 MHz， pulse-width (FWHM) 6.5 ns] as the pump source, which is radically different from previous reports where the output from a longer pulse-width (30 ns) pump laser is mainly stimulated Brillouin scattering (SBS). This is because that in authors' experiments the pulse-width is comparable with the SBS phonon lifetime that results in a transient SBS process. According to authors' calculation, the gain coefficient for the transient SBS in the present case is already somewhat lower than that of BS1, and SBS is suppressed by BS1. Under a CH4 pressure of 1.1 MPa, a pump laser repetition rate 2 Hz and laser pulse energy 95 mJ, the quantum conversion efficiency of BS1 is as high as 73%. Due to the relaxation oscillation, the BS1 pulses are narrowed to about 1.2 ns. As a result, the BS1 peak power turns out to be 2.7 times that of the pump. Furthermore, the beam quality of BS1 is much better than that of the pump. The relaxation oscillation in the experimental BS1 temporal waveform is well reproduced by a calculation using a quasi-two dimensional computer program.

To form the theory and method of hybrid refractive-diffractive (HRD) optical design that can join to the Chinese conventional optical design system, the primary aberration theory expressed by PWC about HRD systems has been investigated, and the function relation between Seidel sums and P, W, C, between ∞, ∞, ∞ and the diffractive lens structure have been established. Using the PWC primary aberration theory and the ultra-high index design method, the original structure of the achromatic HRD medium-power Lister-type microscope objectives has been obtained. The result shows that its design of Seidel sums tallies with the theoretic value, so the PWC primary aberration theory and the ultra-high index design method about HRD systems proved to be effective. Comparing with the conventional two doublet design, it is shown our hybrid objective has a larger work distance and a higher imaging quality, and is suitable for mass production because the front is a singlet plastic lens.

In space laser communication, the composite axis system (CAS), consisting of the coarse tracking system (CTS) and the fine tracking system (FTS), accomplishs the wide ranging and high accurate tracking tasks. The digital simulation of CAS based on Z transform is given to study the bandwidth design of the CAS. According to the trajectory of the satellite and the mechanical error of the CTS, the bandwidth design of CTS is obtained. Moreover, the cooperation of CTS and FTS is given, and the bandwidth design of FTS is gained. At last, the bandwidth design of CAS with disturbance of medium and high frequency is presented. The results above present the theoretical basis for the design of CAS control system.

A polarization independent high speed electro-optical switch by using quadratic electro-optic effect of lanthanum-doped lead zirconium titanate (PLZT) ceramic incorporated in a Sagnac loop is described. Analysis by Jones matrix method shows the output property of the fiber loop mirror and the relation of its extinction ratio and parameters of the configuration. Experiments show its typical performances of 25 dB extinction ratio and less than 3 μs response time, and give Kerr coefficient of the lanthanum-doped lead zirconium titanate κ～1.1×10-16 m2/V2. Experiments also show good stability of the device without any special measures, which is attributed to the mutual optical paths in the fiber loop for the two beams of interferometer, and is regarded superior than other structures like Mach-Zehnder interferometer.

An optical microcavity is a structure with at least one dimension on the order of an optical wavelength. The interaction between material and light within microcavity was changed greatly compared with the condition of free space. Cavity effects such as spectral narrowing and intensity enhancement of spontaneous emission have been observed. These effects can be used to improve the electroluminesence (EL) performances of organic light emitting device (OLED). The microcavity structure of green light OLED was optimized. The device was fabricated and studied with a structure of Glass/DBR/ITO/NPB/Alq:Rubrene/Alq/MgAg. The maximum luminance of 40100 cd/m2, maximum luminous efficiency of 6.44 cd/A, and a full width at half maximum (FWHM )of 28nm were obtained in microcavity OLED, whereas the maximum luminance of a noncavity device is 22580 cd/m2, maximum luminous efficiency2.98 cd/A, and FWHM120 nm. The EL peak intensity of the microcavity device is 4.2 times stronger than that of noncavity device of the same current density. It shows that introduction of microcavity into OLED improves not only the color purity, but also the luminous efficiency and luminance of the device.

To improve the quality of the spectral instruments, two ellipsoidal mirrors and an ellipsoidal aberration-reduced grating are used. Reflected by the first ellipsoidal mirror, the wave front of the light is changed, and after passing the grating the light is focused on different positions of the detector according to the difference of the wavelength by the second mirror. The spectral instruments are with complex aberration and imaging characteristics. According to Fermat principle, consulting the second generation varied-line-space grating's fabrication optical layout and theory, the coordinate system is setup, the angles of incidence and reflection are defined, the elements are located, and the basic geometrical relation is established. The optical path function is divided into two parts: one is associated with the diffraction, and the other is associated with the optical path change by the geometrical position. After the power series of the above two optical path are expanded, the analytical expressions of the high-order aberration are derived, which are useful in the plane, cylinder, spherical gratings, as well as reflector. And the expressions of the aberrations and the optical layout can be used in other spectral instruments.

The optical performance of the fused-taper fiber device is determined by microstructure and morphology of the fused taper region, which are determined by technological conditions. In order to find out the effect of technological conditions on microstructure and morphology, experiment has been carried out. Taking the couplers manufactured at different drawing velocities as samples, the wave number of the fused regions and taper regions is measured with micro infrared spectrometer, and the morphology of the same point is recorded with scanning electron microscope (SEM). It is discovered that there are two characteristic peaks: one is at about 1100 cm-1, the other at about 810 cm-1. The wave number of the former peak is highest at the taper region, the one at the fused region followed, and the one of bare fiber lowest. The 1100 cm-1 characteristic peak shifts to higher wave number while the drawing velocity increases. In the taper region, cracks are discovered. And the cracks are more obvious while drawing velocity increases. In the fused region, there appear crystallines on the surface of the coupler, and the crystalline grain becomes smaller while the drawing velocity increases. A higher quality of coupler cannot be obtained until less change of the microstructure and less defects under the proper drawing velocity (here is 150 μｍ/s) are realized.

A type of two-dimensional photonic crystals with a quasi-honeycomb structure, which is formed by introducing two consistent dielectric cylinders into a triangular lattice, is investigated here. The relationship of its band gaps with the position of cylinders is numerically calculated and the photonic energy band is evaluated by the frequency domain method. The calculated results demonstrate the existence of large band gaps in quasi-honeycomb-structured two-dimensional photonic crystal. At a certain filling ratio, the largest absolute bandgap can be obtained by optimizing the position of the cylinders introduced.

By expanding the circular function into a linear superposition of a set of complex Gaussian functions, the approximative analytic expression for the far-field of Gaussian beam with phase aberrations diffracted by a circular aperture is obtained. For different parameters, the diffraction patterns of the expression of the Fraunhofer diffraction are identical, indicating the correctness of the expression. For its relatively simple form, the calculation becomes more convenient. Based on these, the far-field divergence of Gaussian beam with phase aberrations is completely studied. The divergence of the certain beam does not change with the transmission distance, but with the beam-waist radius and the phase aberrations under the limit of circular aperture, and the result indicates that two parameters are the key factors affecting the divergence.

Orthogonal signal correction (OSC) and net analyte preprocessing (NAP) were respectively used to pretreat the near infrared (NIR) spectra of apples ranging from 1300 nm to 2100 nm, and the models of sugar content were developed from these two pretreated spectra by the partial least square (OSC/PLS models and NAP/PLS models). Results showed that the optimal number of factors used in OSC/PLS models and NAP/PLS models would reduce as the number of OSC factors and NAP factors increased on by one, even to 1 finally. The best OSC/PLS model with 2 PLS factors was obtained when 10 OSC factors were used in pretreatment; and the best NAP/PLS model with 1 PLS factor was obtained when 11 NAP factors were used in pretreatment. Although the best OSC/PLS model and the best NAP/PLS model didn't improve precision to a great extent, they needed fewer factors and became more parsimonious, compared with the model before NAP pretreatment. In a conclusion, OSC and NAP pretreatment could simplify the prediction model of sugar content.

The flurorescecne spectra of lycopene in different solvents including hexane, ethyl acetate, terrahydrofuran, chloroform, acetone, benzene and of lycopene-terrahydrofuran solution with different concentrations are measured by a fluorescence spectrophotometer. It is found that the wavelength of fluorescence spectra maximum (λmax) change along with the variation of the polarity of solvent. The transition energies (ET) of lycopene calculated in different solvents are 220.5 kJ/mol, 218.2 kJ/mol, 215.6 kJ/mol, 215.2 kJ/mol, 214.9 kJ/mol, 211.5 kJ/mol. When the concentration of lycopene-terrahydrofuran solution is lower than 50 μg/ml, the intensity of fluorescence increases with the increase of solution concentration, when the concentration is higher than 50 μg/ml, the intensity of fluorescence decreases with the increase of the concentration, which is caused by the interaction between molecules in the excited state and the ground state. Three weak peaks appear in the fluorescence spectra of solutions when the concentration of which is lower than 80 μg/ml. The corresponding transition energies can be calculated as below: E(1)T=278.2 kJ/mol、E(2)T=260.2 kJ/mol and E(3)T=239.3 kJ/mol.

That the support vector machine network is applied to recognize the nonlinear fluorescence spectrum of impurities of different concentrations in air is proposed. Because the number of spectrum channel of the original spectrum data is large, it is cleaned up and compressed through wavelet trausform firstly, and then the principal component analysis (PCA) is used to extract the character information twice in series. It not only ensures the character of original nonlinear fluorescence spectrum, but also compresses the data number the nonlinear fluorescence spectrum from 3979 to 514, and extracts 9 principal components, which reduces the number of the input vector and improves the training speed of the network. The simulation results show that the correct recognition rates for both training spectrum samples and unlearned test spectrum samples reach 100%. So, the training and testing speed is fast enough to monitor the atmospherical impurity in air in real time.

The technological process of polymer nanoporous antireflection film fabrication is discussed in detail. And the effects of molecular weight of the polymer, experimental temperature, humidity and the volatility of common solvent on the polymer nanoporous antireflection film are also analyzed. Larger polymer's molecular weight, lower condition temperature, higher relative humidity and lower solvent's volatility will lead to increasing the porous size and the scattering loss. So the transmission ratio of the film will be decreased. Lower molecular weight polymer (25 ℃), relative humidity lower than 30%， and lower boiling point solvent such as tetrahydrofuran are the key points for a small porous size and high transmission ratio.

The superprism effect was demonstrated experimentally in a thin film Fabry-Pérot filter based on its large group delay at the wavelength of peak transmittance. The group velocilty delay and the shift caused by the spatial dispersion were numerically simulated by characteristic matrix method used widely in optical thin film theory. The device was fabricated and tested and the results showed that the maximum shift of spatial dispersion due to the superprism effect reached 65 μm at the wavelength of peak transmittance, which consisted well with the design. Compared with the conventional grating element and prism, the device had smaller size and higher spatial angular dispersion, nearly 30°/nm in the range from 784.5 nm to 786.5 nm according to the measurement.