The measurement facility based on parametric down-conversion technique was established to measure the quantum efficiency of the near-infrared band single-photon detector. The correlated photons were generated via spontaneous parametric down-conversion in a periodically poled potassium titanyl phosphate crystal pumped by a 518 nm picosecond laser. The detection efficiency of Si-avalanche photodiode and InGaAs-avalanche photodiode single-photon detectors were measured via correlated photon pairs at 778 nm and 1 550 nm, respectively, and the measrement uncertainty was analyzed. The experiment results indicated that the facility can measuer near inferad single-photon detectors quantum efficiency with 1% measrement uncertainty.

To obtain the Depth Of Interaction (DOI) information for Positron Emission Tomography (PET), we proposed a novel detector design based on the light sharing technique. The detector used scintillators and silicon photomultipliers (SiPMs) with one to one coupling, employed the serpentine-light-path reflector scheme and the single-end Anger logic readout technique. Monte Carlo simulations were carried out to verify the design via GATE toolkit. An 8×1 LSO array with a crystal size of 3.1×3.1×20 mm3 was built. A flood source for generating position look-up table and five beam sources irradiating at different depths of the crystals were simulated and then calculated. The result shows that the average DOI resolution is 3.2 mm (ranging from 1 to 7 mm). In conclusion, we propose a novel detector design based on a serpentine-light-path scheme to realize DOI decoding which is a low-cost and practical direction for clinical PET systems.

In order to realize a Long-wave Infrared (LWIR) optical wireless communication, a long-wave infrared optical wireless communication system model based on pulse width modulation was estabished, the performance of this system was analyzed, and an expression of bit error rate was proposed. An experimental system was designed, the change of average output optical power of CO2 laser was analyzed, the average of received laser pulse width and random variation of pulse width affected by the noise under different duty ratios were analyzed. The relationship between the average of output optical power and pulse width with duty ratio was obtained, and the distribution law of pulse width was also obtained, and then the experimental results and theory analysis results were compared, based on this, the optimal judgment threshold of pulse width and the relationship between system working parameters with (Bit Error Rate)BER were obtained. It is proved that a long-wave infrared optical wireless communication can be achieved based on CO2 laser by the pulse width modulation.

A theoretical model of the eccentric core fiber was established based on traditional optical transmission theory. By using conformal mapping, the asymmetric three-layer structure of the eccentric core fiber was converted into a three-layer structure of coaxial symmetry. Characteristic equation about propagation constant was obtained. Bending loss formula of eccentric fiber was given. The relationship between bending loss and bending radius of the eccentric fiber was analyzed by simulation. When the bending radius reaches 3 mm, the bending loss is almost zero. Meanwhile, the relationship between bending loss and eccentricity distance change was obtained. Optical model of eccentric fiber bending was established by using BeamPROP module in Rsoft software. Mode field distribution in the curved eccentric core fiber was simulated. The simulation result shows that bending loss of the eccentric optical fiber in the same direction as that of the core deflection center axis is smaller than that of the opposite direction.

An ultra-high speed imaging method based on normal CCD sensor was proposed. The light sensitive area of the chip was covered by the mask and turned into storage area.The influence of extra time that CCD charge transfer took was eliminated. So the frame rate of masked CCD could achieve million frames per second. The different types of masks were introduced, and the restoration method of masked images was discussed. The method of strip-hole array mask was used to build the ultra-high speed imaging prototype base on normal CCD, and the performance was verfied by xenon lamp. After the location of mask pixel center and mask direction are both known, the masked images is finally separated and reformed into several low resolution result images. Results show that a performance of 2 million frames per second is achieved and 14 images with a resolution of 79 pixel×79 pixel are got.

In order to realize the push-staring imaging with a certain overlap rate of the long-strip region along the track of plane array camera, a pendulum type searching imaging model was designed with a high resolution CMOS sensor. Based on the region division of the long-strip layer by layer and the construction of gradual change of imaging overlapping rate, the real-time push-staring pointing gesture of the statellite and the matching parameters of pendulum type searching imaging were calculated. Besides, a PD(Proportional and Derivative) controller with multiple small approaching was used to analyze the gesture control precision of searching imaging and image shift mismatch quantity. Finally, a ground geometric scaling simulation experiment of the long-strip pendulum type searching imaging of the system of P5 surface LED(Light Emitting Diode) targets was carried out through the CMOS(Complementary Metal Oxide Semiconductor) principle prototype and the small satellite gesture control simulation platform. The results show that, the area of pendulum type searching imaging is larger about four times than that of the staring imaging when the inter frame overlap ratio is more than 85%. The gesture control accuracy is better than 0.05°, the gesture stability is better than 0.003°/s and the corresponding transfer function can reach up to 0.141 1 when the control frequency is 10 Hz.

Doppler Asymmetric Spatial Heterodyne(DASH) spectroscopy is a new hyperspectral remote detection technology, which requires the data processing technology should also have corresponding super high precision. From the viewpoint of digital signal processing, an adaptive frequency tracking method was proposed, which is based on the spatial frequency of the signal to compensate the phase shift of the fringe signal, and recursive iterate to the phase information most close to the real value. Verified by simulation experiment contrast, the results show that, under the low intensity noise condition, the frequency tracking algorithm can improve the accuracy of signal frequency and phase extraction by about more than 100 times to compare with the traditional Fourier transform method, which proves it can effectively reduce the system error of DASH spectroscopy.

To decrease CCD dark current noise and improve the signal to noise ratio in airborne ultraviolet imaging spectrograph, CCD working temperature should be decreased as CCD is temperature sensitive device. So CCD cooling circuit was designed by adoping analog PID algorithm, and its parameters were determined by Ziegler-Nicholas method. The max cooling rate is less than 5℃/min and its temperature stability is about ±0.05℃, satisfing the requirements of maximum cooling temperature difference.The cooling circuit has been tested on airborne system. The experimental results show that change of temeperature could not affect the cooling results. And after reaching the target refrigeration temperature of －20℃, dark background noise of spectral dimension CCD detector is 1 702 on average and dark signal response non-uniform decreases to 0.5%. The results meet the requirements of spectral data.

A fast three-dimensional object measurement method was proposed by combining the Fourier profilometry with the phase shifting profilometry. By accurately extracting the mean envelope of the fringe in time domain and simply filtering in frequency domain, the fringe can be pre-processed. Combined with the two-step phase-shift algorithm base on 90° phase shift, an accurate phase value can be obtained from two fringe patterns. The results of simulation and experiment of comparing the proposed method with the existing methods show that this method can achieve a higher accuracy, which the RMS (Root Mean Square) is less than 3×10－3 rad. The proposed method has the following advantages of good robustness for its insensitive to noise and surface mutation, fast operation and high accuracy. Therefore, it will have a high application value in the field of fast measurement.

Using Opto-electronic detection technology, marking miRNA and isothermal rolling circle amplification and detecting markers which excite fluorescence intensity, we have designed a new portable device for the rapid detection of miRNA, the recognition system of characteristic fluorescence analysis was established eventually. we have got the point that there is the linear relationship (R2=0.999 1) between miRNA concentration and fluorescence intensity at the time that the miRNA concentration range the instrument can measure is in the range of 0.01~0.1 μmol by changing the excitation light intensity, miRNA reagent concentration and other parameters, meanwhile, we also know that it is 7 copies that the lowest values the instrument can measure in theory.

In order to meet the requirement that the dynamics star simulators based on liquid crystal on silicon splicing reduce the stray light in the background, the traditional optical engine were optimized. A method of combining multiple polarization beam splitter into one components was proposed to optimize optical engine, and the design of the optical-mechanical structure was made. This paper expounded the design scheme of lighting system for optical engine, and we elaborated the method of reducing the field-of-view angle and increasing the uniformity. The lighting optical system was simulated by Tracepro, and the feasibility of the programme was verified. According to the results we can get that the stray light irradiance of the dynamic star simulator was reduced by 2.93 times after optimization, hence, the optimized optical engine effectively suppresses the stray light of background, and improved the consistency of the contrast ratio between the two pieces of liquid crystal on silicon.

In order to realized the rapid and accurate calibrating of the standard scattering plate uesd in calibrating visibility meter, calibrateing the simulated atmospheric visibility by measuring the scattering coefficient of multiple scattering angle of the standard scattering plate, and the calibration system was calibrated. According to the composition and working principle of the calibration system, the calibration method was determined, the calibration chain of the calibration system was established, and the calibration method for each step of the the calibration chain was designed. By calibrating the standard scattering plate with a given scattering coefficient, the veracity of the calibration result of the calibration system and the correctness of the calibration method were verified. The experimental results prove that the error of the calibration results of scattering coefficient of standard scattering plate is 7.93%, the error of the simulating visibility of the standard scatting plate calibrated by the calibration system is 8.61%, satisfied the requirements of the standard scattering plate used in calibrating visibility meter.

Based on the period-one dynamical state emerging in a 1 550 nm vertical-cavity surface-emitting laser with parallel optical injection, high-quality photonic microwave was acquired after further introducing optoelectronic negative feedback. The experimental results show that, under parallel polarized optically injection, 1 550 nm vertical-cavity surface-emitting laser can exhibit rich nonlinear dynamics behaviors such as stable injection locking, period-one, period-two and chaos through adjusting injection strength and frequency detuning. Under suitable injection condition, a parallel polarized optically injected 1 550 nm vertical-cavity surface-emitting laser can generate a photonic-microwave signal with a single sideband optical spectrum structure and a frequency over 10 GHz, but the linewidth of the microwave signal is relatively wide (on the order of MHz). After further introducing optoelectronic negative feedback, the linewidth of the microwave signal can be narrowed to about 100 kHz (reduced to more than two orders of magnitude) by choosing proper photoelectric feedback intensity. Furthermore, under the case that the feedback strength is set at the optimized value, the frequency of the narrow-linewidth microwave signal can be tuned continuously within a certain range through simply varying the injection strength.

A state equation of absorptive optical bistable system driven by colored multiplicative noise was built by using the unified colored noise approximation. The results were discussed and compared with the system with white noise. The results show that, when the multiplicative noise and the additive noise are in positive correlation, the area of optical bistability broadens outstandingly, that also is the hysteresis cycle width is broaden with the increasing of self-correlation time τ; whereas when the multiplicative noise and additive noise are in negative correlation, and only when the multiplicative noise strength is weak, the change of self-correlation time τ can induce the change of the area of optical bistability; when the self-correlation time τ of multiplicative noise is equal zero, the model of the system with colored noise falls back to the model of the system with white noise.

In order to detect the states of multiple switches simultaneously, a multi-switch states detection system based on a semiconductor laser with optical feedbacks is presented. The system consists of a semiconductor laser, multi-switch sensing network, data acquisition and processing unit. The sensing network includes multiple feedback optical circuits made of couplers, optical fibers and intensity reflection optical switches. When all switches are in ‘off’ state, the system has no feedback. Hence the laser outputs continuous light. Once some a switch is in ‘on’ state the feedback light makes the system work at chaotic state. By comparing the sub-peak values, corresponding to each switch position, of the autocorrelation curve of the signal out of the system with a preset threshold value, the switch state can be determined. The interference sub-peaks are eliminated by programing to set the values of sub-peaks in other locations as zero. An experimental system is constructed in which the sensing network is composed of three switches in series. The system is detected in different switch states respectively. The results confirm that the system can accurately detect the states of single or multiple switches.

In order to analyze the far-field propagation characteristics of the new non-diffractive Airy beams deeply in theory on the condition of finite energy, firstly, the one-dimensional paraxial wave equation which decides the propagation characteristics of light waves in free space was analyzed by taking advantage of the Fourier analysis method. Moreover, the special characteristics of Airy function and the initial condition of finite energy with exponential decay were also taken into consideration. As a result, the analytical solution of the wave equation can describe the propagation characteristics of one-dimensional Airy beams accurately in free space under the condition of finite energy. Then, the propagation characteristics of one-dimensional and two-dimensional Airy beams were studied by utilizing the acquired analytical solution. What′s more, the influence of various parameters on non-diffractive propagation and transverse acceleration characteristics of Airy beams were studied in detail. Study results indicate the non-diffractive propagation distances of two-dimensional Airy beams are 1014, 624, 455, 338, 193 mm separately when the arbitrary transverse scale is 100 μm and the decay factors are 0.03, 0.05, 0.07, 0.1, 0.2, separately. The non-diffractive propagation distances will become longer if the decay factor becomes smaller when the arbitrary transverse scale keeps invariant. Moreover, the transverse acceleration of Airy beams will become larger with the decrease of arbitrary transverse scale when the decay factor keeps invariant. The adopted study method is also fit for analysis of Airy beams′ propagation characteristics in the media with refractive index.

In order to improve the diffraction efficiency, a diffractive telescope system with 300 mm aperture was designed and fabricated. The primary lens of the proposed system is made form a four-step phase Fresnel zone plate which is fabricated on fused silica substrate by laser direct writing and Ar ion beam etching. The diffraction efficiency of primary lens of the diffractive system was tested. The tested results show that the diffraction efficiency of the diffractive lens at wavelength 632.8 nm is 66.3%, about 82% of the theoretical value. The light path of diffraction imaging system was built, and the imaging performance was tested by both the star image test and the resolution test. The diameter of the star image is 44 μm and the limit resolution is 84 lp/mm, which are very close to the theoretical values, showing the diffractive system has a good imaging performance.

According to the disadvantages of traditional Foucault test, the wavefront retrieval technology of off-axis aspheric surfaces is proposed. According to the mathematical analysis of Foucault test, a mathematical model is established and the wavefront integrating algorithm is proposed to reconstruct wavefront of the off-axis aspherical surface through integrating gray values of two Foucault pattern, removing tilt of wavefront map, and wavefront integration etc. When the circles of confusion in two directions reach to 0.152 mm and 0.284 mm, the surface error tested by interferometer is that PV value is 1.110 μm and RMS value is 0.194 μm, and the wavefront patterns are consistent. Experiment proves that the technology of wavefront retrieval is correct and can be used to guide the manufacture of off-axis aspherical surface until to be coherent with interferometer test of the fine grinding stage.

According to the characteristics of too large root mean-square value acceleration response of random vibration at edges of rectangular space mirror with large aspect ratio, this paper propose a topological optimization based on ground structure method to design the structure of rectangular space mirror, with the goal for minimizing the acceleration root-mean-square values of random vibration. Firstly, after the finite element method analysis to the initial rectangular space mirror structure, the author find that the acceleration root-mean-square values of random vibration at the edge of mirror is too large to meet the requirements of design. Based on continuum structural topology optimization thought, by using the acceleration root-mean-square values of random vibration at rectangular space mirror's edge as an optimization object, the optical aberration PV and the primary frequency of mirror as the design restraint, the rib plate type ground structure of mirror as the design variable, a topology optimization function is established. Futhermore, the new mirror based on the structure obtained by topology optimization was designed. Finally, by the finite element analysis and vibration test, the new structure of mirror is proved with a good dynamic performance and lightweight rate. Especially, mirror mass has reduced 13%, and the acceleration root-mean-square values of random vibration at mirror’s edge in Z direction reduced by 58%. The results demonstrate the validity of proposed approach properly.

A kind of hybrid surface plasmonic waveguide working in the terahertz frequency range and based on un-flat substrate structure was proposed. The relative permittivity of InSb changed with the working frequency was investigated. The mode properties of the proposed structure and two other un-flat substrate structures, such as the effective mode area, propagation length, figure of merit and energy ratio, were investigated in detail by varying with the working frequency and geometrical parameters. The results show that the proposed structure can reach the same propagation length with the traditional structure, while the mode area of the proposed structure is smaller by more than an order of magnitude compared with that of the traditional hybrid waveguide, which indicates stronger mode confinement. The proposed waveguide satisfies the requirement of the application in high density circuits in the terahertz range.

A tunable long-period waveguide grating coupler based on LiNbO3 was proposed. The light can be coupled effectively from the input waveguide to output waveguide through the cladding because of the long period gratings′ unique feature. The effective refractive indexes of the core and cladding of the waveguide grating will change accordingly as a result of electro-optic effect when a tunable voltage is applied, making the resonance wavelength and the coupling efficiency tuned by controlling the applied tuning voltage. The effects of coupler length and the grating period on bandwidth and the coupling efficiency tuning range, and the influence of the waveguide size on the resonant wavelength tuning sensitivity were analyzed. The results show that the shorter grating period and the greater coupler length lead to the narrower bandwith and the greater coupling efficiency tuning range. In addition, the resonance wavelength tuning sensitivity decreases with the increase of waveguide width, and the influence of thickness on the resonance wavelength tuning sensitivity can be ignored. For the coupler with the grating period of 94 μm and the coupler length of 3.52 cm, the simulation results show that the resonance wavelength tuning sensitivity can reach 26.2 pm/V, and the bandwith of 3 dB as 4.5 nm, when the applied voltage varying from 0 to 200 V, the resonant wavelength change 5.24nm, and the coupling efficiency can be tuning from 1 to 0.15.

The multiple quantum well structure for 850nm vertical cavity surface emitting laser device based on GaAs/AlxGa1－xAs has been design in this paper. The major process parameters of Low Press-Metal Organic Chemical Vapor Deposition (LP-MOCVD), such as the growth temperature, the reaction chamber pressure, total carrier gas flow rate and so on, have been optimized and the growth of complete epitaxial structure has been carried out. The experimental results showed that components of the GaAs/AlxGa1－xAs multiple quantum well structure devices was obtained under the condition of 700℃, the best composition x is 0.24 by the Photoluminescence (PL) spectrum comparison test and a good surface morphology of the multiple quantum well structure had been obtained. The optimum growth rategrowth rate was 0.34~0.511 nm/s ultimately.

In order to fabricate the phosphor-free InGaN/GaN multiple quantum wells (MQWs) with white light emission, the GaN microfacets with trapezoidal structure were grown using SiO2 stripe mask patterns, and then the InGaN/GaN multiple quantum wells (MQWs) were regrown on the GaN microfacets, forming dual-wavelength emissions in a single chip. The results indicate that, the trapezoidal GaN microfacet is composed of (0001) and (11-22) planes, which is attributed to the differences of surface polarity and surface energy. Also, the different color emissions on (0001) and (11-22) planes are due to the different diffusion rates of In and Ga adatoms; this property allows microfacet MQWs emit blue light (emission peak at 420 nm) from the (11-22) plane and yellow light (emission peak at 525 nm) from the top (0001) plane, the mixing of which leads to the perception of white light emission.

Carbon Nanotube (CNT) film was grew on a Ni/Au buffered Si wafer by the pyrolysis of iron phthalocyanine(FePc). The intense pulsed emission stability of as-prepared CNT film was studied by measuring the intense pulsed emission characteristics repeatedly with diode configuration under the same voltage of Marx generator. The results show that, at the peak values of pulsed voltage ranging from 1.60 MV to 1.74 MV (the corresponding electric field intensity range of 11.43～12.43 V/μm), the peak current of the first emission cycle reaches to 331.2 A. The Ni/Au composite buffer layer can not only improve the emission current, but also the emission stability of as-prepared CNT film. When the number of emission cycles is up to 7 times for cold cathodes, the current of the Ni/Au-CNT cathode is 72% for the first current point, and while the Ni-CNT cathode and Si-CNT cathode is 62% and 32%, respectively.

A novel infrared reflective device was proposed. The alignment direction of cholesteric liquid crystal particles in the device are driven by the diversion of host liquid crystal, which can realize the work state switch from infrared reflection to transmission. So the cholesteric liquid crystal particle is the key of this device. The preparation method of cholesteric liquid crystal polymer film was introduced, and the liquid crystal polymer particles were obtained by ultrasonic processor. The influence of different proportions crosslinker liquid crystal 1 on liquid crystal film brittleness and reflection band was studied. It was found that when the proportion of crosslinker liquid crystal 1 is 100%, the brittleness of liquid crystal film is maximum and the clear point is the highest. In addition, the reflection band of liquid crystal films is getting closer to shortwave with crosslinker liquid crystal 1 concentration increasing. At last, the influence of the liquid crystal film thickness on liquid crystal polymer particles was conducted. It was found that the size of liquid crystal polymer particles became smaller and evener with the film thinning. This work will support the optimization of electricity-responsive infrared reflective windows performance.

CdxZn1-xS thin films were prepared by chemical bath deposition in a solution containing cadmium chloride, zinc chloride, thiourea, sodium citrate and ammonia. The morphological, structural and optical properties of CdxZn1-xS films were investigated by X-ray diffraction, scanning electron microscopy, energy disperse spectroscopy and UV-vis-NIR spectroscopy. The photo-current response curves were also tested to study their electrical property. It was found that CdxZn1-xS thin films were successfully prepared at pH values from 10 to 13. With increasing of pH value, the ratio of zinc atom in the film and optical band gap decreased. All the films showed obvious photoconductive phenomenon. The CdxZn1-xS films prepared at pH=11 and 12 showed uniform and dense surface morphology and presented good crystallinity. Their optical band gaps were about 2.92eV and 2.72 eV respectively, with light and dark conductivity ratio of 1.20 both. The photo-currents in the samples fell more quickly when the light was turned off, with current reduction of about 68.55% and 69.39% after 10s.

Using the pure N,N-dimethyl formamide, pure dimethyl sulfoxide and mixed solvents with different proportion as precursor solvent, respectively, the perovskite thin film samples were prepared. The samples were divided as two sets, one set was processed by thermal annealing in N2 atmosphere, while the other set was processed by solvent annealing in dimethyl sulfoxide vapor atmosphere. The microstructure and optoelectronic properties of the samples were systematically analyzed. The results show that, compared with the thermal annealing, the grain size and uniformity of samples can be significantly improved by solvent annealing, which reduces volume fraction of grain boundaries or interfaces in the films. By solvent annealing and mixed solvents, the absorption and utilization ratio of visible light are enhanced, the morphology is improved, and the crystallization quality is optimized. The increasing of photoluminescence intensity indicates that the defect densities of thin film are reduced. Thin film solar cells were fabricated using the optimized perovskite film as the absorber layer, and a conversion efficiency of 15.7% was achieved.