As one of important technical specifications of diffraction gratings, stray light directly affects Signal Noise Ratios(SNRs) of spectrometers; especially in spectrum analysis at UV spectral bands.In order to search the stray light from applications of ruled gratings in the spectrometers, this paper analyzes forming causes of the stray light in theory based on a scalar diffraction equation.The numerical simulation indicates that the random errors in the spacing of grating grooves and the random errors in the groove depths for UV ruled gratings are the main sources of the stray light, moreover, the small-scale random surface roughness affects the stray light slightly.Otherwise, it presents a concept of a relative width for the exit slit and establishes the relation between the height and relative width of exit slit and the intensity of the stray light,which provides a theoretical reference for reducing the stray light in the process of fabricating gratings essentially and repressing the stray light in practical applications.Finally, in order to compare theory results with experimental results measured by a filter method, an equation to solve the total stray light is presented to measure the stray light of four wavelengths by a UV spectrophotometer.It proves that the maximal relative error between theory results and experimental results is about 13%, when random errors in the groove periods, random errors in the groove depths and random surface roughness are 0.8 nm, 0.5 nm and 1.2 nm respectively.

In order to discuss the influence of the crosstalk effect on a Grating Light Modulator(GLM) driven by a passive matrix, an electrical model of the GLM array addressed by the passive matrix is established. The operation principle and the driving voltage of the GLM are presented, and the electrical model is analyzed and simplified according to the Kirchhoff's Current Law(KCL) and Kirchhoff's Voltage Law(KVL).The voltage relationships among the half-selected pixel, non-selected pixel and the all-selected pixel are obtained.Then, the operating voltage and the pull-in voltage of the GLM are measured by an experiment and the crosstalk effect is validated by another experiment, so that the relationship between the ±1 order diffractive intensity of those three kinds of pixels and the driving voltages of all-selected pixels is gained.Finally, two methods for restraining the crosstalk effect are proposed.The experiment shows that the operating voltage and the pull-in voltage of the GLM are 8.0 V and 8.5 V respectively.It also indicates that the voltage of half-selected pixel is approximately half of the all-selected pixel, and is much larger than that of the non-selected pixel, which is consistent with the theory of crosstalk effect.These results illustrate that the crosstalk effect can reduce the contrast ratio and light utilization of the GLM, but selecting a lower driving voltage can realize the monolithic integration of the GLM and the active matrix and can eliminate the crosstalk effect.

A new method is presented based on a Stokes vector and a Mueller matrix to measure the retardation of a Liquid Crystal Variable Retarder (LCVR) attached with a compensator, a 2 kHz square wave with the adjustable amplitudes 0~10 V is used to control the LCVR and the retardation is measured at six wavelengths in a visible region.The experimental results show that the LCVR has the advantages of low voltage controlling, simple operation, wide adjustment ranges and no mechanical rotating,and the error of measurement is less than 0.6%.The fitting function is established based on the characteristics of the retarders,and the retardations of all wavelengths and all voltages in the working condition of the retarders are obtained by the least-square curve fitting.The standard error of curve fitting is less than 0.02λ.The measurement method, experimental results and the fitting function provide a reference for the polarization device research.

The motion control system of moving components in the Shanghai Synchrotron Radiation Facility (SSRF) Small-angle X-ray Scattering (SAXS) station is an important part of the facility construction,which is applied to accommodate light spots and find the best position of the scattering sample to satisfy the different experiment demands of the users.The system is a distributed control system recognized as the “standard model” in large accelerator laboratories in the world and is developed based on the large-scale software EPICS (Experimental Physics and Industrial Control System).It implements controls for 23 moving components of two four-jaw slits,two slit-bases,a sample-stage and two Beam-Stops.The moving test results indicate that the resolutions of moving components are from 0.333 33 μm to 20 μm and the error is about 1%.The test results can satisfy the requirements of the SAXS station and show advantages in the stability,repeatability and higher precision of the system.Now,the system has been successfully applied to debugging and running in the station.

A method to detect high speed and parallel CCD control signals and to simulate sequential video signal outputs is presented to realize reasonable design and applications of a space optical remote sensing system with high-performance and multi-spectral TDI CCD image sensors.At the line readout and transfer period of 200 μs,48 clock drivers and 31 DC bias signals are analyzed and detected simultaneously.At the same time,the prepared image data are fastly transmitted from a disk array to a video image generator consisting of a FPAG,a D/A,an amplifier and a denoise circuit through the PCI-X and LVDS bus,so that the sequential and real-time outputs of diversified testing video images are realized.It is shown in experiments that three simulative chromatic video signals at a frequency of 6 MHz and eight simulative panchromatic video signals at a frequency of 12 MHz are generated synchronously with the reference voltage of a video signal of 8 V and amplitude ranges of 0~1.7 V and the valid sampling time and the signal stabilization can meet the requirements of the simulation and detection for the imaging circuits.

To diagnose the electron temperature of a Z-pinch plasma,the ratio between the dielectronic recombination satellite line for argon to the w-resonance line for lithium-like as a function of the plasma temperature is analyzed.The spectral emission pattern of helium-like argon together with the associated satellite line from lithium-like argon have been measured with an X-ray elliptical curved crystal spectrometer on a Yang accelerator.The spectrum of argon-puff Z-pinch plasma is described by a curved mica crystal analyzer and the helium-like resonance line w,satellites j,k,q,magnetic quadrupole M2 line,intercombination line y,forbidden line z and the cold Kα line of argon are recorded with an X-ray film.Then,the line ratios between the dielectronic recombination satellites (j+z,k) to the w-resonance line are calculated,obtained results show that the electron-temperatures of the plasma are 960~1 060 keV.The theoretical models are in agreement with the experimental results well,which supports the conclusion that proposed method can be used in a temperature diagnostic method for plasma.

In order to detect the Acoustic Emission(AE) signals in solid materials, a novel ring cavity all fiber F-P interferometer is proposed.By welding the incidence and the exit of a 2×2 optical fiber coupler directly,an optical fiber ring cavity is constructed,which can be attached or embedded in the measured solid to detect the AE signals.The detection property of the optical fiber sensor is determined through the theoretical derivation and the computer simulation.Otherwise,a piezoelectric ceramic(PZT) is powered by signal generator to generate a continuous known AE signal,and a blast wave is adopted to simulate the burst AE signal in marble slabs.Then,they are detected by the optical fiber sensor,and the characteristic frequencies of AE signals are obtained by the Fourier technique.Experimental results indicate that the system can detect AE signals that cause the optical fiber stretching to 10-8 m in an axial direction and can identify the characteristic frequency of AEs.For it can work without the match of optical path and is suitable for monitoring the large scale component,the sensor provides a new method for the health examination and the monitoring of material structures.

When a Distributed Feedback(DFB) fiber laser is used in underwater acoustic detection,it is found that the acoustic pressure sensitivity of the DFB fiber laser is higher than that calculated under hydrostatic boundary conditions,and its frequency response curve is not smooth.To explore the reason and the mechanism of the above,the dynamical characteristics of the DFB fiber laser are investigated.Firstly,the vibration models of the DFB fiber laser whose both ends are fixed and placed in air and water respectively is analyzed numerically.And then,with specially designed testing assemblies,experimental study on it is also conducted.The results indicate that the first order natural frequency of the DFB fiber laser is related to the length of the optical fiber between the two fixed ends,and the frequency decreases with the length increases.When the length is set as 50,55 and 60 mm respectively,the first order natural frequencies have been found to be about 250,200 and 125 Hz in air,and 200,160 and 120 Hz in water.These data show that when vibration takes place in the DFB fiber laser, its output will be changed to cause the fluctuation of acoustic pressure sensitivity,which shows the first natural frequency is an important factor for the underwater acoustic detection.

In order to improve the detecting precision and robustness in determination of the content of free amino acid in tea by the FT-NIR spectroscopy,a predictive model was established by selecting efficient spectral regions combined with the Partial Least-squares(PLS) algorithm.In this research,the synergy interval Partial Least-squares (si-PLS) and the Genetic Algorithm Partial Least-squares (GA-PLS) were applied to select the efficient spectral regions.The number of PLS components and the number of intervals were optimized according to the cross-validation in a calibration set.The performance of the final model was evaluated according to the Root Mean Square Error of Prediction (RMSEP) and the Correlation coefficient (R) in prediction and calibration sets.Experimental results show that the two methods are able to produce better prediction models in relation to the full-spectrum model,and the performance of siPLS model is better than that of the GA-PLS.The optimal model offers its R in 0.954 2 and RMSEP in 0.256 0 respectively by a prediction set.This study demonstrates that NIR spectroscopy combined with efficient selection wavelength region algorithm can be used successfully to analyze the content of free amino acid in tea.

A phase-locking method for two individual fiber lasers is proposed to realize high power and high brightness fiber laser outputs.By taking a beam splitter with reflectivity of 50% at 45° as a interferomtric element to construct the Michelson cavity,two optical fiber lasers can realize coherent enhancement at output ends,so that the coherent combination output of two laser beams is complemented.The phase-locking principle is analyzed in detail and the phase-locking of two individual double clad fiber lasers is experimentally demonstrated.Experiments obtain a coherent combination laser output of 360 mW,and a power combination efficiency about 73%.Moreover,the on-axis power intensity is greatly enhanced.The phase-locking approach is simple and no power-limitation optical components,so the drawbacks of fiber couplers are eliminated and the output power can be further up scaled.These results show the method is one of the most promising phase-locking technologies in the high-power fiber laser coherent combination field.

In order to investigate the optical properties and the stability of a diffuser panel used for on-board calibration for satellite optical sensors under space environments,some reflective properties including the Directional/Hemisphere Reflectance (DHR),Bidirectional Reflectance Factor (BRF)and the Bidirectional Reflectance Distribution Function(BRDF) were measured,and a series of environment simulation tests were performed.Experimental results show that the reflectances achieve 0.99 at VNIR,0.95 at SWIR;and the differences of the reflectances at VNIR are within 0.01;The variations of BRF(10-60°) and BRDF (0-60°) are less than 0.1 and 0.045,respectively,and the Lambertian property of the diffuser panel is independence on the spectral band and UV radiation.Moreover,the degradation of reflectance at 350 nm is within 3% after 120 h equivalent solar UV radiation and the atomic oxygen exposure or proton bombardment has little effect on the optical properties of the diffuser panel.As characterized by above,the diffuser panel has gotten through on-orbit level mechanical environment tests.For its high reflectance,spectral flatness,Lambertian characteristics and stability in orbit, the diffuser panel can satisfy the requirements for space applications.

A novel deterministic sub-aperture magnetically assisted polishing method,Magnetorheological Finishing (MRF),was explored and some fundamental principles and key problems in MRF process were researched.The KDMRF-1000F MRF polishing machine developed by our group was introduced and two methods to establish the material removing model in polishing process were given.Then,how to realize a path planning based on dewell time was described.By using the KDMRF-1000F MRF polishing machine and KDMRW-1 water based MR fluid,one K4 flat work piece (80 mm in diameter) and one K9 sphere work piece (145 mm in diameter) were polished,experiments show that work piece one has a peak-to-valley (PV) value of 55.3 nm and a root-mean-square (RMS) value of 5.5 nm;and work piece two has a PV value of 40.5 nm and a RMS of 5 nm,which indicates that the surface roughnesses of two work pieces have been improved obviously.It is concluded that the MRF has the abilities to produce optical surfaces with high form accuracy,surface roughnesses and high efficiencies.

With the aim to fabricate nano-silicon-tips with a high aspect ratio,the effect of changing the mask direction on tip shapes was studied.The fabrication processes for nano-silicon-tips using a (100) single crystal silicon wafer by the anisotropic wet etching in the KOH etchant were designed.Based on the experiment and the model of {411} crystal planes,the crystal planes of silicon tips were analyzed.Through discussing the effect of the mask direction on tip shapes,the process parameters to get high aspect ratio nano-silicon-tips were achieved.Experimental results indicate that silicon tips of octahedron and tetrahedron are fabricated with changing mask directions.The direction of mask does not affect the rapid etched crystal planes at the fixed solution concentration and temperature.Nano-silicon-tips are formed by the anisotropic etching in 40% KOH etchant at 78 ℃ and the dry oxidation sharpening in 3 h at 980 ℃,as the square masks are aligning to the 〈110〉 direction.Experiments show that the high aspect ratio of the tip is greater than 2,and the profiles of silicon tips are constituted by {411} crystal planes that intersect (100) crystal planes at 76.37°.

In order to improve the accuracy of a recovered object by machine vision system,the relationship between the camera position and the accuracy of the recovered object is analyzed and the weighted values based on the relationship are evaluated. Then, the weighted least square method is used in the machine vision system.After expressing the machine vision model by parameter equations, a recovering equation for 3D space point recovering is deduced based on the parameter equation.Then,according to the error transmission principle, the transmission from the uncertainty of a camera image plane to the uncertainty of an observe space is analyzed,and the relationship between the error transmission and the camera position is obtained.Finally,the covariance of the observed object is used to evaluate the weighted value for the weighted least square method.Experimental results indicate that the accuracy of the weighted least square method is better than that of the general least square method. When the noise variance is bigger than 0.5 and the number of the photos is less than 30, this method can offer a very good measuring accuracy.According to the experiment,when the measurement distances are 37.031 0 cm,24.970 4 cm and 26.015 3 cm,the weighted least square method can improve the accuracy by 0.4 cm.

A novel four-arm structure single crystal silicon MEMS gripper integrated with force sensors is successfully developed by using a silicon bulk machining method to realize the monolithically fabrication and the control of gripping force.Based on the technology of the piezoresistive detection,an effective technique for fabricating vertical sidewall piezoresistors in planes is used to fabricate the piezoresistor in the end of the end-effectors.The Finite Element Analysis (FEA) is used to analyze the structure of the gripper and the flexible beams of the force sensors.The comb microactuator generates a linear horizontal motion,which is converted into a rotational motion by an S-type flexible beam system to realize the gripping motion.Combined with the four-arm structure,the manipulating range is expanded.The bonding of the silicon and the glass is used to realize the electrical isolation of the gripping arms.At a voltage of 80 V,the deflection of the arm tip is 25 μm and the ranges of the operation are 30~130 μm.Testing results show that the maximum measurement range of the piezoresistive sensors is better than 1 mN and the resolution is better than 3 μN,which can be used effectively in the force feedback.

To realize the high resolution measurement of rotation motion to get the characteristic and dynamic parameters of MEMS resonators in every moment,this paper presents a method to measure the rotation angle based on the combination of Radon transform and phase correlation for the motion image sequences of microstructures collected by a stroboscopic imaging technology.By the Radon transform,the motion in Descartes coordinates is transferred to parameters of polar coordinates,so that the rotation motion is projected to displacements of parameters.Then,the displacements are measured by a phase correlation method to realize a subpixel motion measurement.Experimental results show that the resolution power of rotation angle is superior to 0.01°.MEMS motion process is analyzed by the method,and the offered solutions indicate that the influence of the distortion caused by rotation is deceased,while the repetition of measurement is increased.

With the aim to fit the sin wave curve of a strap-down north seeking system and to calculate the angle between the start position and the real north accurately,the principles of the system are introduced,and the factors influencing the measuring precision are analyzed.Then,the least square method and the Back-Propagation(BP) neural network are adopted to fit curves separately.Experimental results indicate that the BP method is preciser than the least square method,for its error sum is 0.23 μV and the squared error of residual error is 1.3 mV.Moreover,for phase angle measurement,both methods can get the same average values in six measurements,but the squared error of BP neural network is 8″,which is more accurate than that of the least square method.These results show that the proposed BP method can satisfy the precision requirements of curve-fitting north seeking systems.

A novel 6-(P-2P-S) parallel macro manipulator used as the macro system of the cutting equipment for chromosome is put forward,which is characterized by some movement decouplings on the orthogonal posture and its good manufacturability.In order to explore the theoretical foundation of the manipulator,the kinematics performance and the attitude distribution properties in the orientation workspace are investigated. Based on the design feature,the position problem is analyzed to obtain the inversion of forward displacement.Then,the kinematics equation is derived,and the speed transmission evaluation indicators (Kv,Kω) are defined based on the Jacobian matrix.Finally,in consideration of the structure constraints,the distribution of the Kv and Kω in the orientation workspace is analyzed.Simulation results indicate that the Kvmax and Kωmax are close to the edges of orientation workspace,and they can be changed in the ranges of 0.970 9-1.179 1,and 8.586 5~10.978 3,respectively.By choosing appropriate Kv and Kω in the orientation work space,it can satisfy the different requirements of the parallel macro manipulator applied in other precision operating fields.

A digital dispensing technology by using a pulse inertial force drived micro dispensing system without a heat resource and a pneumatic actuator for microfluids was presented to slove the existing problems on the huge external driving device,complicated configuration manufactured by MEMS technology and poor operability and compatibility with biological samples.The principle of digital dispensing technology of microfluids was introduced.Then,effects of the sample uniformity and driven parameters on dispensing technology were researched and the experimental data were analysed.Finally, the medium-density microarray was accomplished with the low viscosity liquid at 4.13 mPa·s.The experimental results show that the average diameter of spots is below 100 μm,and the Coefficients of Variation (CV) for diameters of spots are clearly below the measurement error (8%).It can satisfy the microarray requirements for the density,uniformity and sizes of spots,and can be used in small laboratories.

Adjusting the drift angles of a space camera on real time in photographing could decrease the influence of the attitude change on image quality and could prolong continuous photographic time with more TDI integral progressions.In this paper,the requirements adjusting drift angles on real-time were analyzed in detail and the real-time adjusting tactics were presented.Then,the structure of a drift angle adjusting system and the operating principle of real-time adjusting were introduced.Finally,a batch-type real-time adjusting method was presented and experimented.Experimental results indicate that the partial difference of drift angle can be adjusted to below 4.2′ with the adjusting error no larger than 72.17″ and the adjusting time less than 1 s.Moreover,Modulation Transfer Function(MTF) of the image is 99.67%,while the drift angles are adjusted.These results show that the proposed method can meet the requirements for adjusting drift angles on real-time and long period without essential influence on images in photographing process.

To realize mode matching and high quality factors (Q) and to improve the performance of micro-machined ring vibration gyroscopes,a novel electromagnetic ring vibration gyroscope was designed and fabricated by using a centrosymmetric and enantiomorphous structure consisting of a ring and eight symmetric support springs to match the resonant frequencies of a drive mode and a sense mode.The theoretical models for the sensitivity and the mechanical noises of the gyroscope were deduced,and the effects of gyroscope parameters on the sensitivity and resolution were analyzed.Through the standard bulk MEMS processes,the gyroscope samples were fabricated with (100) silicon.Experiments show that all the processes are easy to realize,for they have no bonding operation.Moreover,a frequency response experiment indicates that the frequency split between the drive mode and the sense mode is less than 0.5 Hz,and the quality factor is about 500 in atmosphere and 14 000 in vacuum with 1 Pa.Phase Locked Amplifier(PLA) tests show that the resolution of the gyroscope is 0.05 °/s and the sensitivity is 0.2 μV/(°/s) over the ranges of -200 ~200 °/s.

In order to realize the steady and reliable operations of microobjects,a hybrid-type electrostatic silicon microgripper integrated with gas pipes for aspirating and blowing operations is developed and its structure and dynamic characteristics are analyzed by the Finite Elemenl Analysis(FEA).Gas pipes integrated in this novel microgripper are used to improve its picking and placing capability and a bulk micromachining technology is employed to fabricate the microgripper from single crystal silicon wafers.The linear horizontal motion generated by an electrostatic microactuator is converted into the rotational motion of the arm by the S-type flexible beam system to realize the gripping motion.Moreover,the grippers with two different dimensions are designed to expand the manipulating range and the gas pressure control system and electrostatic comb drive control system are designed for controlling the microgripper.Experiments show that the deflection is 25 μm at the arm tip of the gripper in a driving voltage of 80 V,and the microobjects can be picked and placed successfully from 100 μm to 200 μm in the experiment,which indicates that this microgripper realizes the steady manipulation and can satisfy the micromanipulation of the microobjects from 100 μm to 200 μm.

In order to match the image motion produced by the movement of ground objects at a focal plane accurately,a calculation method for image motion speed is realized by taking the spacecraft location vector and velocity vector based on the WGS-84 coordinate system and the spacecraft attitude angle and attitude angular velocity based on spacecraft orbital coordinate system as input parameters.Firstly,a inertial coordinate system is selected flexibly based on the original calculation model of image motion speed to decrease the complexity of solving orbital inclined angels and descending node longitudes.Then,the cosine value of central angle between the spacecraft and the descending node is solved by using cosine theorem of spherical geometry to avoid the cumbersome solving process by judging the flight direction of a satellite and the latitude of sub-satellite point in an original calculation model of image motion speeds.Subsequently,the longitude and the latitude of sub-satellite point and the FLASH address are linked organically and skillfully to avoid the cumbersome lookup process when elevation data are visited.Then the feasibility of calculation model of image motion speed is evaluated by analyzing the impact of the residual error of image motion speed on the MTF of a camera.Finally,the whole process of calculating image motion speed is realized on DSP from TI.Analyzing processes and experimental results indicate that the residual error of calculation model is 0.15%,which results in a 1% decrease of the MTF.The calculation time of image motion speed is less than 2 ms.These data satisfy the system requirements of the camera on the calculation accuracy and computing time.

According to the requirements of calibrating multi-cameras in motion parameter optoelectronic measurement systems used in big space motion analysis,a novel approach of calibrating the internal and external parameters of multi-cameras is proposed. Firstly,a principal point is assumed at a certain location around the image plane center,and the fundamental matrix is found out by Hartley's eight-point-algorithm.Then,different focal lengths of different cameras are obtained linearly through a epipolar geometry and it's constraint.After educing the essential matrix,the singular-value-decomposition is carried out to obtain the rotation matrix and translation vector with a scale factor.Moreover,the coordinates and the distances of the ball bar are computed by a triangulation method,and the scale factor is found out through the true distance of the ball bar.Finally,the estimation of internal and external parameters of the multi-cameras is transformed to a nonlinear minimization problem by a cost function,and the optimal principal points of two cameras are found out iteratively by using the improved simulated annealing evolution strategy.The experimental results show that the precision of focal length and principal point is up to 0.1 pixel and the stardard deviation bar length is 0.046 mm.Compared with traditional methods,the novel calibration approach expands the applications of the multi cameras and can find out the internal and external parameters simultaneously without the constraint for the ball bar movement.These results prove that the improved simulated annealing evolution strategy accelerates the convergence speed and improves the global convergence performance.

The mature ion source in a commercial mass spectrometer is sealed and works under vacuum or low atmospheric pressure environment,which limits the application of the mass spectrometer.In order to solve this problem,an ambient negative corona discharge ion source with line-cylinder electrodes is put forward in this paper.The gas discharge part in the ion source is composed of inner and outer electrodes with the diameters of 0.16 mm and 4 mm respectively.There are four slots in the outer electrode for injection of samples and traction of ions.The experiment results show that when samples are carried by N2 and the voltage on the inner electrode reaches -3 800 V,the corona discharge occurs between the inner and outer electrodes and negative substances,such as acetic acid,can be ionized under ambient conditions.The mass spectrogram suggests that N2 and O2 in the air maybe go into a reaction to generate NO－2 (m/z:46), NO－3(m/z:62) and [M+NO-H]－ (m/z:89).The corona onset voltage of the line-cylinder electrodes is analyzed with COMSOL and the theoretical result consists well with the experiment result.The simplicity of the interface of the ion source makes it available in the ambient mass spectrometer,Ion Mobility Spectrometer(IMS) and high-field Asymmetric Waveform Ion Mobility Spectrometer(FAIMS).

The essence of in-band transmission loss of plane wave in the Frequency Selective Surface(FSS) with an infinite planar single layer was illustrated by distinguishing the mismatch loss and the dielectric loss.The effects of some parameters on loss characteristics were studied based on a modal analysis approach.Research results show that the dielectric loss is higher when FSS is illuminated with TE polarization in a large incidence angle.While array parameters and aperture widths affect the bandwidth,the dielectric loss always shows the reverse tendency with the bandwidth.The in-band transmission loss of one side dielectric loading is induced mainly by a mismatch loss or by a dielectric loss that depend on the dielectric thickness and the way of dielectric loading.When the dielectric thickness is 1/4 of wavelength,the mismatch loss reaches its maximum;when the dielectric thickness is 1/2 of wavelength,the dielectric loss reaches its maximum.Moreover,with the same dielectric thickness,the mismatch loss of the FSS with dielectric superstratum is less than that of the FSS with a substrate.The loss of both-side symmetric loading is from dielectric loss,which is lower than that of one-side loading.

Electrostatic stretching is a new technique developed in recent years for fabricating membrane mirrors with light,good flexible and collapsible properties.Based on the electrostatic shaping principle of a membrane mirror,the force field of the stretched membrane was calculated by a single electrode mode,and the support structure of the membrane and its control circuit were designed.Then,several kind of technological parameters for controlling membrane shape were given and the focus of reflective surface of the membrane was measured by a knife-edge.The experimental results show that the PV and RMS values of the membrane mirror are 11.14λ and 1.86λ without applying an electric force respectively,and the focus of reflective surface of the membrane is about 2.1 m with a 10 000 V high voltage.Finally,the Zernike fitting was used to verify the reliability of the data, and it suggests that the surface accuracy can be improved further by optimizing the holding structure and selecting excellent membrane materials.

To measure the flow field distribution parameters on the aerodynamic surfaces of airplanes and flow pipes, a hot film sensor array with the sensing material of nickel is fabricated on a polyimide substrate by the MEMS technology.The flexible and thin (whole thickness less than 100 μm) sensor array can be attached upon a curved surface (such as airfoil) to measure the temperature distribution or flow parameters in a minimal invasion rapidly.The structure of sensor array is designed and the fabrication techniques are presented based on the magnetron sputtering and heat annealing.Then,the influences of the argon pressure,substrate temperature,sputtering power,and the annealing temperature on the temperature coefficient of resistance(TCR) for the sensor is studied.Finally,a high TCR of 4.64×10-3/℃ with good linearity is obtained.

Flight parameter computing platforms are important supporting technologies in the aircraft development process.In order to accurately simulate the wind model and calculate the flight parameters,a simulation modeling method for a flight parameter simulation platform under wind field environments is established.Based on the several influence factors of the flight,some mathematical models of the flight parameter simulation platform is established.Then, the Simulink toolbox in the Matlab is used to establish the model to implement the complex kinetic equations based on the principles of hierarchy and C MEX S-function.After the models are established, these models are encapsulated for several subsystems and the flight parameter computing platform based on a unified interface is assembled.The simulation tests,including the performance test of a certain gliding aircraft and the wind influence test of a certain uncontrolled aircraft,prove that the computing capability of the platform can obtain a calculating step of 1 ms.The comparion on the simulation and the field test data of a certain uncontrolled aircraft shows that the relative deviation of range is 0.44%,the relative deviation of maximal altitude is 1.8%.Taking QJ 1997-90 as a reference standard,it is demonstrated that the platform scheme is resonable and reliable,which meets the engineering accuracy requirement.

The high vacuum acquisition and vacuum maintenance of device-level vacuum packaging for MEMS gyroscopes are studied to increase their quality factors and stability. The relationship between error signals and quality factors is analyzed based on the dynamics equations of Z-axis MEMS gyroscopes and the air damping of gyroscopes with high quality factors is analyzed using rarefied gas dynamics theory.The changes of quality factor of an early vacuum packaged gyroscope are analyzed,and the result shows that residual gasses in vacuum cavity result in the decreasing of vacuum level.A temperature programmed desorption-mass spectrometry (TPD-MS) is used to measure the contents of the evolution gases for the ceramic packages and the lids.According to the contents of trapped gases,the reasonable getter is chosen to absorb the gases and to keep the vacuum level of the cavity.Finally,the device-level vacuum packaging process is improved.The test results show that the quality factor of the gyroscope packaged with a new device-level vacuum packaging process is about 162 660,which is about 14 times those of previous vacuum packaged MEMS gyroscopes,and the change of quality factor is less than 0.05% within one year.

Angular disturbance isolation evaluated by a isolation index is an important function of a seeker.In order to realize a quick test of the isolation index, the principle of disturbance isolation of the seeker is analyzed,and a model of the isolation index is deduced.Then,by analyzing the structure of stabilization loop of the seeker,an isolation index test method based on the dSPACE hardware-in-the-loop simulation system is proposed to be used in the stage of controller adjusting.According to adding a disturbance signal in the stabilization loop,this method can test the isolation index quickly without simulating missile advantages in disturbance by the flight simulation turntable.Compared with the traditional methods,this method has simple structures and fast test velocities.Using this method,the isolation index of stabilization loop of an infrared imaging seeker is tested and the results show that the test error is less than 2 dB compared with that of the traditional method,which indicates that this method is feasible to test isolation indexes of seekers or other optoelectronic platforms on moving vehicles.

A color-difference detection algorithm based on the CIELAB color space is proposed to solve the problem of detecting the color-difference of an image with random texture in the actual production environment. Firstly,the number of clusters are calculated in the HSV color space and the initial cluster center is determined in the CIELAB color space.Then,the color-difference is detected with a fuzzy C-Means clustering.Taking the pebble tiles for examples,some experiments in inteference of light,dust and vibration are done in this paper.The experimental results show that the proposed color-difference detection algorithm has anti-interference capability and obtained result is consistent with that of the human eye.When it is affected by the dust and vibration interference,the color-difference value is less than 1.5.While it is affected by the light interference, the color-difference value is greater than 1.5.It suggests that the stability of light power should be controlled reasonably in the practical application.These results also prove that the algorithm can be used to detect the surface color of other random texture.

The Orthogonal Subspace Projection (OSP) algorithm is a supervised classifier that needs the information of the classified objects.To expand its application, a local OSP (LOSP) is design to apply to detect the hyperspectral image.The anomaly detection algorithms are usually used to extract the isolated man-made objects in the nature background,where the substances in the small local region are usually uniform.Based on the principle,the LOSP is constructed by choosing the detected pixel as the interested object and the mean of its nearby pixels as the suppressed object.The experiments show that the LOSP can detect the sub-pixel targets with a content greater than 30%,and can also detect the targets occupying more pixels by enlarging the window size.In addition,LOSP is proved not to be affected by the Hughes phenomenon,and the computing time is less than 1/10 that by RX detector when the number of wavelengths is 80.LOSP is effective both in precision and in efficiency,and is applicable to the real-time detection of the hyperspectral image.

In order to research the specific applications of a CMOS image sensor with a rolling shutter,this paper researches the specific CMOS image sensor.Firstly, the principle and characteristics of the global shutter and rolling shutter of the CMOS imager are introduced,and the impact of the rolling shutter on measurement precision of the imaging system is analyzed.An imaging experiment is taken to analyze the rolling shutter.Then,a new method for computing instantaneously 3D poses and velocities of fast moving objects is presented by using a single view.Finally,a general perspective projection model of a moving 3D point is presented,and a method to estimate the pose and velocity is discussed.Experimental results indicate that distortion appears when moving objects are imaged by the CMOS imager,and the degree of distortion has close relations with some parameters of CMOS imager such as integration.Moreover,the post and velocity parameters of a motion object can be obtained which shows in the case of minimizing errors. The calculation error is under 2.5% and the measurement accuracy is 0.01 rad/s.After analysis,the experimental results have proved that not only the approach is practicable but also it can transform a low cost and low power CMOS camera into a new velocity sensor.

The basic principle and algorithm of the complex valued Independent Component Analysis(ICA) are introduced,and a new target recognition algorithm based on the complex valued ICA is proposed to be applied to the multi-sensor fusion target recognition.Two images from different cameras are composed into a complex value training group matrix,then the Fast ICA(FICA) algorithm is performed on the matrix to get independent components (ICs).After extracting the features of training set and testing set based on ICs,the linear discriminant analysis is adopted to train target features to find out a reasonable classification threshold.Finally,the targets in testing set are classfied and recognized.Experiments show that the recognition rate obtained by the proposed algorithm is 92.1%,which is more excellent than those of the traditional ICA in 78.1% and PCA in 76.2%.

For the difficulty of recognizing infrared small targets in sky backgrounds and its low recognition reliability,an improved method based on the unfair D-S evidential theory for small target recognition is presented combined with the previous methods.To improve the recognition confidence degree,the velocity and gray information of small targets is regarded as the major evidence,and the track information of the targets is treated as the supplement evidence.The proposed algorithm firstly recognizes the small target by using the velocity and gray information.If the targets cannot be recognized only by using the velocity and gray information,the target track information should be added into the fusion evidences to complete the fusion by the unfair D-S evidence method.If the added target track information can strongly support the decision of target velocity or target gray,the proposed algorithm can evidently improve the recognition reliability of infrared small targets in sky backgrounds.The simulation results show that recognition confidence degrees are raised by 125% without adding new sensors.

Changing a variable-length code into an equal-length code is required in the feature match process of a conventional algorithm based on the zero-crossing of the wavelet transform proposed by Boles.The factor for the variability of the code length is complex,so it is difficult to change the variable-length code into the equal-length code compulsively and to avoid the effect of feature matching on the final recognition rate.To solve the above problem,an iris recognition algorithm based on the zero-crossing position of wavelet transform and matching using a variable-length code is proposed.Firstly,The zero-crossing position of the wavelet transform at each level from the 4th to the 8th decompose level is calculated,and then shift-matching algorithm is used to directly calculate the distance of the variable-length code.That is to say that the distance between the two samples is calculated at each relative shift.Finally the minimal distance is chosen as the distance between the two sample.Experiments are done with iris database containing 800 images,and the experimental results show that the recognition rate obtained by this method reaches 97.70％,which is about 5.16％ higher than that of Boles' method.For the feature extracting time and the match time is only 0.2 s,the proposed method can satisfy the system requirements of real time applications.

To realize a common and feasible design method for a high speed communication card,this paper introduces the characteristics of the CompactPCI bus and describes the relations of bus latency and communication bandwidth simply.A kind of hardware architecture based on dual-port RAM is proposed to isolate different agreements and different frequency buses of the PCI bus and the local bus.By using this method,the hardware design and drive programming are simplified greatly,and the interface chip and the drive chip are integrated to be modules for the development of function cards.At the same time,a experimental card is produced to test the feasibility of the method,and obtained optimal communication bandwidth by the experimental card is 117.97 MB/s at different communication modes.Experimental results show that the communication bandwidth of the experimental communication card meets the requirements of high speed and the method is easy to realize.

In consideration of the characteristics that loads of nodes and hops to Sink are interrelated in data gathering traffic model,a grade-based unequal clustering algorithm is proposed to solve the load-balancing problem of wireless sensor networks.A traffic-mode-based clusterinig topologic structure is built to realize the load balancing of network and to relieve the traffic congestion around Sink(“Hot Spot” problem). Based on the analysis of the data gathering traffic model,the relationship between the hops to Sink and the load of Cluster Heads(CHs) is given.Then,a grade-based load-balancing clustering algorithm for a wireless sensor network is prensented,in which the cluster size is based on nodes’ grade and the cluster head selection uses mulit-weight.Simulation results show that the standard deviations of CHs’ load with different grades are reduced by 9 times,so that the load balance of the wireless sensor network is improved effectively.

In order to improve the accuracy of the diagnosis with medical signs in medical imaging diagnostics,a computer-aided diagnosis system is developed quickly and accurately to find out the differentce of benign and malignant Solitary Pulmonary Nodwles(SPNs) in chest CT images based the image features of SPNs.Firstly,SPNs are extracted from chest CT images using the interactive segmentation,and the multi-resolution histograms of SPNs are directly calculated to receive a high-dimensional feature sample set with spatial information of SPNs.Then,the classifier for differentiating benign and malignant SPN is constructed by using a Support Vector Mechine(SVM).Finally,the performance of classification is evaluated by testing the trained SVM with the test sample set.The test results of 214 cases show that it takes 4.83 s for computing 768 dimensional features of 240 SPNs and 2.24 s for training and testing the SVM classifier.The receiver operating characteristic (ROC) analysis of classification performance of the proposed approach shows that the sensitivity is 73.33%,specificity is 70%,accuracy is 71.67%,and the area under curve (AUC) is nearly 0.786 4.Obtained results show that the image spatial information can effectively express the characteristics of SPNs.The system classification accuracy of benign and malignant SPNs is up to 71.67% without medical signs,and the classification speed is about 50 times faster than that of traditional texture methods.It provides a feasible,simple and objective method for solving the problem in medical imaging diagnosis of the SPNs.

The design and fabrication processes of a DC-contact shunt RF MEMS switch were studied to improve its performance.A low stress electroplated Au membrane was taken as an upper electrode to implement the Au-Au direct contact.BorofloatTM glass was used as a substrate and a inside resistance was used to isolate the crosstalk between radio-frequency signal and bias voltage.Then,the distance between Au membrane and Coplanar Waveguide(CPW) was optimized to lower the insertion losses of the switch.Experimental results show that the contact resistance is 0.1 Ω under the pull-down voltage of 60 V.The insertion losses are -0.03 dB@1GH,-0.13 dB@10 GHz,-0.19 dB@20 GHz in the frequency ranges from DC to 30 GHz,which is less than -0.5 dB;the isolations are -47dB@1 GHz,-30 dB@10 GHz and -25 dB@20 GHz in the same frequency ranges,which is all better than -23 dB.The measurement results show that the fabricated RF MEMS switch is suitable for the applications of frequency ranges from DC to 30 GHz.