2018 Vol. 47, No. 7
column
- Outstanding doctoral dissertation in optical engineering
- Invited column-“Deep learning and its application”
- Laser technology and application
- Infrared technology and application
- Photoelectric measurement
- Optical design and simulation
- Photoelectric device and application
- Optical communication and optical sensing
- Information acquisition and identification
2018, 47(7): 734001.
doi: 10.3788/IRLA201847.0734001
A switchable and stable single- and dual-wavelength C+L band Erbium-doped fiber laser based on FBG array fabricated by femtosecond laser was proposed and experimentally demonstrated. Fiber grating period of 538, 542, 547 nm was realized on polyimide fiber core through fiber protection and cladding by femtosecond laser, grating length was 3 000 m, and reflection spectrum of each FBG was 1 555.5, 1 569.6, 1 583.8 nm, respectively; the length of 3 m C-band Erbium-doped fiber and 10 m L-band were used as the gain medium, and the proposed fiber laser was composed of gain medium, pump source, FBG, PC and broadband reflection mirror. In the experiment, the laser working threshold was 35 mW, through adjusting the polarization controller, a switchable single-wavelength of 1 555.4, 1 569, 1 583.2 nm laser was obtained, 3 dB linewidth was 0.05 nm, and SNR was more than 35 dB; The spectral stability of single-wavelength laser was tested in the experiment, and it was found that the power fluctuation of each single-wavelength laser was less than 0.98 dB over 10 min; by adjusting the polarization controller, the proposed fiber laser had the capability to produce switchable 1 569, 1 583.2 nm and 1 555.4, 1 569 nm dual-wavelength laser simultaneously, and the power shift was less than 1.14 dB and 4.48 dB, respectively over 10 min monitoring time.
2018, 47(7): 703001.
doi: 10.3788/IRLA201847.0703001
The traditional classification methods of PolSAR image generally required the feature extraction in the early stage, involving more human participation, and the classification accuracy needed further improvement. In addition, when using supervised classification method, there were sometimes small sample problems. In view of these problems and combining the requirement of PolSAR image fine classification, a PolSAR image classification method based on 3D convolution neural network was proposedr. The traditional convolution neural network was extended to three dimensions and applied to PolSAR image classification, and the classification method was described in detail. Thus, the characteristics of the multichannel PolSAR image could be fully excavated and improve the classification performance. Moreover, the method of virtual sample expansion was used to improve the small sample situation of certain category and get better classification results. Experimental results showed that 3D convolution neural network could get better performance than 2D convolution neural network in PolSAR image classification and the virtual sample expansion method could effectively improve the small sample classification problem.
2018, 47(7): 703002.
doi: 10.3788/IRLA201847.0703002
In recent years, deep learning techniques have achieved great breakthrough for its powerful non-linear computations in the tasks of target recognition and detection. Existing deep networks were almost designed based on the precondition that the visual data reside on the Euclidean space. However, many data in computer vision have rigorous geometry of manifolds, i.e., image sets can be represented as Grassmann manifolds. The deep network was devised based on the non-Euclidean structure of the manifold-valued data, which combined the differential geometry and deep learning methods theoretically. Furthermore, a deep network for image-set recognition based on the Grassmann manifold was proposed. In the training process, the model was updated by the use of the backpropagation algorithm derived from the matrix chain rule. Learning of the weights can be transformed as the Riemannian optimization problem on the Grassmannian. The experimental results show that this method not only improves the accuracy of recognition, but also accelerates the training and test process in one magnitude.
2018, 47(7): 703003.
doi: 10.3788/IRLA201847.0703003
In order to improve the accuracy of infrared fault image recognition for electrical equipment, an infrared fault image recognition method for electrical equipment based on double supervised signal deep learning was proposed. Firstly, a Slic super pixel segmentation algorithm was adopted to merge the similar pixel regions into blocks. According to the luminance information provided by the improved HSV space transformation, the temperature abnormal regions were determined. Secondly, the connected areas and the corresponding device of this region were separated. Finally, based on the GoogLeNet convolution neural network model, fault features of infrared images for electrical equipments were extracted, then trained and supervised by two kinds of signals, i.e., the softmax loss and the center loss signal. Among an established 700 infrared fault of electrical equipment images dataset, 500 of which are for training, and 200 for testing. Experiments results show that the test accuracy rate can reach to 98.6% which enhanced 1% when being compared with the classic method simply using the single softmax loss. The algorithm can accurately locate five kinds of electrical equipments which include the transformer bushing, current transformer, surge arrester, isolating switch, insulators, as well as identify the corresponding faults.
2018, 47(7): 703004.
doi: 10.3788/IRLA201847.0703004
Since the current image quality assessment methods are generally based on hand-crafted features, it is difficult to automatically and effectively extract image features that conform to the human visual system. Inspired by human visual characteristics, a new method of full-reference image quality assessment was proposed by this paper which was based on convolutional neural network (DeepFR). According to this method, the DeepFR model of convolutional neural network was designed which was based on the understanding of the dataset by itself using the human visual system to weight the sensitivity of the gradient, and the visual gradient perception map was extracted that was consistent with human visual characteristics. The experimental results show that the DeepFR model is superior to the current full-reference image quality assessment methods, its prediction score and subjective quality evaluation have good accuracy and consistency.
2018, 47(7): 703005.
doi: 10.3788/IRLA201847.0703005
In order to improve the small object detection ability of SSD object detection algorithm, the transposed convolution structure in SSD algorithm was proposed, the low resolution high semantic information feature map was integrated with high resolution low semantic information feature map using transposed convolution, which increased the ability of low level feature extraction and improved the average accuracy of SSD algorithm. At the same time for the problem that SSD algorithm model being large, running memory consumption high, without running on the embedded equipment ARM, a lightweight feature extraction minimum unit was proposed based on DenseNet, combining depthwise separable convolutions, pointwise group convolution and channel shuffle, running on the embedded equipment ARM cloud be realized. The comparative experiments on PASCAL VOC data set and KITTI autopilot data set show that the mean average is significantly improved by improved network structure, and the number of model parameters is effectively reduced.
2018, 47(7): 706001.
doi: 10.3788/IRLA201847.0706001
In order to study characteristics of CO2 laser spectrum in 9R band and satisfy the need of laser interaction with matter, technology of coating for selecting wavelength was used firstly to make laser in 9R band output lonely from TEA CO2 laser. Then through single pulse discharge experiment, wavelength of CO2 laser and quantity of laser spectrum was analyzed by CO2 laser spectrum analyzer. Energy meter was used to measure the proportion of each laser spectrum in the single pulse energy. The changes about wavelength, quantity and energy proportion of CO2 laser spectrum in 9R band were studied by changing transmittance of output mirror, curvature radius of total reflector and proportion of working gases. Experiment results show that several adjacent CO2 laser spectrums in 9R band have been output meanwhile, and these spectrums can be changed with the change of transmittance of output mirror, curvature radius of total reflector and proportion of working gases.
2018, 47(7): 706002.
doi: 10.3788/IRLA201847.0706002
The effect of laser power, scanning speed and powder feeding rate on feature sizes of weld height, weld width, weld depth during laser deposition repair single track GH738 alloy were investigated by orthogonal test; the parameters were optimized and defect-free microstructure was obtained; the reason of defects such as ill bonding cracks in deposition zone and the characters of microstructure microhardness were analyzed. The results indicate that the microstructure of GH738 deposition zone is made of columnar dendrites which grow epitaxially from the substrate. Few carbides are precipitated in the interdendrite. Compared with the substrate, the carbide in the heat affected zone is obviously reduced;meanwhile the size of ' phase in heat affected zone increases obviously. The microhardness of deposi-tion zone(350-470 HV0.3) and heat-affected zone(450-480 HV0.3) is lower than substrate(480-510 HV0.3),and microhardness decrease gradually along the deposition height direction.
2018, 47(7): 706003.
doi: 10.3788/IRLA201847.0706003
In order to meet the needs of length on-line measurement in the process of cable production, a cable length on-line meter calculating system based on laser Doppler velocimetry was designed. The basic principle of laser Doppler velocity measurement was described, a compact double-beam double-scattering optical path was designed. A filter band adaptive selection algorithm based on finite impulse response (FIR) filter was proposed, which effectively filtered out the base signal and partial noise, reduced the influence of Doppler signal noise and improved the signal-to-noise ratio. The use of energy barycenter correction technology for spectral correction reduced the measurement error of the system. The experimental device was set up, and the contrast experiment was carried out. The experiment results show that the system could measure the length in real time and accurately, and the relative error is less than 0.1%.
2018, 47(7): 706004.
doi: 10.3788/IRLA201847.0706004
The laser cleaning experiment of 38CrMoAl samples was studied by pulsed laser. The influence law of cleaning threshold, damage threshold and laser power density of surface pollution of 38CrMoAl samples was studied by microscope, white light interferometer, step profiler and infrared thermometer. The result shows that when the laser power density increases to 1.22 J/cm2, the surface contaminants begin to be effectively removed, corresponding to the cleaning threshold, when the power density increases to 3.11 J/cm2,the substrate begins to be damaged, corresponding to the damage threshold; When the energy density between the cleaning threshold and the damage threshold, the cleaning mechanism is vibration effect; When the power density equals to 2.36 J/cm2, all the contaminants on the sample are removed, the roughness is not changed after cleaning, the substrate is not damaged. Laser cleaning can satisfy the fine cleaning needs of 38CrMoAl parts.
2018, 47(7): 706005.
doi: 10.3788/IRLA201847.0706005
The thermal expansion of the object reflects the property of the material itself, when solid is being heated, linear expansion is usually described as a one-dimensional changing in the length. Measurement of linear expansion coefficient of materials was not only important in their developments, but also one of the most important standards in choosing them. Combining Doppler effect with laser heterodyne technology, a novel measurement method of multi-beam laser heterodyne for metal linear expansion coefficient was proposed, which was converted into the measurement of the length variation of linear expansion coefficient. Based on Doppler effect of oscillating mirror, the information of length variation was loaded into the frequency difference of the multi-beam laser heterodyne signal by the frequency modulation, and many value of length variation could be acquired simultaneously after the multi-beam laser heterodyne signal was demodulated. Processing these values by weighted-average, length variation of the sample versus temperature could be obtained accurately, and eventually the measuring of linear expansion coefficient of metal was improved. Simulations for linear expansion coefficient of metal rod under different temperatures had shown that the relative measurement error of this method is just 0.1%. The measuring accuracy was improved by one order of magnitude compared with traditional measurement methods.
2018, 47(7): 706006.
doi: 10.3788/IRLA201847.0706006
In a laser emission system, the high accurate docking of laser and theodolite ensures the high-precision orientation laser emission. For the under-the-machine docking, the Coude mirror space normalization theoretical model of the overall coordinate system based on the Coude optical path which transmits laser was established, and it accurately calibrated the position and variation of the laser in each Coude mirror space. Based on this theoretical model, a high accurate docking technology of laser and theodolite was proposed, which achieved the minute level of inner field and the second level of outside field. In a experiment of a long wave laser emission system, the technology realized a fast accurate docking which has an error of less than 2. Its correctness and practicality are verified from the engineering, it also provides the technical support for this type of docking of laser and theodolite.
2018, 47(7): 706007.
doi: 10.3788/IRLA201847.0706007
In order to solve the problem of generating optical frequency comb by mode-locked laser seed source, a new method to generate optical frequency comb by continuous light source was presented and the problem of phase mismatch in four-wave mixing and cascaded four-wave mixing with wide spectral range was solved by using dispersion-flattened highly nonlinear optical fiber. We experimentally demonstrated efficient generation of an equalized optical comb with nearly 40 nm bandwidth. The comb was generated by low-power, low-cost, continuous-wave seeds(FP-LDs) without needing for pulsed laser sources. The CW frequency comb were generated by four-wave mixing(FWM) and cascaded FWM whose bandwidth of the spectrum was expanded by nearly 10 times and the linewidth of the frequency comb was 4.3 MHz with 420 m zero-slope dispersion dispersion-flattened HNLF.
The Al2CrFeCoCuNixTi high entropy alloy coating was prepared by laser cladding on Q235 steel surface. The microstructure of Al2CrFeCoCuNixTi high entropy alloy coatings were analyzed, and the corrosion resistance of Al2CrFeCoCuNixTi high entropy alloy coatings in 0.5 mol/L HNO3 solution and 0.5 mol/L HCl solution were tested. The experimental results show that the Al2CrFeCoCuNixTi high entropy alloy coating can be divided into cladding zone, bonding zone and heat affected zone. The microstructure of cladding zone is mainly composed of equiaxed grain, with micron-grade particles distributed on it, the phase structure of Ni1.0 high entropy alloy coating was simple for FCC and BCC structure. Due to the passivation effect of Cr and Ni elements and formation of Al2O3 and Al2O3H2O films by Al element, the Al2CrFeCoCuNixTi high entropy alloy coating exhibits good corrosion resistance in 0.5 mol/L HNO3 solution and 0.5 mol/L HCl solution. Compared with Q235 steel, the corrosion current density decreases 1-2 orders of magnitude. Pitting corrosion appears because of Cl- in 0.5 mol/L HCl solution can penetrate the passive film on the Ni0.5 high entropy alloy coating surface.
2018, 47(7): 704001.
doi: 10.3788/IRLA201847.0704001
Taking a fighter aircraft as the research object, the flow field calculation model of the aircraft was established by the three-dimensional modeling and meshing process, and the numerical simulation of the flow field characteristics of the aircraft was carried out based on the commercial CFD software ANSYS Fluent 16. The effect of solar radiation on the temperature field of the fuselage was considered by using the solar ray tracing algorithm, the discrete ordinates method for DO radiation model was used to calculate the iterative calculation of the radiation transfer equation, simulating the after-combusting tail flame injection process with component transport model without chemical reaction (Species Transport Model), and the temperature, concentration and component distribution data of the outflow field of the aircraft were obtained. The influence of solar radiation on the whole temperature field distribution, the influence of the Mach number on the infrared radiation of the flow field and the distribution of tail flame flow field were simulated and analyzed. Analysis shows that the effect of solar radiation on heating of the skin is only about 5 K. With the increase of the number of Mach aircraft fuselage and the difference of infrared radiation intensity is highest in the abdomen, abdominal radiation intensity is about 2 times the maximum radiation intensity. Under the shock wave, there will be the core high temperature region with the highest two 450 K and 580 K discontinuity at the rear of the tail flame. The infrared radiation intensity distribution of the plume conforms to the pear shape distribution trend.
2018, 47(7): 704002.
doi: 10.3788/IRLA201847.0704002
At present, the missile recognition algorithm based on the space-based hyperspectral image is considered from the overall analysis of the spectral curve. The required data is large and the algorithm is redundant. In order to solve the problem of insufficient algorithm, the influence of missile launching device, fuel composition and temperature pressure on the tail flame spectrum from the influence factors of missile tail flame spectrum was analyzed firstly, and the radiation data at the characteristic interval under certain assumptions was used to obtain the conclusion of missile identification. On this basis, the fuzzy recognition algorithm was introduced. The algorithm made full use of the radiation intensity and the spectral line information. The recognition result is the membership probability of each type missile. Through analyzing the spectral angle, it is found that it is too dependent on the line type, and the effect of the line recognition is poor by comparing the fuzzy recognition result with the spectral angle measurement, the superiority of the fuzzy algorithm in the spectral recognition is proved.
2018, 47(7): 704003.
doi: 10.3788/IRLA201847.0704003
In order to get the low transmission rate in both mid and far infrared atmospheric windows, an infrared frequency selective surface(FSS) composed of two layers based on hexagonal ring structure was designed. Simulation analysis using CST electromagnetic software shows that the frequency selective surface has two stopbands in both 3-5 m and 8-14 m, of which the average transmission is lower than 2.5%. Therefore, this frequency selective surface realizes two stopbands in infrared wavelengths; the filtering mechanism of the frequency selective surface was analyzed based on the method of surface current model analysis and the effective medium theory. And it was found that the inducted current on the surface of units was symmetrical in different resonance points, which enhanced the total scattering-field and led to the enhanced reflection. That was the ideal infrared stopband. The inducted current was produced by the coupling between screen and the coupling between unit in a screen; In the end, the effect of electromagnetic wave polarization mode, incident angle, dielectric layer properties and structural parameters on the transmission properties of the frequency selective surface was researched. The results reveal that the transmission of FSS is insensitive to polarization mode and incident angle, while the dielectric layer properties and structural parameters are important factors to the transmission.
2018, 47(7): 704004.
doi: 10.3788/IRLA201847.0704004
Surface source infrared decoy is regarded as the new type weapon to confront infrared guided missile. In comparison with traditional infrared decoy, it has obvious superiority in burning rate, diffusion range, jamming effectiveness and so on. It's of great significance to study the anti-interference mechanism of infrared guided missile for improving the jamming effectiveness. Firstly, the models of missile motion, seeker tracking and the optimization algorithm of tracking were introduced. Secondly, the models of flight motion and radiation were established, as well as models of surface source infrared decoy diffusion and infrared image. Finally, the tactical parameters of missile was selected and the indicator of effectiveness evaluation was formulated, based on which, comparing the simulation results with testing data to verify the reasonability of simulation result. It's concluded from simulation that the simulation results of the influence of tactical parameters of missile on jamming effectiveness of surface source infrared decoy match the testing data, being of great reference value.
2018, 47(7): 704005.
doi: 10.3788/IRLA201847.0704005
In order to establish a method to quickly and accurately distinguish and inspect disposable chopstick plastic packaging bags extracted from the site, an infrared spectrometer and an experimental method of Smart Performer sampler were used to analyze the collected samples of disposable chopstick packaging bags of 40 different brands, different places of sale and different batches. According to the difference of the peak number, peak position and peak strength of the absorption peak in the infrared spectrogram, the collected samples of the chopsticks plastic packaging bag can be distinguished scientifically and effectively by analyzing and summarizing the differences of the infrared spectrum characteristics of various chopsticks plastic packaging bag samples. The reproducibility of the samples, the minimum amount of detection and the influence of contamination were also discussed. The experimental results show that the test method of disposable chopsticks plastic bags samples is simple rapid, accurate and reliable, reproducible, without sample preparation and nondestructive examination.
2018, 47(7): 717001.
doi: 10.3788/IRLA201847.0717001
In order to solve the problems of complicated process, bulky and expensive apparatuses in multi-dimensional position detection of a moving body, an optical non-contact measurement method of simultaneously measuring four parameters of a moving object was proposed. The system consisted of a combined surface reference composed of a rotating paraboloid and a plane, and an optical probe. Based on the principle of plane mirror reflection and the features of combined surface reference, the angle measurement models of the pitch angle, the yaw angle and the two-dimensional displacements of the moving object were established respectively, and the accuracy evaluation and the repeatability experiments of the system were completed. The experimental results show that the angle accuracy of the system is 1.5, and the displacement accuracy is 1 m, laying a foundation for the error identification of machine tools.
2018, 47(7): 717002.
doi: 10.3788/IRLA201847.0717002
According to the statistics, 39.65 billion tons of coal has been consumed up from China in 2015. Obviously, CO2 emission from power generation and thermal supply industry segment contributes a large part to global warming. Therefore, it is of great importance for online monitoring of CO2 emissions from coal-fired power plants. The CO2 concentration was measured by tunable semiconductor laser absorption spectrum(TDLAS), which had the benefits of high accuracy, rapid response, non-contact, etc. In order to prevent interference from absorption of other gases, a distributed feedback laser with a center wavelength of 1 580 nm was chosen as the light source. Simulated flue gas was made referring to the concentration of CO2 in the flue gas of the power plant. The direct absorption(DA) and wavelength modulation(WM) methods were used to measure the CO2 concentration. The results show that the relative standard deviation(RSD) for DA and WM are 0.94% and 0.22%, while the maximum relative errors and detection limits for DA and WM are 2.64% and 1.65%, 0.013 6% and 0.001 4%. It is found that WM is better than DA in measurement qualities. But the WM method is influenced by the calibration method in field application and the line width is interfered by the fluctuations of the parameters such as pressure and temperature, resulting in reduced long-term stability. In the case of the DA method, the concentration can be calculated directly by measuring the absorbance, line strength and pressure without calibration. It has a high signal-to-noise ratio for the measurement of gas with high concentration. Meanwhile, the measurement accuracy is high enough to meet the requirement for CO2 online measurement in power plants. Therefore, the DA method is a better choice for the CO2 measurement in the flue gas of power plant boiler.
To effectively monitor the insulation state of the optic-electric composite submarine cable, the IEC60287 cable' ampacity and temperature calculation standard were improved by iteration algorithm and the temperature field model of 110 kV YJQ41300 mm2 submarine cable was established with the finite element analysis, solver and simulation software package COMSOL based on the thermoelectric coupling model. The results were validated by the improved IEC60287 and the optimal model size of the submarine cable at different values of ampacity was obtained. The effects of ampacity and ambient temperature on the optical fiber's temperature rise were investigated in normal insulation state and at different levels of insulation degradation of the submarine cable. Dielectric loss factor was used to characterize the insulation states, the change of optical fiber's relative temperature rise under the typical value of dielectric loss factor was investigated. The results reveal that, under normal circumstances, the temperature rise of fiber is approximately proportional to the square of ampacity and it increases approximately linearly with increasing ambient temperature with a slow velocity. When the cable's insulation degraded, the relative temperature rise of the optical fiber is proportional to the tan. Comparatively, the ambient temperature has little influence on it. According to above research, a formula for estimation of dielectric loss factor based on relative temperature rise and ampacity of cable is presented and a method for insulation degradation evaluation of the submarine cable based on relative temperature rise of optical fiber is proposed. Compared with the existing insulation monitoring methods based on electrical measurands, the error caused by electromagnetic interference can be avoided in proposed method, which can be applied in the online monitoring of the optic-electric composite submarine cable.
2018, 47(7): 717004.
doi: 10.3788/IRLA201847.0717004
In order to predict the imaging quality of space cryogenic optical system, a method of high accuracy measuring the wave front aberration of the optical system with low F number and short back focal in cryogenic vacuum environment was presented. Firstly, the optical path was designed. The layout of cryogenic optical system, interferometer and flat mirror were prepared for wave front aberration testing. Then, the key components such as cryogenic vacuum standard lens, standard flat mirror and atmospheric window glass were analyzed and designed. The testing wave front aberration was removed as a system error term. Finally, the wave front aberrations of normal pressure and temperature and cryogenic vacuum environment (Temperature:100 K, Pressure:110-4 Pa) were obtained by debugging the optical path. The accuracy test showed that the deviation between the measured value and the standard value was 0.010(=632.8 nm), and the difference was very small, which proved the feasibility of the test method. The wave front aberration difference between the two stages was small. This method solved the difficult problem of the cryogenic vacuum optical system wave front aberration testing with low F number and short back focal in the cryogenic vacuum environment, which couldn't be accurately tested or even tested.
2018, 47(7): 717005.
doi: 10.3788/IRLA201847.0717005
The optical axis parallelism was an important parameter for the multi optical axis optoelectronic device. It was very necessary to detect the parallelism. With the gradual increase in the complexity of optoelectronic devices, the distance between optical axes was increasing. The traditional small aperture detection methods could no longer meet the requirements of existing photoelectric devices. In order to detect the parallelism between large space optical axis, a expanding component based on double pentaprism structure was designed to realize the translation of the beam emitted from the collimator, and the error of the expanding component was modified by double wedge structure. The effective aperture of the collimator was extended from 300 mm to 1 200 mm. Mathematical modeling analysis and actual adjustment of the expanding component were carried out and the precision experiment was done. The experimental results of the detecting system show that the beam emitted by the collimator could maintain good parallelism after expanding. The parallelism deviation is within 11, which could meet the precision requirements of large space optical axis detection.
2018, 47(7): 717006.
doi: 10.3788/IRLA201847.0717006
SCIATRAN, providing more reliable simulation data for retrieval of CO2 concentration, is a high spectral resolution radiative transfer model with rich parameters and easily calibrated. In this article, SCIATRAN was used to simulate the CO2 radiation intensity of different aerosol types, aerosol optical thickness and surface types, and the regularity of influence of CO2 radiation intensity under different conditions was analyzed. The results show that:first the influence ratio of aerosol type on CO2 radiation intensity is less than 8%, because the urban aerosol has complex composition, the effect on CO2 radiation intensity is slightly larger than that of country type and ocean type aerosol; then, the influence of aerosol optical thickness on CO2 radiation intensity is -3.52%-+42.97%, and the smaller the optical thickness, the greater the radiation intensity; Last, the surface types have the greatest influence on the CO2 radiation intensity, and the difference between the result and the reference value is up to 166.43%. The CO2 radiation intensity increases when the surface albedo increases. This article also verifies the feasibility and effectiveness of the simulation signal based on SCIATRAN by comparing its analog signal to the measured signal.
2018, 47(7): 717007.
doi: 10.3788/IRLA201847.0717007
Raman lidar has been designed for the measurement of vertical and temporal distribution of aerosol optical properties, atmospheric temperature and water vapor. In order to investigate characteristics of aerosol boundary layer(ABL) height in Beijing, lidar system had been installed in the University of Chinese Academy of Sciences from November 2014 to January 2015. The data obtained by Raman lidar had been used to derive the ABL height based on the gradient method and the ABL height was compared with particulate matter(PM) data provided by the Ministry of Environmental Protection of the People's Republic of China. A total of 15 days of haze, 27 days of pollution and 24 days of clean occurred through the entire period of observation. On haze, pollution and clean days, the average heights of the ABL were 0.6-0.9, 0.9-1.3 and 1-1.9 km, respectively. And the concentrations of surface PM2.5 were 143-203, 77-90, 17-34 g/m3, the PM10 concentrations were 170-271, 103-153, 33-78 g/m3. The measurement results show the height of ABL has a negative correlation with the concentration of surface PM. The rate of PM concentration variations increase gradually with the height of ABL in clean, pollution and haze days. The rate of PM2.5 average concentration in haze days(-242.4 gm-3/km) is more than 2 times than the rate in pollution days(-114.8 gm-3/km), 3 times than the rate in clean days (-77.4 gm-3/km). The rate of PM10 average concentration in haze days(-224.2 gm-3/km) is more than 2 times than the rate in pollution (-117.6 gm-3/km) and clean days (-90.4 gm-3/km).
2018, 47(7): 718001.
doi: 10.3788/IRLA201847.0718001
In order to meet the needs of wide area search using agile satellite, the aerospace camera circular scanning imaging model was designed. Based on analyzing the principle of the circular scanning imaging, the optimal ground track was designed and the relationship between the critical spin speed of the satellite and the orbital velocity, the critical circular scanning coefficient, the overlap rate was analyzed, the circular scanning model was established. At the same time, the exposure time, frame rate and other imaging parameters were determined based on the ground track design. The imaging model by STK software was simulated, and the imaging geometric parameters were analyzed. The result shows that when the orbital height H is 500 km, the pixel size a is 4 m, the focal length f is 1 m and the axial pixel number M is 50 000, the ground sample distance and swath width increase gradually with the increase of the camera inclination , when equals to 10, 20, 30, 40, the swath width raises to 1.96, 3.10, 4.58, 6.85 times compared with nadir imaging.
2018, 47(7): 718002.
doi: 10.3788/IRLA201847.0718002
Based on the fabricated and assembled lidar telescope, the structure of its primary mirror subsystem was presented. The effects on the primary mirror surface of the mismatch among the assembling points on the second supporting board and the malfunction of the flexible supporting foot on the first board were analyzed. The optomechanical analysis was conducted through establishing the finite element model of the primary mirror, the analyzing results were compared with the measured wave front map during the assembling. According to the analysis, the mismatch of the assembling points on the second baseplate was the main reason to deform the primary mirror surface. The primary mirror subsystem structure was optimized through integrated optomechanical analysis. And the expected root mean square deformation of the primary mirror should be decreased from 0.3 to 0.087, which meets the specification requirement of 0.15.
2018, 47(7): 718003.
doi: 10.3788/IRLA201847.0718003
In order to meet the requirement of high precision and mass manufacturing of large aspheric lens for high power laser device, a novel manufacturing method combining the bonnet polishing(BP) technology and flexible pitch polishing(FPP) technology was explored. Firstly, after the aspheric generating by ultra-precision grinding, the bonnet polishing technology, in keeping the aspherical surface shape, was used to quickly remove the grinding defect layer and improve the roughness of the element so that it can be directly tested by interferometer. Then, the bonnet polishing technology was used to quickly correct the low frequency error. Finally, the flexible pitch tool was adopted to smooth the middle and high frequency error. During the polishing process, the self-built wavefront detection system and the roughness instrument were used to detect the full frequency error of the aspheric element. Based on the above-mentioned fabrication and testing method, a 430 mm430 mm off-axis apherical lens was manufactured, and the results indicate that, after polishing, the PV, GRMS, PSD1 RMS, PSD2 RMS and Rq are 0.1, 5.7 nm/cm, 1.76 nm, 1 nm and 0.6 nm, respectively. In addition, the power spectral density(PSD) curves are below the required evaluation curve. The experimental results show that the full frequency specifications of the off-axis aspheric lens meet the requirements. The manufacturing method is also applicable to the high-precision manufacturing of other types of large aspheric optical elements.
2018, 47(7): 718004.
doi: 10.3788/IRLA201847.0718004
With the increasing requirement of space information acquisition ability, the dynamic high-resolution remote sensing has become a new hotspot research in space optical fields. The field-bias coaxial three-mirror optical system has the advantage of long focal length, small volume, high level of lightweight and high quality image, which can satisfy the high-resolution, multi-spectrum, versatility and low cost requirement for the LEO video satellite, so it is widely used in the field of the dynamic high-resolution space sensing field. Based on the Gaussian optics and three-mirror aberration theory, the common aperture optical system with area array imaging in visible spectrum and linear push-broom imaging in NIR and MIR spectrum was designed. The focal length for visible spectrum is 4.1 m, the focal length for NIR spectrum is 2.6 m and the focal length for MIR spectrum is 1.85 m, the aperture of three systems is 520 mm, field of view is 0.60.6. The optical structure is convenient for area array imaging, and the imaging quality of the optical system approaches the diffraction limit. The total length of the optical system is less than f'visible/3.7, and is easily implemented because of the relatively loose tolerance about processing and assembling.
2018, 47(7): 718005.
doi: 10.3788/IRLA201847.0718005
Fresnel lens is one of the most common solar concentrators. The performance of the curved Fresnel lens is usually better than flat Fresnel lens. Based on the principle of non-imaging optics, a new curved Fresnel lens design method was proposed in this study. The curved Fresnel lens was on conical surface. In the premise of meeting the mechanical requirements of the Fresnel lens, the slope of the second surface of the Fresnel lens was solved. Based on the objective of manufacturability, different shapes of Fresnel lenses were obtained by this method to analyze the effect of structural parameters of curved Fresnel lens on concentration ratio, acceptance angle and the illumination uniformity by optical simulation. The effect of machining error on optical efficiency in ultra-precision machining was also analyzed. The design method provided a new way for the parametric analysis of curved Fresnel lens. The simulation results show that the small aspect ratio of curved Fresnel lens will get good uniformity and high energy efficiency.
2018, 47(7): 718006.
doi: 10.3788/IRLA201847.0718006
In order to meet the requirements of velocity control performance of servo system for K mirror in the 2 m telescope system, a new active disturbance rejection-controller(ADRC) based on adaptive control law parameter was proposed. Firstly, the second-order linear extended state observer based on first-order velocity model was designed; And then, in order to improve the dynamic and steady-state performance of speed-loop, proportion controller's parameter based on regression analysis method was designed into a function and it will be automatically adjusted with the variation of input speed. Finally, experiment system based on K mirror servo control was set up, the experiment was performed under the excitation of step signal. Actual results show that compared with the PI and ADRC controller, the regulation time for 0.001 ()/s step response is reduced from 7.3 s, 3.2 s to 0.9 s. The overshoot for 10()/s step response is reduced from 8%, 62% to 0; The best disturbance inhibition ability is increased 23 dB among the low and middle frequency band, satisfying the high accuracy speed control performance requirements of K mirror servo system.
2018, 47(7): 718007.
doi: 10.3788/IRLA201847.0718007
For the space camera on orbit active optics requirement, the position and orientation of secondary mirror relative to the primary mirror need to be adjusted. Based on the parallel robot joint space method, the motion control system of 6-PSS Stewart platform secondary mirror adjusting mechanism was designed. By use of DSP and FPGA as the core processor, linear encoder as the feedback element and the integrated three-phase bridge as driving element, the motion control circuit was completed. Based on the top level inverse kinematics model and the bottom level link control system, the motion control algorithm of the secondary six DOF adjusting mechanism was accomplished, the control parameters were easy to be adjusted and the algorithm was easy to be used in the engineering, the requirement of high reliability adjustment for space motion mechanism was satisfied. Experimental results indicate that the 0.7 m and 3 motion accuracy can be achieved, the demand of space camera active optics can be satisfied.
2018, 47(7): 720001.
doi: 10.3788/IRLA201847.0720001
The Joule heating inside organic light-emitting devices(OLEDs) is part of the device degradation factors. Therefore, it is very important to study the thermal characteristics of OLEDs for effective heat dissipation. So the thermal characteristics of OLED were simulated via the solid heat transfer module of COMSOL finite element analysis software. It is found that the device temperature increases linearly with the input power. At driving current of 150 mAcm-2, the highest temperature of the Alq3 emitting layer and bottom surface of the glass-substrate are 82.994 3℃ and 77.392 6℃, respectively, the highest temperature at the central region of the cathode surface is 82.994 2℃, and its average temperature is 78.445℃. Thermal analysis simulated results show the temperature distribution and thermal characteristics of an OLED can be affected by changing the thermal conductivity and thickness of functional layer, convection heat transfer coefficient, and surface emissivity. When increasing the substrate thermal conductivity, the temperature of the OLED is significantly reduced, and the surface and internal temperature gradients are greatly reduced; When improving the air convection heat transfer coefficient and the surface emissivity of the substrate, the temperature of the OLED can be greatly reduced. However, other parameters have no significant effect on improving thermal characteristics for the OLED.
2018, 47(7): 720002.
doi: 10.3788/IRLA201847.0720002
A low-light-level CMOS image sensor with Capacitive Trans-impedance Amplifier(CTIA) pixel circuit and Delta Double Sampling(DDS) was proposed. By using CTIA circuit, stable bias voltage of the photodiode and high injection ratio can be realized, and the weak signal in low-light-level condition can be readout; Meanwhile, an off-chip digital DDS was used to reduce the fixed pattern noise (FPN) which was realizing the subtraction algorithm between the reset signal and pixel signal after A/D conversion off-chip, and improve the image quality of the low-light-level CIS. This low-light-level CMOS image sensor (CIS) based on CTIA pixel circuit was implemented in the 0.35 m standard CMOS technology. The pixel array was 256256, and the pixel size was 16 m16 m. The experimental results show that this low-light-level CIS can capture recognizable images with the illumination down to 0.05 lx.
2018, 47(7): 720003.
doi: 10.3788/IRLA201847.0720003
The connected defective elements identifications has always been the research difficulty of focal plane array (FPA) detector. It is difficult to identify connected defective elements by FPA test-bench because the response voltage of connected defective elements is basically the same as that of normal elements. Novel optical filter for connected defective elements identifications was proposed. The presented filter had sorted elements of FPA detector into two kinds of detection units. The two kinds of detection units were designed in pairs and staggered arrangement closed to each other. The response voltage of the connected defective elements was 50% of that of normal elements. The connected defective elements were identified markedly by using the proposed filter.
2018, 47(7): 720004.
doi: 10.3788/IRLA201847.0720004
A new optical salinity sensor was proposed, which was designed by an optical collimator, based on a two-channel back reflection technique, and especially suitable for the long distance measurement of salinity (salinity was demonstrated by different concentrations of sodium chloride in de-ionized water). To demonstrate the performance of this new sensor, a comparative experiment was conducted. The experimental system was based on the structure of two Fresnel-back-reflection channels, one of which was designed with the measuring sensor, and the other was a fiber end with protective cladding. The main innovation and advantages of this sensor were that the common collimator could be directly used as a salinity sensor. Especially, the experimental results showed the signal to noise ratio(SNR) of this measuring method by using the collimator-sensor was at least 6.11 dB more than the common method by using fiber-sensor and the equivalent extended measurement distance of the system was at least extended by 30 km. The experimental results also show the temperature influence was estimated to be 0.005 16 dB/℃, which meant the temperature has little influence on the measurement, and also the influences resulted from fluctuation of light source and environment can be effectively eliminated.
2018, 47(7): 722001.
doi: 10.3788/IRLA201847.0722001
Phase-sensitive optical time domain reflection system(Ф-OTDR) is a new type of distributed fiber perturbation sensing system. With the continuous improvement of application requirements, the detection and positioning of external intrusion can not satisfy the actual needs, the accurate detection of the signal identification is urgently needed. When detecting the intrusion signal, it is a key problem in the research of distributed optical fiber disturbance sensor system how to accurately distinguish the classification of intrusion events and reduce the false alarm rate and false negative rate. The principle of distributed fiber perturbation sensing system was briefly introduced, the existing methods of feature extraction and classifier design were summarized and classified, and the recognition results were summarized and compared to facilitate the research. According to the difference of the application environment and the characteristics of the signal to be measured, the accurate selection of the appropriate signal pattern recognition method is adopted to promote the researchers to study the pattern recognition method of distributed fiber perturbation sensing system more deeply.
2018, 47(7): 722002.
doi: 10.3788/IRLA201847.0722002
In order to meet the requirement of atmospheric pressure measurement in the aviation field, based on the principle of the optical fiber Fabry-Perot (F-P) sensing and low coherence interference, an optical fiber F-P multi-channel pressure sensing system was proposed. First of all, the demodulation algorithm and basic principle were introduced, the calibration method and temperature compensation of the optical fiber F-P pressure sensor were analyzed in theory. The fitting error of the sensor under non-constant temperature condition was reduced to 0.134% F.S.. Then, in the wind tunnel environment within -4-4 sideslip angle range, the pressure measurement experiment was carried out on three monitoring points of the aircraft model. The results were compared with the simulation results using Ansys-Fluent software. In conclusion, the results show that the system using the optical fiber F-P pressure sensor and the Ansys-Fluent numerical simulation have the same trend of demodulation results, with the full range error of 0.38% F.S.. The experiments confirm that the system can provide reliable pressure data accurately and reflect the pressure of the aircraft model's monitoring points in the wind tunnel.
2018, 47(7): 722003.
doi: 10.3788/IRLA201847.0722003
The transport of intensity equation(TIE) offers an experimentally simple technique for computing phase information directly from several defocused images. In this work we developed the traditional TIE intensity acquisition system. In order to avoid the shifting of CCD, a quadratic phase pattern was displayed on the SLM to provide a lens effect for realizing different defocus distance by varying the focal length of the lens. Two kinds of phase imaging experimental configurations guided by two different theories were designed. In most instances, since a camera is used to capture images in which the phase modulation of lens cannot be ignored, the former intensity acquisition system was designed based on the single lens optical propagation system, and the relationship between the focal length and the image distance in the lens law was used to facilitate the varying of the defocus distance by changing the phase pattern displayed on the SLM. The latter was designed by locating a SLM in the Fourier domain of the 4f setup, according to the Fourier transform property of the lens and the Fresnel diffraction theory, the relationship between the defocus distance and the focal length of the quadratic phase pattern can be derived. The experimental results verify that the proposed phase retrieval methods are reasonable and correct.
2018, 47(7): 726001.
doi: 10.3788/IRLA201847.0726001
A positioning and monitoring system for the objects with long distance, large view field and binocular vision was a three-dimensional coordinate measuring system based on binocular stereo vision principle. There was a common problem that the low correct matching rate of multiple non-feature targets randomly appearing led to the reduced accuracy of target positioning. The key was judging the trigger start time correctly. For the situation of the same target appearing continuously in the same side of the image sequence, or the morphological characteristic of the target changing from large to small or splits, the use of centroid method was to eliminate pseudo targets and detect target centroid position correctly. Then image sequences had time synchronization, and multiple targets realized sparse matching based on polar constraint at the same time. The experiment proves that 9 frame live continuous images contain 8 targets, using the above method to complete the targets detection and automatic positioning, and improve the accuracy rate.
2018, 47(7): 726002.
doi: 10.3788/IRLA201847.0726002
The accuracy of the star image coordinates is closely related to the positioning accuracy of the digital zenith camera. On the basis of the principle of digital zenith camera imaging, the star image coordinate expression was deduced strictly. The error equations of the image coordinates under four error factors and composite error factor were deduced based on the star image coordinate expression. The analysis results showed that the errors were independent of each other. Star map data was obtained through the simulation under four error factors and composite error factor. The simulation results show that:The influence of the image coordinates of each star point caused by the focal distance error and the transposition error is different. Compared to average value of the coordinate change values of x and y, the maximum fluctuation values of x and y caused by the focal distance error were about 2.38 pixel and 3.04 pixel. The maximum fluctuation values of x and y caused by the transposition error were about 1.06 pixel and 1.41 pixel. The error of the optical axis tilt and the main point offset made the change of the star image coordinates as whole migration. In addition, a method for solving the error parameters was proposed. The calculated error parameters were used to compensate the star image coordinates. The longitude accuracy of the digital zenith camera was improved by about 1.98 m, and the latitude accuracy was improved by about 1.65 m.
2018, 47(7): 726003.
doi: 10.3788/IRLA201847.0726003
A method of vehicle identification in natural scene was proposed for the target recognition of moving vehicles. Firstly, the image difference technique was used to study the significant characteristics of the target vehicle, and encode the local features and the image of the learning target, then the salient detection of the target vehicle was realized according to the above two information. Secondly, aiming at the complexity of vehicle movement, a block projection matching method was used for global motion estimation and compensation, and the difference technique was used to detect the motion feature. Then the target vehicle's saliency features and the motion features were fused to obtain more accurate candidate target areas. Finally, the candidate region was further used to discriminate the target by using the visual features. The experiment shows that the method has good performance of target discrimination and can solve the problem of vehicle motion recognition in natural scene.
2018, 47(7): 726004.
doi: 10.3788/IRLA201847.0726004
In view of the end point of BOF smelting, there are many uncertain and unavoidable errors in the traditional judgment of flame by human eye. A method of BOF endpoint estimation was studied by recognition of light radiation with fuzzy support vector machine. A non-contact system was designed for light radiation acquisition of furnace mouth. Based on the analysis of the radiation, three parameters characterizing the overall spectral fitted by Gauss function and two parameters corresponding to emission peaks were extracted respectively and then used as inputs of support vector machines. Oxygen consumption, oxygen gun vibration amplitude, oxygen gun vibration time and feeding quantity of the process of production were chosen to construct subsample, the membership factors were calculated and the prediction model was built by using fuzzy support vector machine. The experimental results show that the proposed method has better recognition accuracy than the manual method and the traditional SVM method, and can provide reference for converter operator to determine the end point accurately.
2018, 47(7): 726005.
doi: 10.3788/IRLA201847.0726005
It is difficult and important to de-noise nonstationary signal. To this end, a new noise attenuation method for nonstationary signal was proposed based on sparse representation and Particle Swarm Optimization(PSO). A redundant dictionary which is insensitive to useful signal was developed for the representation of cultural noises. PSO was used to improve the search strategy of Matching Pursuit(MP). Simulated experiments and real MT data were used to test the proposed scheme. As a conclusion, not only charge-discharge-like noise can be effectively removed, spikes and some other irregular noise can also be well suppressed. The apparent resistivity and phase curves obtained after applying our scheme are greatly improved over previous.
