Office Online  
    Office Online
    Manuscript Tracking
    Editor Work
    Office Work
    Editor-in-chief
Journal Online
    Current Issue
    Advanced Search
    Archive
    Read Articles
    Download Articles
    Email Alert
Traffic
Quick Search  
  Adv Search
2021 Vol. 42, No. 3
Published: 2021-03-28
 
265 Research on High Accuracy Mass Measurement of Silicon Sphere in the Dissemination of Mass Unit
PENG Cheng,WANG Jian
DOI: 10.3969/j.issn.1000-1158.2021.03.01
Based on the differences on material and geometry properties among silicon spheres and other regular mass standards, an improved method to determine the air buoyancy correction in the silicon sphere mass measurement was proposed. Six types of air buoyancy fluctuations were concluded by studying the real-time data of air density during the mass measurement of silicon sphere. Since the environment conditions are changeable, it was observed that the air buoyancy fluctuations of 1×10-4kg/m3 were existed between any two adjacent weighing processes in mass measurement of silicon sphere. About 15.1～30.2μg errors in the air buoyancy correction can be produced by ignoring the air density fluctuation in the mass measurement of a 1kg silicon sphere (using a 1kg platinum-iridium weight as reference standard).
2021 Vol. 42 (3): 265-270 [Abstract] ( 346 ) HTML (1 KB)  PDF (684 KB)  ( 376 )
271 Research on High Accuracy Large Torque Standard Machine Based on Air Bearing Technology
ZHANG Zhi-min,ZHANG Yue,MENG Feng,JIANG Ji-le,HU Gang,ZHANG Wei
DOI: 10.3969/j.issn.1000-1158.2021.03.02
A set of high accuracy 20kN·m torque standard machine based on air bearing technology was newly developed at National Institute of Metrology (NIM). The static air bearing is adopted to support the arm lever at the fulcrum, a multi-part frame structure is used in the arm system, the arm is made of Invar alloy material with a low thermal expansion coefficient. A combination of multiple weight groups, rotating platform and lifting mechanisms is adopted in the weight loading system which can realize the measurement of torque values in different ranges without reverse loading. The working principle and mechanical structure of the machine are introduced, the results of performance test and uncertainty assessment are described. The theoretical analysis and performance tests show that the repeatability is smaller than 2×10-5, the uncertainty is less than 2×10-5(k=2) in the range from 100N·m to 24kN·m.
2021 Vol. 42 (3): 271-275 [Abstract] ( 321 ) HTML (1 KB)  PDF (1134 KB)  ( 467 )
276 Research of Piston Gauge Based on Two-way Fluid-structure Interaction Technique
ZHANG Zhong-li,WANG Can,LIN Zheng-hao,Ni Yu-shan,HONG Bian,QIN Ting-ting,LI Xue-jing
DOI: 10.3969/j.issn.1000-1158.2021.03.03
According to the deficiency that the parameters cant be singly controlled in the traditional experiment, a 3D rotating model of 1000MPa simple piston gauge has been established to probe the effects of fluid viscosity, piston rod modulus and piston-cylinder gap on the pressure reproduction of piston gauge based on two-way fluid-structure interaction technique. The results show that the pressure distribution along the piston-cylinder gap from the inlet to the outlet generally experiences first approximate-linear decrease of pressure with the quick decline down to 30% of the balance pressure, then through a small pressure fluctuation, and finally approximate-linear slow decrease of pressure till to zero. The piston rod shrinkage decreases due to the increase of the elastic modulus of piston rod, the decrease of the fluid viscosity and piston-cylinder gap, while the effective area of the piston gauge increases due to the increase of the elastic modulus, the fluid viscosity, and the decrease of the piston-cylinder gap. Moreover, the effective area increases approximately 1.3% in maximum when the fluid viscosity doubles, which affects on the effective area of piston gauge most, while the effective area increases less than 0.04% in maximum when the elastic modulus of piston rod increases 10 times, which affects on the effective area least.
2021 Vol. 42 (3): 276-281 [Abstract] ( 275 ) HTML (1 KB)  PDF (2852 KB)  ( 530 )
282 Vibration Calibration Method Based on the Decoding of Heterodyne Interferometer Output S/PDIF Digital Signal
WANG Zhao,CAI Chen-guang,YANG Ming,KONG Ming,GUO Tian-tai,WANG Dao-dang
DOI: 10.3969/j.issn.1000-1158.2021.03.04
An accelerometer calibration method based on Polytec heterodyne interferometer output S/PDIF digital signal decoding is proposed, which optimizes the calibration process and guarantees the sensitivity magnitude and phase calibration accuracy of the accelerometer. Comparing with the calibration results of the heterodyne interferometry recommended by ISO 16063-11 in the frequency range from 5Hz~20kHz, the relative error of sensitivity magnitude and difference of sensitivity phase by the proposed method are less than 0.969% and 0.165° respectively; the relative standard deviation of sensitivity magnitude and the standard deviation of sensitivity phase are less than 0.560% and 0.356° respectively.
2021 Vol. 42 (3): 282-286 [Abstract] ( 331 ) HTML (1 KB)  PDF (1192 KB)  ( 643 )
287 Analysis and Elimination of Half-wavelength Error in Homodyne Laser Interference Signal Processing System Based on CORDIC Algorithm
FAN Hong-wei,DIAO Xiao-fei,ZHANG Fu-min,XUE Zi,DONG Xin-yuan,TIE Mi-mi
DOI: 10.3969/j.issn.1000-1158.2021.03.05
In order to meet the nano-scale resolution and high-speed processing, a laser interference signal processing system based on CORDIC algorithm was designed. A gross error of half-wavelength in the signal processing system was founded during the research process. The characteristics and formation mechanism of the gross error were studied, and the research showed that the root cause of the half-wavelength error was the existence of a certain error interval in phase measurement, thus, near the zero-point of interferometric signal phase, the error interval caused the decimal phase to be in ［2π-,0+］, which was uncertain. While the decimal phase combining with integer phase, the half-wavelength error might occur. On the basis of in-depth research of the gross error, a CORDIC algorithm phase compensation technique based on the correction table was proposed to eliminate the influence of the half-wavelength error on the laser interferometry system. The experimental results showed that the phase compensation technique could effectively eliminate the half-wavelength gross error in the signal processing system, and the total quantization error introduced by the CORDIC algorithm was less than ±0.05nm.
2021 Vol. 42 (3): 287-293 [Abstract] ( 352 ) HTML (1 KB)  PDF (1460 KB)  ( 570 )
294 NURBS Surface Fitting Based on Fractional Fourier Transform
KONG De-ming,HUANG Zi-shuang,WANG Shu-tao,SHI Hui-chao
DOI: 10.3969/j.issn.1000-1158.2021.03.06
In order to realize high quality non-uniform rational B-splines (NURBS) fitting surface of free-form surface model, a NURBS surface fitting method based on fractional Fourier transform is proposed. Firstly, the elevation image of the point cloud data of the free-form surface model is analyzed by means of fractional Fourier transform, and the feature points representing the three-dimensional structure of the free-form surface are extracted from the elevation image of the point cloud data of the free-form surface model. Then, the data points for NURBS surface fitting are selected by using the outer tangent circle extraction method combined with the extracted feature points. Finally, the fractional Fourier transform filtering and inverse interpolation node method are used to optimize the shape of the fitting surface to improve the fitting accuracy. The experimental results show that compared with the traditional NURBS fitting method, the root-mean-square error of the fitting results is reduced by 28% under different adjustment times, and the fitting effect is better.
2021 Vol. 42 (3): 294-300 [Abstract] ( 278 ) HTML (1 KB)  PDF (2665 KB)  ( 741 )
301 Research on Lens Automatic Detection System
HE Rong,JIN Tao
DOI: 10.3969/j.issn.1000-1158.2021.03.07
Aiming at the problems of lens eccentricity due to manual adjustment during lens detection, a technical solution for measuring the lens by combining Hartmann detection method, machine vision technology and stepper motor control system is proposed. Based on this scheme, an automatic lens detection system is designed and manufactured. By analyzing the real-time size and position of multiple light spots, the host computer communicates with the microcontroller to control the movement of the mechanical module, thereby automatically calibrating the lens position and measuring the lens. -15m-1~+15m-1 spherical lens is used for repeatable measurement. The experimental results show that the system achieves full-automatic lens measurement, and the test results meet the requirements of national metrological standards. Measurement error is less 0.02m-1 than manual measurement results, the accuracy is improvemented.
2021 Vol. 42 (3): 301-307 [Abstract] ( 325 ) HTML (1 KB)  PDF (3388 KB)  ( 381 )
308 Agriculture Image Dehazing Based on Mean Clustering and Loop Guided Filtering
TIAN Yuan
DOI: 10.3969/j.issn.1000-1158.2021.03.08
Image dehazing is researched based on the mean clustering and loop guided filtering(MCLGF). Firstly, image dehazing optical model was established. Secondly, two sub feature vectors were selected randomly as initial cluster center, image was divided into the sky and non sky area from the classification results based on mean clustering, atmospheric light value was obtained from the sky area pixel. Thirdly, loop guided filtering was achieved the details of smoothing and edge orientation, and difference output and input image of guided filtering was minimize with minimization cost function. Finally, the algorithm process was given. Simulation results show mean clustering and loop guided filtering could reduce the haze of image effectively, and structural similarity is 0.98, information fidelity is 0.96, information entropy is 9.12 and regular gradient mean is 0.85, the four evaluation indexes are better than the other algorithms.
2021 Vol. 42 (3): 308-313 [Abstract] ( 293 ) HTML (1 KB)  PDF (4069 KB)  ( 213 )
314 Development of Large-aperture and High-emissivity Surface Blackbody Radiation Source
HU You-hua,HAO Xiao-peng,SIMA Rui-heng,XIE Chen-yu,SONG Jian,LIU Yang,WU Qiang,YANG Yan-long
DOI: 10.3969/j.issn.1000-1158.2021.03.09
As a primary standard, the black body radiation source plays an important role in the calibration of infrared measurement equipment. In order to response the needs for the radiometric calibration of large aperture infrared measurement equipment, a surface blackbody source whose area is 400mm×400mm has been designed. The temperature control of the blackbody is carried out through the multi-channel temperature control method and the method of connecting a fixed cold source. The appropriate heat transfer model has been determined by thermal simulation. Combining high emissivity coating process and structural design of the radiating surface makes the black body have high emissivity, the effective emissivity of the blackbodys radiating surface is 0.992. The temperature maximum uniformity deviation of blackbody radiation source measured by standard platinum resistance thermometers is 0.101K and the average stability is about 0.018K/10mins, it can satisfy the requirement of blackbody radiation sources for large aperture infrared measurement equipment.
2021 Vol. 42 (3): 314-320 [Abstract] ( 257 ) HTML (1 KB)  PDF (3132 KB)  ( 578 )
321 Study on the Realization of Aluminum Freezing Point in the Three-zone Temperature Controlled Fixed Point Furnace
HE Pei,LIN Lin,YAN Xiao-ke,ZHANG Ming-yu,WANG Ning
DOI: 10.3969/j.issn.1000-1158.2021.03.10
The vertical temperature uniformity of the fixed-point furnace is an important factor affecting the plateau quality of the fixed-points of the International Temperature Scale of 1990 (ITS-90). In order to improve the reproducibility of the aluminum freezing point, a three-zone temperature-controlled fixed-point furnace was designed, and the aluminum fixed point cell with a metal shell was used to study the vertical temperature uniformity and its influencing factors. On this basis, the continuous heat flux method was adopted to precisely realize the aluminum freezing point. The experimental results show that the temperature uniformity can be improved by adjusting the settings of the top, middle, and bottom furnace temperatures, and the temperature stability of the freezing plateau within 4 hours is less than 1mK. The phase transition supercooling of the high purity aluminum is approximately 0.14℃ under the experimental conditions.
2021 Vol. 42 (3): 321-326 [Abstract] ( 296 ) HTML (1 KB)  PDF (2470 KB)  ( 289 )
327 Multi-Factor Coupling Effect of Subsonic Nozzle on Gas Flow Distribution
ZHANG Han,JIA Li,CUI Li-shui,LI Chun-hui
DOI: 10.3969/j.issn.1000-1158.2021.03.11
In order to study the interaction of gas flow field evolution in subsonic nozzle, the coupling effect of pressure-velocity-contraction ratio on the velocity profile was quantitatively evaluated. Based on the standard k-ε turbulence model, the numerical simulation was utilized for the uniform gas flow field in the subsonic nozzle under coupling effect and the key parameters of top-hat flow field distribution were defined. The results show that within multi-factor coupling, the subsonic nozzle can change the quality of the flow field distribution by influencing the key points of the strengthening process including the equilibrium point and the critical position point of the pressure gradient. Increasing the gas pressure can effectively reduce the boundary layer thickness in the tectonic flow field, reduce the dependence of the key points on the flow velocity in the process, and improve the stability of the boundary layer thickness under different flow velocities. The increase of contraction ratio leads to the further deviation of the flow field from the ideal distribution.
2021 Vol. 42 (3): 327-333 [Abstract] ( 266 ) HTML (1 KB)  PDF (5276 KB)  ( 215 )
334 Identification of Two-phase Flow Based on MO-PLP-ELM and Electrical Capacitance Tomography
ZHANG Li-feng,ZHU Yan-feng
DOI: 10.3969/j.issn.1000-1158.2021.03.12
Identification of two-phase flow based on multi-objective optimized parallel layer perceptrons extreme learning machine (MO-PLP-ELM) and electrical capacitance tomography (ECT) is proposed. Firstly, the random training method is used to generate the training and testing sets for the studied seven two-phase flow regimes, which assures the representativeness of the samples. Secondly, the capacitance data of the sample are normalized. Finally, the MO-PLP-ELM algorithm is used for flow regime identification, and the results are compared with those of BP neural network, support vector machine, extreme learning machine algorithms and extreme learning machine with parallel layer perceptrons. The results show that the average recognition rate can reach 96.1% using MO-PLP-ELM, which is obviously higher than other algorithms.
2021 Vol. 42 (3): 334-338 [Abstract] ( 301 ) HTML (1 KB)  PDF (704 KB)  ( 218 )
339 Capacity Improvement of Natural Gas Flow Standard Device
YANG Yan-ping,LIU Bo-tao,XU Ming
DOI: 10.3969/j.issn.1000-1158.2021.03.13
Introduced the composition,measurement principle,traceability method of measurement value and key measures to improve the calibration capacity of the working flow standard device of Wuhan Branch of National Station of Petroleum & Natural Gas flow Measurement,and evaluated its uncertainty. The uncertainty of absolute pressure and differential pressure measurement, the uncertainty of experimental evaluation of compression factor ratio and the uncertainty of standard turbine flowmeter correction are analyzed. The uncertainty of the standard device is 0.13%~0.16%(k=2),which can meet the verification requirements of the natural gas flowmeter of level 0.5 and below.
2021 Vol. 42 (3): 339-345 [Abstract] ( 278 ) HTML (1 KB)  PDF (613 KB)  ( 508 )
346 Analysis of Influencing Factors on Measurement Accuracy of Particle Image Velocimetry Technology
LIU Bing,CUI Li-shui,LI Xiao-ting,WANG Yi
DOI: 10.3969/j.issn.1000-1158.2021.03.14
The instantaneous full flow field measurement method for particle image velocimetry(PIV) is reviewed.The working principle of PIV, the key link of the technology(the quality of the obtained image), and the main factors influencing the measurement accuracy (the performance of tracer particles, light source performance,background noise, imaging distortion, time resolution of synchronizer and image processing algorithm) are systematically introduced.Based on the above analysis, specific suggestions are proposed on the guarantee of the performance of tracer particles, the selection of imaging components and the determination of image processing method.
2021 Vol. 42 (3): 346-351 [Abstract] ( 420 ) HTML (1 KB)  PDF (560 KB)  ( 1044 )
352 Research on Torque Ripple Suppression of Switched Reluctance Motor Based on Peak-Valley Complementary Method
HE Wei-fu,WANG Ning-li,ZHAO Hao,FENG Hao
DOI: 10.3969/j.issn.1000-1158.2021.03.15
Obvious torque pulsation of switched reluctance motor is produced for its nonlinear electromagnetic characteristics and pulse working mode. In order to solve this problem, a method based on the principle of peak-valley complementary superposition is proposed, which through analyzing the switched reluctance motor torque ripple reason theoretically and measuring the torque ripple waveform. In this method, two sets of motor systems with the same stator rotor parameters are used, to generate torque ripple waveform with 180°phase difference, and the same amplitude approximately. Finally, the experimental results show that the pulsating torque measured by the above method can be reduced by 50%~60% compared with the single switched reluctance motor.
2021 Vol. 42 (3): 352-357 [Abstract] ( 225 ) HTML (1 KB)  PDF (2892 KB)  ( 238 )
358 The Sinewave Fit Algorithm Based on Total Least-Square Method with Partial Period Waveforms and Non-uniform Sampling
LIANG Zhi-guo
DOI: 10.3969/j.issn.1000-1158.2021.03.16
To introduce a four-parameter sine wave curve-fit method, it can attain the curve-fit results only with the partial wave period. it based on the three-parameter sine wave curve-fit method, and it aims at the parameter estimation of ultra lower frequency sinusoidal with non-uniform sampling series, where, normally one cant get the whole period waveform due to time limit, and there are only partial sine wave period. The speciality of the arithmetic is that it turns the optimization of four parameters (amplitude, frequency, phase and offset) into the optimization of one parameter (only frequency), and without any original parameter pre-estimation. Both the simulation and experiments have proved the validity and feasibility, this method can be applied to the four-parameter sine wave curve-fit and the ultra low frequency sine wave parameter control.
2021 Vol. 42 (3): 358-364 [Abstract] ( 266 ) HTML (1 KB)  PDF (1794 KB)  ( 421 )
365 Investigation into Definition Method for On-wafer 16-term Error Model Calibration Kits Below 110GHz
WANG Yi-bang,ZHOU Rui,CHEN Ting,WU Ai-hua,LIU Chen,LIANG Fa-guo
DOI: 10.3969/j.issn.1000-1158.2021.03.17
The conventional lumped parameter circuit model has been widely used in the definition of calibration kits for on-wafer probing. However, the accuracy of the defined model parameter is not great, due to the non-ideal circuit model and fitting algorithm used in extracting circuit parameters. A more accurate definition method is presented for 16-term calibration kits that is designed to correct the crosstalk at high frequencies. The new definition method is based on measurements of fabricated assistant multiline TRL calibration kits and full-wave simulations. The measurement system corrected by such defined 16-term calibration kits was compared with two-tier calibration method of Multiline TRL proposed by NIST. The results shows excellent agreement between these two techniques, i.e. the difference in transmission magnitude is within 0.30dB and the phase is within 1.0°. The proposed definition method offers appealing advantages over the NIST technique, in terms of smaller number of standards required and fewer movements of probes.
2021 Vol. 42 (3): 365-369 [Abstract] ( 265 ) HTML (1 KB)  PDF (1923 KB)  ( 333 )
370 Research on Lead Equivalent Measurement of Narrow X-ray Beam Shielding Material
HAN Lu,ZHAO Rui,WU Jin-jie,LU Chun-hai,WEN Yu-qin,YAN Yong-qiang
DOI: 10.3969/j.issn.1000-1158.2021.03.18
The shielding performance of radiation protection materials is usually expressed in lead equivalent. In order to measure the shielding ability of different protective materials accurately, according to the IEC 61331 standard, the X-ray radiation quality for lead equivalent measurement has been researched and established. The radiation quality spectrum was simulated by using EGSnrc software to study the influence of different thicknesses of lead on the radiation quality spectrum. The narrow beam condition measurement method of X-ray lead equivalent has been discussed, and the lead equivalent of 16 different protective materials under narrow beam condition was tested. The results showed that the established radiation quality meets the standard requirements. Under the condition of 120kV tube voltage, the lead equivalent measurement result of the protective material is equal to or greater than the nominal lead equivalent value, indicating that the shielding effect of the tested material meets the requirements.
2021 Vol. 42 (3): 370-374 [Abstract] ( 320 ) HTML (1 KB)  PDF (1039 KB)  ( 497 )
375 Research on the Detection Method of Key Parameters of Oral and Maxillofacial CBCT
HUANG Jin-feng,CHEN Xiao-zhao,WANG Hong
DOI: 10.3969/j.issn.1000-1158.2021.03.19
The imaging principle of CBCT is different from the traditional CT. The detection method of the original spiral CT is not suitable for the oral CBCT. For the characteristics of oral CBCT, the detection methods of key parameters are studied, including dose area product, spatial resolution, contrast-to-noise ratio index, uniformity. The advantages of the new method compared with the traditional method are expounded, and the corresponding analytical calculation methods and the actual test results are given. Except that the dose area product method is not suitable for Newton VG equipment, the test results of other equipments meet the following requirements: The corresponding measurement performance indicators of these parameters should be satisfied: the normalized dose area product should be no more than 250mGy·cm2; the measured value at 10% of MTF should be no less than 1LP/mm; CNI should be less than 20%; the uniformity should be greater than 5.
2021 Vol. 42 (3): 375-379 [Abstract] ( 280 ) HTML (1 KB)  PDF (639 KB)  ( 344 )
380 Granularity of Terrigenous Clastic Rock Using Sieving Method and Laser Particle Size
YOU Fan,GENG Xiang
DOI: 10.3969/j.issn.1000-1158.2021.03.20
The particle size of the terrigenous clastic rocks is determined by the combination of the sieving method and the laser particle size method, in order to ensure the concerted synergy between the sieve method and the laser particle size method, the analytical data of the laser particle size method is corrected, comparing the probability cumulative frequency curve of sediment rocks before correction and after correction in different geological ages, it is found that the corrected curve can make better characterize the particle size distribution of terrigenous clastic rock. The L9 (34) orthogonal optimization scheme is carried out for the three factorsof the sampler stirring speed (A), pump speed (B) and ultrasonic intensity (C) which influenced the background measurement of the Malvern MS 2000 laser particle size analyzer, The optimal measurement conditions are: A is 600r/min, B is 2000r/min, which makes the background measurement stable, and concluding that the influence degree on the background measurement can be expressed by B>A>error>C. The particle size standard curve is calibrated with 9 reference materials, the residual error between the standard curve of the particle size and the standard curve of instrument was small, ranging from 0.795% to 2.934%, the relative standard deviation is between 0.24% and 0.33%. This method is suitable for the particle size analysis of terrestrial clastic rocks in various geological ages.
2021 Vol. 42 (3): 380-387 [Abstract] ( 265 ) HTML (1 KB)  PDF (2436 KB)  ( 406 )
388 Research on Construction and Method of Particle Size Recognition and Detection Device for Light Scattering Particles Monitor
WU Dan,ZHANG Guo-cheng,ZHAO Xiao-ning
DOI: 10.3969/j.issn.1000-1158.2021.03.21
A particle size identification device for light scattering particles monitor was designed and built. The particle size identification error of a PM2.5 sensor with particle size recognition function was tested by using SHIBATA LD-5 light scattering particles monitor as a reference (TSP) with different size particles (0.6, 2.0, 2.3, 2.5, 2.7, 3.0, 3.2, 3.4, 3.7μm). If the particle size identification function of the sensor is accurate, the concentration value TSP=PM10>PM2.5=0 when the particle size is larger than 2.5μm. However, the results show that PM2.5/PM10 decreases significantly until the particle size is larger than 3μm, where the particle size identification is realized. However, the value of PM2.5/PM10 risesafter 3.7μm.
2021 Vol. 42 (3): 388-394 [Abstract] ( 315 ) HTML (1 KB)  PDF (3297 KB)  ( 1048 )
395 Denoising Method of EEG Signal Based on EEMD-ICA
FAN Feng-jie,BAI Yang,JI Hui-fang
DOI: 10.3969/j.issn.1000-1158.2021.03.22
The electroencephalograph (EEG) signal is the electro physiological activity of brain cells, it is reflected in the scalp surface. However it is usually interfered by noises during signal acquisition process. In order to reserve the effective information and eliminate as much noise as possible, a method of ensemble empirical mode decomposition (EEMD) combined with independent component analysis (ICA) is introduced. Firstly, the EEMD decomposition can get a certain number of intrinsic mode function (IMF) of EEG signals. The virtual channel is reconstructed by the IMF components with more noise components which are selected based on correlation coefficient and de-noised by ICA algorithm. Secondly, the denoised results and the IMF components with multiple signal are reconstructed. Finally, the reconstructed signal is denoised by ICA again, and the final denoised signal is obtained. The experimental results show that the mentioned method has better signal-to-noise ratio and smaller RMSE than the other denoising methods, including wavelet denoising, EEMD denoising and ICA denoising. It shows that the mentioned method can denoise better.
2021 Vol. 42 (3): 395-400 [Abstract] ( 450 ) HTML (1 KB)  PDF (2504 KB)  ( 538 )
News
Download
Links
Copyright © Editorial Board of Acta Metrologica Sinica
Supported by:Beijing Magtech