With the continuous progresses of microwave technology, it is urgent to improve the frequency stability and phase noise index of microwave signal source. Optoelectronic oscillator has attracted wide attention because of its extremely low phase noise and good frequency expansibility. However, the sideband of the output signal is greatly increased by the long fiber inside the optoelectronic oscillator. The method to solve the contradiction between single-mode output and phase noise has become a hot topic in the field of optoelectronic oscillator. Parity-time-symmetry method can greatly optimize the sideband suppression of the output signal on the premise of ensuring the phase noise of the output signal. Based on the description and analysis of the working principle and characteristics of parity-time-symmetric optoelectronic oscillator, the structure and schemes of parity-time-symmetric optoelectronic oscillator designed by domestic and foreign research teams are summarized, and the basic principles and performance indicators of various schemes are analyzed and introduced. And the future development directions of parity-time-symmetric optoelectronic oscillator technology are discussed.

As a typical representative of the third generation semiconductor materials, SiC has great application prospects in the field of power devices due to its excellent physical and chemical properties. In this paper, the basic principles of laser drilling are briefly introduced. The research progresses of long pulse and ultra-short pulse laser drilling of SiC are summarized. The processing characteristics of long pulse and ultra-short pulse laser are compared, and the drilling effects of SiC under different pulse widths are analyzed. At the same time, the research status of laser drilling of SiC power devices, such as GaN/SiC and AlGaN/GaN/SiC are introduced. Finally, the challenges of laser drilling of SiC and its power devices are pointed out, and the future development directions are prospected.

The nearly infrared enhanced PIN photodetector with high absorption efficiency black silicon micro-structure was fabricated by wet etching process based on nitric acid/hydrofluoric acid/phosphoric acid/sulfuric acid mixture. And the parameters of the fabricated photodetector were tested and compared with that of the PIN photodetector without black silicon. Test results show that the quantum efficiency of the black silicon photodetector reaches 80.7% and its response rate reaches 0.69A/W at 1060nm, which is 116% higher than that of the non-integrated black silicon devices. The dark current of the black silicon detector is less than 8nA, the response time is less than 8ns, and the capacitance is less than 9pF, which is equivalent to that of the non-integrated black silicon devices. Due to the process compatibility, the black silicon technology has the potential to be widely used in silicon-based near-infrared PIN, APD, SPAD, SPM and other photodetectors, so as to significantly improve the response rate, quantum efficiency, response speed, temperature coefficient of breakdown voltage and other performance.

Using GaN-based micron-sized LEDs (Micro-LEDs) as the light source for visible light communication (VLC) systems has become a hot research topic in recent years. In this paper, Micro-LED devices with different sizes from 40~10μm have been designed and fabricated by deep UV lithography. The optoelectronic characteristics of GaN-based Micro-LEDs as well as modulation bandwidth performance were systematically investigated. It is found that the modulation bandwidth can be significantly increased by decreasing the device size. By optimizing sidewall passivation and electromagnetic shielding, high brightness as much as 1×108cd/m2 has been acquired by the green Micro-LED chip with 10μm in diameter. Furthermore, the modulation bandwidth of a single blue Micro-LED of 20μm can reach 372.6MHz. The results show that higher modulation bandwidth can be expected for the GaN-based Micro-LED chip, and it is expected to promote the application of high-speed visible light communication system in the near future.

A hybrid mode-locked fiber laser based on nonlinear amplifying loop mirror and molybdenum diselenide saturable absorber is reported in this paper. By optimizing the pump power and polarization state, five types of different mode-locked states, such as Q-switched mode-locked pulse, basic mode-locked pulse, 2~5 multi-pulse, pulse beam, and 2~4 harmonic of pulse beam have been obtained in the experiment. Compared with the single mode-locked pattern, the hybrid mode locking pattern can not only realize the self-starting mode locking, but also enrich the pulse dynamic state.

Combining optical and genetic methods, optogenetics can precisely control the activity of specific neurons, which provides a powerful means for neuroscience research. The optrode can introduce light into animals and record the activity of neurons under the control of light through the electrode. In order to reduce its volume and improve its function, a kind of implant optrode with the cross-section size less than 0.1mm2 was designed and fabricated based on high-density integrated silicon microelectrode and bare optical fiber. Compared with the traditional single-channel electrode, the designed optrode is composed of two optical channel and 32 recording sites. The two channels can configure different excitation wavelengths more flexibly to activate or inhibit neurons at different sites at the same time. Compared with the traditional wire electrode, the 32-channel silicon electrode has higher integration and higher spatial resolution to record the activity of neurons before and after stimulation.

Ridge-waveguide superluminescent diodes (SLD) with compressive strain multi-quantum-well were designed and fabricated. The TE mode reflectivity of the TiO2/SiO2 four-layer broadband antireflection film is about 10-6, and the effects of the deviation of the angle of the ridged waveguide and the thickness of the film on the reflectivity of the antireflection film are analyzed. The experimental results show that the single-tube output power of the designed SLD chip can reach 22.7mW, the output spectrum FWHM is about 37.3nm, the spectral ripple coefficient is less than 0.15dB, and the TE mode output intensity is dominant with a polarization degree of about 19.2dB under 250mA DC current.

The influence of the factors such as the LED chip size, the emission half Angle and the distance from the luminous surface to the LCD panel layer in Mini LED backlight modules, on the illumination uniformity of the LCD crystal panel was studied and analyzed with Taguchi experimental method and Tracepro software. The results show that the distance from the luminous surface to the liquid crystal layer has the greatest influence on the illuminance uniformity, and the weight of the influencing factors is 82.88%, and that of the luminous half angle and the chip size is both 14.33%. According to the research results, the optimal illumination uniformity is obtained by fixing the luminescence half angle and the chip size and adjusting the distance between the luminous surface and the LCD panel layer.

The ITO film with the thickness of 35nm was plated on the glass substrate, and the linear permittivity of the ITO film was measured via spectroscopic ellipsometry. Owing to the large doping concentration of ITO, its permittivity can be quantified by Drude model, and the epsilon-near-zero wavelength is obtained as 1100nm. The varying rules of electronic temperature and lattice temperature with time can be calculated through the double-temperature model. The change of plasma frequency can be calculated with the increasing electronic temperature, and then it is taken into the Drude model to get a new permittivity. Finally, by calculating the variation of the refractive index, the nonlinear refractive index n2 can be obtained. The calculated results show that the maximum nonlinear refractive index of n2=4.66×10-15m2/W can be obtained at epsilon near zero wavelength. It is expected to apply in the nano photonic devices such as optical storage, all-optical switches and so on.

The gate of thin film transistor (TFT) in liquid crystal display is fabricated by photolithography process. It is found that in addition to the exposure and development process, the development inspection critical dimension (DICD) and the profile angle (Taper) are also affected by the vacuum drying parameters. In this paper, the influence of vacuum drying process on DICD and Taper of photoresist (PR) was studied by a full factor experiment, in which the slow pumping time, holding time and bottom pressure were taken as independent variables and DICD and Taper were set as the dependent variables. The results show that: in the vacuum drying process, slow pumping time and bottom pressure has little effect on DICD and Taper, and holding time is the key parameter. With the increase of holding time, DICD increases but Taper decreases. This can be explained as follows: with the increase of holding time, the total amount of solvent volatilization in the PR increases, the PR becomes denser, then the development speed decreases, so the DICD enlarges; meanwhile the amount of solvent volatilization at the top of the PR increases, and the concentration of photosensitizer at the top increases, resulting in the increase of lateral development at the top, and finally the Taper reduces. Furthermore, the regression equation of the DICD and Taper with the holding time was established, through which the DICD and Taper could be predicted or the required holding time corresponding to the expected lithography effect could be deduced. This work provides a reference for the parameter optimization and product yield improvement of TFT lithography production line.

The solar cells and modules were fabricated with gallium doped and boron-doped monocrystalline silicon wafers respectively. Light-induced degradation (LID) treatment and soldering without welding rod were conducted on the two kinds of solar cells. The photovoltaic performance changes of the solar cell and modules induced by LID and soldering without welding rod were tested with Halm I-V tester. Test results indicate that, the LID ratio of the Ga-doped silicon solar cells was 0.91% lower than that of the B-doped samples. Moreover, after the soldering without welding rod, the consistency of the optoelectronic performance of the Ga-doped silicon wafer has little change, and its cell-to-module (CTM) ratio is higher than that of the B-doped monocrystalline silicon solar cell module. In conclusion, Ga-doped monocrystalline silicon solar cells can better suppress the photo-attenuation effect, reduce the influence of the series welding process on the photoelectric performance of solar cells, and obtain higher power of solar cell component.

The polarization potential generated by thermal variation can change the mechanical and electrical physical quantities in the piezoelectric semiconductor structure, which is of valuable application in artificial intelligence and MEMS. By utilizing a developed two-dimensional model together with the accurate thermoelectric physical boundary conditions, it is systematically investigated the temperature gradient-dependent physical fields such as polarization, electric potential, electric field, carrier distribution and current in a GaN piezoelectric pn junction. It is found that, due to the coupling between the thermal-gradient fields and polarization charges, the electromechanical field of a piezoelectric pn junction has a quick response to thermal-gradient. Furthermore, gate voltage and carrier transport characteristics can be effectively tuned with thermal-induced and piezoelectric charges. This may provide an alternative approach and theoretical guidance to manipulate the carrier transport in intelligent heterojunction devices.

Aiming at the demand of the backside leads of MEMS devices, a process for the fabrication of 10.16cm (4inch) wafer substrates based on the through-glass-vias (TGV) processing method is presented. Firstly, the conductive silicon wafer was deeply etched, then anodic bonding was performed between the silicon wafer and the glass wafer, and then the bonded glass-silicon wafer was heated at a high temperature to fill the glass into the silicon wafer. Then the front glass and back silicon of the glass-silicon wafer was grinded and polished until the silicon and the embedded glass were on the same plane. Finally, a 4inch substrate with a thickness of 258μm was obtained. The arithmetic mean deviation of the profile, the maximum profile height and the microscopic unevenness are 13, 71 and 49nm respectively. In addition, the measured resistivity of the silicon via in the wafer is 0.023Ω·cm.

For the heat dissipation problem with uniform background heat flow, a leaf-like microchannel rectangular heat sink model was constructed. Based on the constructal theory, under the constraints of a given heat sink volume and the total volume of the liquid cooling channels, with the minimizations of maximum temperature and pressure loss of the heat sink as the objectives, the optimal designs are carried out with the element number of microchannels, the angle between the main channel to the branch channel, and the tube diameter ratio of the main channel to the branch channel as the design variables. The results show that, by increasing the element number of microchannels, reducing the angle between the main channel and the branch channel, and adopting a smaller tube diameter ratio of the main channel to the branch channel, the maximum temperature can be reduced, but the pressure loss increases.

WO3 nanoplate films with highly active (002) facets were prepared on the FTO substrate with one-step hydrothermal method by using sodium tungstate dihydrate (Na2WO4·2H2O) as the raw material and oxalic acid (H2C2O4) as the structure-directing agent. The phase and morphology of the samples were characterized by XRD and FESEM. The energy band structure and carrier separation ability of the film were characterized by UV-Vis and PL. The photoelectric properties of WO3 nanoplate films were characterized by electrochemical workstation. The effects of oxalic acid dosages on the crystal orientation, morphology, size and photocatalytic performance of WO3 nanoplate films were investigated. The results show that the WO3 nanoplate films with 0.3g oxalic acid dosages present the highest diffraction peak on the (002) facets and possess the best photocatalytic performance.

An ultra-wideband omni-directional metamaterial absorber based on dome-cone-shell array was proposed. The influences of the geometric parameters of dome-cone-shell, the shape of the top and the different packing modes on the wideband absorption were studied. The ultra-wideband absorption characteristics of a metal-coated dome-cone shell array in the solar radiation spectrum are studied. The results show that the average absorption rate in the range of 250~2500nm reaches 95.39%. The average absorptivity for TE and TM polarization is 92.7% and 94.59%, respectively, at a large incidence angle from 0° to 65°, indicating that the proposed structure is polarization independent and incident angle insensitive. The weighted average absorption rate of air mass 1.5 is 95.24%, proving that it can effectively absorb solar radiation on earth. The proposed absorber was successfully prepared by combining colloidal ball lithography and surface coating. Primary test results show that the absorption performance still needs to be improved.

The band structures and electronic density of states of alkali metal doped SnO2 systems were calculated by using the first-principles theory. The results show that with the increase of the energy levels of the doped SnO2 system, the band gap can be adjusted well. The energy levels of the valence band top cross the Fermi line in the Li, Na, K, Rb doped SnO2 system, and the system shows the characteristics of semiconductors. More energy levels were introduced to the Fermi line in Rb-doped SnO2 system, because the valence band top moves to the low energy region in Cs, Fr doped SnO2, the energy levels of the valence band top do not cross the Fermi line. The density of states of the doped system is hybrids near the Fermi line. Compared with the Cs-doped SnO2, the energy levels in the Fr-doped SnO2 disperse. The change of reflectivity of doped SnO2 is mainly reflected in the visible and ultraviolet region, and the absorption edge is red shifted, which plays an important role in the realization of SnO2 photocatalysis.

In order to solve the problems of large distortion of ultra-wide field of view (FOV) infrared image and obvious difference from human vision, in this paper, an ultra-wide FOV infrared image distortion correction algorithm is proposed based on accurate model and back-projection. The algorithm first describes the relationship between object point and image point in the ultra-wide FOV infrared camera imaging by using the accurate model. Then, aiming at the disadvantage of low sampling rate of infrared image pixels, the more accurate cubic convolution interpolation method is used to interpolate the image for completing the imaging information. Finally, according to the point coordinates of corrected image, combined with the correction model and the accurate model, the corresponding coordinates on the ultra-wide FOV infrared image point coordinates are calculated by backward projection. The pixel value of the nearest image point is used as the value of the corrected image point. The experimental results show that the proposed algorithm has clear boundary and no sawtooth effect, and the average restoration error of the straight line in the scene is less than 0.35 pixels, which indicates that the algorithm has excellent applicability for the distortion correction of the ultra-wide FOV infrared image.

Aiming at the problem that the temperature measurement accuracy of the online infrared thermometer is easily affected by the environmental temperature, which causes big measurement error, an infrared temperature measurement compensation method based on the environmental temperature is proposed to improve the measurement accuracy. First, based on the principle of infrared temperature measurement, an infrared temperature measurement experiment platform was built, and temperature measurement data under different ambient temperatures were obtained. Then, using the principle of least square method to fit the extracted mean value of temperature measurement error to the curve of error-ambient temperature, the infrared temperature measurement compensation model was obtained. Finally, experiments were carried out to verify. The results show that using the compensation model to compensate the infrared temperature measurement data, the maximum relative error between the infrared temperature measurement value and the true temperature value of the measured target is 0.29%. It shows that the compensation method effectively reduces the measurement error of the infrared temperature measurement system caused by environmental temperature changes, and improves the measurement accuracy.

An APD bias voltage module with voltage boost and automatic compensation for temperature drift was designed. The module includes a positive feedback oscillator, a high frequency full wave rectifier circuit and a temperature compensator. Positive feedback oscillator is used to achieve DC/AC conversion, and the transformer is used to amplify voltage. High-frequency full-wave rectifier circuit outputs DC high voltage through AC/DC conversion. The temperature compensator automatically compensates the temperature drift of APD by the simulate temperature sensor. The size of the module is 1.7cm×2.7cm, the maximum power consumption is 100mW, the output voltage is continuously adjustable in 0~372V, and the maximum ripple is not more than 0.004%. By adjusting the parameters of components in the module, the bias voltage and temperature compensation can be carried out for any type of APD, and the maximum deviation of bias voltage doed not exceed 0.005%.

In order to solve the problems of slow scanning speed of traditional dispersive hyperspectral imagers and the problems of information loss, signal distortion, and high reconstruction complexity in the coded aperture spectral imaging system, in this paper, a slit spectral imaging system based on a uniformly distributed array is designed. On the basis of the coded aperture spectral imaging system, a displacement motor is used to control the slit array code plate to perform field-of-view spectrum non-aliasing micro-scanning to realize the non-destructive detection of dynamic scenes. The simulation results show that the collection efficiency of this method is 17 times that of the single-slit system, and the systems spectral data reconstruction result is consistent with that of the single-slit system. The spatial structure similarity is 1.8 times and the spectral fidelity is 1.17 times of that of the 50% sampling rate coded aperture system. Finally, the system was verified by experiments.

Visual measurement is a promising approach for locating the space deployment mechanism due to simple structure, low cost, and noncontact. Visual measurement data needs to be converted to object space coordinate system in practice, therefore providing guidance for the control quantity of the positioning system. In this paper, a data processing method is proposed by using motion sample data space to construct coordinate system transformation, which eliminates the cumbersome operation of external parameter calibration in on-orbit implementation. After correction by the compensation equation, the positioning accuracy of the single-camera measurement system can reach 0.02mm and 0.003° under the measurement distance of 1.7m.

Aiming at the large-size shutter of the large-area-array optical telescope in the space environment, a high-fidelity life test program for the space environment is designed. A large-caliber vacuum tank is used as the platform, and the complex environmental factors of large shutter in space operation are fully considered, providing a very real on-orbit operating environment for it. Aiming at the life index of the shutter opening and closing 1 million times on orbit, a specific test plan was designed, and a high-fidelity space environment test platform was built, which used the principle of accelerated life test to make the shutter mechanism operate in a vacuum tank. The results show that the test program can provide a very real space on-orbit operation environment for the shutter, and can support it to complete 1 million opening and closing actions.

According to the relevant characteristics of point cloud acquisition of time-of-flight method (TOF), an improved registration algorithm suitable for TOF point cloud was proposed. Firstly, FPFH features were used to perform rough registration of point cloud. And in the fine registration stage, the angle feature sampling method of normal vector was used to reduce the number of points of the point cloud, while retaining the key information points of the point cloud. At the same time, KD tree and RANSAC method were introduced to improve the registration efficiency of ICP. Experimental results show that this method has good registration efficiency and accuracy, and has a wide range of application.

In order to extract lung precisely, aiming at difficulty in segmentation of lung caused by interfering factors such as image noise, blood vessels and bronchus, an algorithm based on logit adjustment in multi-class residual network was proposed. The algorithm divided the image area into three categories: lung, background and boundary, which improves the segmentation accuracy by expanding the difference between different types of images. Firstly, the image was divided into regions with fixed size, then, a residual network was then trained to extract the texture features for classification and tested to achieve coarse segmentation. Finally, refining segmentation was conducted on regions which were marked as boundary based on threshold method. The segmentation performance of the proposed model was tested and verified by using a public dataset. The recall rate, precision and intersection over union of the algorithm were obtained as 99.79%, 98.13% and 97.83%, respectively, and the overall segmentation performance was higher than that of U-Net, one of the most cutting-edge segmentation networks. According to the experimental results, the proposed algorithm provides a reference basis for subsequent clinical diagnosis of lung diseases.

For multi-source high speed parallel data real-time transmission problems in data association network, a multi-source data optical transmission system based on sampling clock recovery is presented in this paper, and its operating principle and design idea are introduced in detail. The system combines the high speed transceiver in Field Programmable Gate Array (FPGA) with the special digital phase-locked loop chip, and gives the specific scheme of sampling clock recovery and data flow control. Compared with the existing high speed parallel data transmission system, the one presented in this paper has software-defined access capability and supports more flexible dynamic range for the different sampling clock. What is more, through the design of simplified and reasonable frame structure, and the constraint relation between the data bit width and the sampling clock is deduced, the system can improve the effective transmission bandwidth obviously. The test results show that this system can work stably and reliably with an excellent real-time transmission effect, and have the high precision of clock recovery. Meanwhile the system extends the application field of serializer and deserializer technology, and it provides a high reference value for other transmission systems.

Aiming at the problems of poor real-time performance and low accuracy of the existing UV imagers in UV and visible image registration, a method of UV and visible image registration fusion based on convolutional neural network (CNN) and wavelet fusion (WF) is proposed and applied to high sensitive UV imager. Firstly, the parameter model of the collected image data is pre-trained by combining the rigid body transformation and convolution neural network, and the optimal spatial transformation parameters are found by self-mining image features to achieve accurate registration of UV image and visible image. Secondly, two-dimensional wavelet decomposition and reconstruction algorithm is used to realize the fusion of UV and visible images. Experimental results show that the proposed method has fast registration speed, high overlay accuracy and good stability.