Aiming at the insufficient stiffness of the bonnet polishing system of the six-degree-of-freedom tandem joint robot, which may cause vibration and further mid-spatial-frequency errors, used we the IRB 6700 robot as the research object, established the modal analysis model based on Ansys Workbench and combined experiment to analyze the dynamic characteristics of the robot bonnet polishing system in the working condition frequency range. The experimental and simulation results together show that the robot bonnet polishing system has at least 5 modes in the working condition frequency range, and the jitter amplitude at the end of the robot is millimeter-level when the resonance occurs. Robot processing is severely restricted. In addition, for the application of advanced optical component polishing technology in the robotic bonnet polishing system, a vibration suppression bonnet tool was designed, and the fixed-point polishing and whole-surface polishing comparison experiments were carried out with the ordinary bonnet tool. The results show that the RMS and spectral amplitude of the fixed-point polishing spot of the vibration suppression bonnet are generally lower than that of the ordinary bonnet, and the introduced mid-spatial-frequency errors PSD is 40% lower than that of general bonnet polishing.

The single-pixel imaging system is a computational optical imaging technology that obtains the two-dimensional distribution information of the target through a single-pixel detector without spatial resolution. Compared with traditional direct imaging technology, it has a series of advantages such as high energy collection efficiency and high sensitivity. In the field of high-energy physical diagnosis technology it has broad application prospects. Aiming at the problem that the actual single-pixel compressed sensing imaging system has large reconstruction noise in complex diagnostic environments, this paper proposes and implements a single-pixel imaging system based on the block smooth projection Landweber quadratic reconstruction algorithm. According to the distribution characteristics of the algorithm's observation matrix and the digital micromirror hardware input requirements, the transformation of the actual projection observation matrix is realized, and the simulation analysis and experimental test of the single-pixel diagnosis are realized by using the quadratic reconstruction algorithm. The simulation results show that under the condition of a 20% to 30% sampling rate, the peak signal-to-noise ratio of the reconstructed image is greater than 20 dB, and the structural similarity is higher than 0.8. The single-pixel imaging platform is further built to complete the experimental research and verification. The experimental results show that the recovery effect of the target scene using the quadratic reconstruction algorithm model is better than the other two traditional algorithms. The quadratic reconstruction single-pixel imaging system can reconstruct a clear original image with a sampling rate of only 20%, and has good noise suppression characteristics.

With the progress of semiconductor manufacturing technology and the expansion of semiconductor laser applications, the demand of semiconductor laser for high power and miniaturized driving power becomes more and more urgent. It drives power supply towards the direction of higher power density and modular. This paper introduces the design of an all-brick drive power module based on BUCK circuit. The overall size of the power module is 116.8 mm×61 mm×12.7 mm, and the maximum output current is 12 A, the maximum adaptive output voltage is 50 V. The current ripple is less than 35 mA and the current stability reaches 0.67‰. At the same time, the power module is designed with overvoltage comparator and overcurrent protection circuit to ensure the reliability and the safe operation of laser load. The all-brick drive power module has been applied successfully in a project.

For the low efficiency of phase stereo matching in binocular structured light 3D reconstruction, a fast matching method with polar line approximation is proposed. Firstly, the polar lines are described based on the intersection line between the optical centers of the two cameras and the left pixel forming plane and the right imaging plane; the corresponding polar lines of some regions in each row of pixels are approximated, and the continuity constraint of stereo vision is combined, so that the isolated search for matching phase along the respective polar lines is replaced by the continuous search for matching phase along the approximate polar lines in the region; meanwhile, the global uniform partitioning is realized by combining the bit pose characteristics of binocular cameras to avoid repeated partitioning for each row of pixels; the calculation is assisted by the table look-up method. The global uniform partitioning is combined with the binocular camera pose characteristics to avoid repeated partitioning for each row of data. The experiments show that the average error of the point cloud obtained after stereo matching is 0.436 mm, which is within the acceptable error range, and the average speed of stereo matching calculation is increased by 10.18 times.

The grating feedback characteristics of distributed feedback (DFB) quantum cascade lasers with TM mode were simulated by difference time domain method. Comparative analysis was mainly focused on the optical properties of lateral coupled grating and ridge waveguide grating. The causes of differences, the effects of side wall angle on reflection spectrum and loss of grating were also investigated. The results show that the main factor influencing the Bragg wavelength is the effective refractive index, the optical limiting factor is the reason for the great difference of coupling coefficient between the two grating structures, the specular loss is minimum when the side wall angle of the grating is 90°. The relationship between grating period, duty cycle, etching depth and the coupling coefficient shows that the parameters not only affect the relative dielectric constant of grating, but also affects the light limiting factor, thus affecting the coupling coefficient; the coupling coefficient is proportional to the etching depth, and the variation of the coupling coefficient with duty cycle is smaller. Theoretical research on grating optical feedback characteristics is beneficial to improve the understanding of DFB quantum cascade lasers and promote the improvement and development of laser performances.

We propose a weakly coupled polarization-maintaining few-mode fiber (PM-FMF) design with elliptical-core and bow-tie stress-applying areas. Using a high refractive index core, the proposed fiber can support 32 independent eigenmodes in the 1505-1585 nm band. The combination of the elliptical-core and bow-tie stress-applying area effectively separates the adjacent eigenmodes. The structural parameters of the elliptical-core and the bow-tie stress-applying area of the PM-FMF are optimized using the finite element method. The effects of fiber parameters on the number of modes, the minimum effective refractive index difference (Δneff, mim) between modes, the mode birefringence Bm, the stress birefringence Bs, and the bending loss are evaluated. The bandwidth performance of the fiber is also analyzed, including the effective refractive index neff, Δneff and differential mode delay (DMD) between adjacent modes. The results indicate that 32 eigenmodes supported by the fiber are completely separated with Δneff, min between adjacent modes larger than 1.295×10-4 in the 1505-1585 nm band. The fiber proposed can improve the transmission capacity and has potential applications in eigenmode multiplexing transmission.

In high power fiber laser system, the cladding power stripper can remove cladding light to ensure the quality of the output laser beam. The fiber end-cap reduces the optical power density of the output fiber end by expanding the output laser beam, so as to protect the fiber end face. Both of them are important core devices for stable operation of high power fiber laser system. In this paper, the integrated design of cladding power stripper and fiber end-cap is carried out. The integrated high power cladding power stripper and fiber end-cap is applied to a 20 kW beam combination system and a single fiber system. When the output power is 20 kW, the maximum temperature of the end-cap is about 40 ℃ and the temperature rise rate is about 0.8 ℃/kW.

Power scaling of high-brightness fiber laser is limited by the mode instability and nonlinear effects. To overcome these limiting factors, a large mode area confined-doped fiber is designed and fabricated. By utilizing the homemade confined-doped fiber and backward tandem pumping scheme, fiber laser with output power of 10.1 kW is successfully achieved, where the corresponding beam quality factor （M2） is 2.16.

To realize a high power microwave source of longtime operation with a low guiding magnetic field, an S-band, GW level multi-beam relativistic klystron amplifier (RKA) has been investigated by means of theoretical modeling, numerical simulation and experiment. Firstly, a four-cavity multi-beam RKA was optimized with a one-dimension large signal code, and optimal working parameters are obtained. Under the conditions of 530 kV voltage, 4.7 kA current, and 14 beams, a 1.1 GW averaged microwave power with efficiency 43% was generated with the code. Subsequently, the beam-wave interaction parameters obtained from the code were verified with a PIC code, and a 992 MW output microwave power with efficiency 37% was obtained. At last, a long time operation experiment was conducted. In such an experiment, a 934 MW averaged microwave power with 69 ns pulse width and 33% efficiency was generated under the conditions of 530 kV voltage, 5.4 kA current, 20 Hz repetition frequency for 1 s, 0.39 T guiding magnetic field and 1.7 kW input microwave power. In addition, for the experiment of 20 Hz repetition frequency and 10 min operating time, a 889 MW averaged microwave power was obtained with 42 ns averaged pulse width. The investigation results make a strong foundation for the S-band RKAs of low guiding magnetic field and longtime operation.

Observation of the return stroke channel is a key approach in understanding return stroke's developing process and mechanism, and therefore to improve the engineering model. A lightning progression feature photic observation system (LiPOS) is used to observe the bottom luminosity from 21 m to 309 m in the artificially triggered lightning. The output-error (OE) model in system identification is applied to establish the upward propagation transfer function between different luminosity pulses. The curves of phase velocity and group velocity between 1 kHz to 1 MHz are obtained. Time-domain analysis shows that the risetime of the leading edge changes from 1.1 μs at the bottom to 1.84 μs at 309 m height, and the dispersion in velocity is the key factor to induce this effect. The frequency-domain analysis shows that the OE model can suppress the measurement noise and give a clear frequency distribution curve. Below 100 kHz, the group velocity curve has a range with strong non-monotonic variation, and the curve becomes flat and reaches 58% of light velocity above 500 kHz. The analyzed results are finally compared to the research work in the reference literature and some discussions are presented. The results may find further applications in evaluating dispersions in return strokes quantitively.

Aiming at the loaded effect of reverberation chamber, the lossy approaches in cavity are analyzed, and it is concluded that the loaded lossy is the only controllable approach in the testing process. Then five test scenarios are constructed to estimate the quality factor of reverberation chamber by using the time domain method. The result shows that the lossy effect caused by metal antenna supports is the least, while the non-metallic antenna supports could significantly load the reverberation chamber and reduce the quality factor of reverberation chamber. With the increase of the number of non-metallic supports, the effects would become more obviously. In addition, the average absorption cross section of the load is also investigated, taking all the lossy objects in the reverberation chamber as one lossy absorption cross section. It is shown that the absorption cross section of the metal antenna support is the smallest, and the loading absorption cross section of the non-metal support is significantly increased.

In view of the limitations such as close effective distance and low efficiency of the existing radioactive source detection methods, high-power microwave (HPM) is used to detect radioactive materials over a long distance because of its good spatial radiation characteristics. In this paper, the principle of microwave pulsed plasma breakdown and the influence of free electrons on breakdown characteristics are explained, the process of free electrons generated by the decay of radioactive source 137Cs is analyzed, and the HPM breakdown time and breakdown threshold are presented. Based on the HPM atmospheric breakdown plasma experimental apparatus, HPM irradiation experiments were performed multiple times in the low-pressure environment of 6000 Pa, 7000 Pa and 8000 Pa with presence and absence of radioactive sources. Experimental results show that the presence of a radioactive source lowers the HPM breakdown threshold by approximately 10% and can reduce the breakdown time by approximately 50%.

Two-dimensional PIC (Particle-in-Cell) simulation is used to investigate the electron acceleration process when a vacuum channel is filled with pre-plasma. Using a tightly focused ultra-intense short-pulse laser to interact with a hollow plasma channel is an effective way to obtain a relativistic electron beam with high power and high collimation. In the experiment, the pre-plasma generated by the laser pre-pulse ablation of the target wall will expand and fill the vacuum channel, resulting in changes in the quality of the electron beam. The simulation results show that under the condition of short-pulse laser with a power density of 5.0 $ \times {10}^{20}\;{\mathrm{W}/\mathrm{c}\mathrm{m}}^{2} $, the laser field preferentially interacts with the low-density plasma when the channel is filled with plasmas. The laser pulse interacts with the low-density plasma and the interaction of the channel wall is weakened. The electron acceleration mechanism is transformed from the vacuum electron acceleration dominated by the longitudinal field to the plasma electron acceleration dominated by the transverse electric field, resulting in an electron beam with a larger amount of charge, but with a lower energy and an increased divergence angle.

The manual classification methods for surface defects of precision optical elements are inefficient and the accuracy is easily affected by manual factors such as fatigue. And the accuracy based on traditional machine learning methods needs to be further improved. We propose an inspection method for surface defects of large-caliber optical elements using deep learning convolutional neural network. Firstly, collect and catalog a dataset of the surface defects of large-caliber optical elements through field tests. Then, for mining deeper feature expression, creat gradient-based three channels fusion image by the single-channel grayscale image. Finally, put forward the ICFNet which aims at Inertial Confinement Fusion (ICF) based on typical LeNet. The ICFNet does not require sophisticated manual design or feature extraction, only uses grayscale image to realize efficient inspection for surface flaws of large-caliber optical elements. Experiments show that ICFNet has better classification accuracy than traditional methods using multiple features and support, vector machines for three types of defects, including scratch, dust, and pits. This method has certain application value in surface defects identification of optical components.

By comparing the continuity and resolution of virtual image at “point convergence criterion” and “phase difference criterion”, it is proposed that the former is more suitable for virtual image model of wide-angle velocity interferometer system for any reflector (VISAR). Based on “point convergence criterion”, the effects of ellipsoidal mirror parameters, image recording method and shock wave tilt on the virtual image of the wide-angle VISAR target are analyzed. It is discovered that the ratio of outer and inner diameter of the virtual image is about 8 and the ellipsoidal mirror processing error has little influence on image surface when $ {10^{ - 5}}\left( {k - 400} \right) \lt m \lt {10^{ - 5}}k $ (k and m are the shape parameters of wide-angle VISAR target), it is difficult to obtain dynamic interference fringes in plane detection and wide-angle VISAR diagnosis allows the average inclination angle of shock wave not to exceed 2°. Issues such as virtual image interference of wide-angle VISAR, processing of special-shaped fiber optic panels, selection of ellipsoidal mirror parameters and image reconstruction are discussed, and some suggestions for diagnostic experiments are given. The research on properties of virtual image lays a foundation for the improvement of wide-angle VISAR diagnostic capability, and is of great significance for the symmetry diagnosis in inertial confinement fusion.

For multi-MeV X-ray flash radiography, the areal density of object can be obtained by the primary direct X-ray. Objects of flash radiography often have very high areal densities which greatly attenuate the intensity of direct X-rays emitted by the source. At this time, the direct penetration signal that can transmit the region of interest inside the object will be smaller than that of the scattered X-ray “noise”. If the captured image is reconstructed directly without scatter correction, it will affect the accuracy of reconstruction. The main method to reduce the scatter X-ray from a physical point of view is to use an anti-scatter grid, that is, an array-type collimation hole. However, the performance may be affected by the stability of the X-ray source spot, and the manufacture of such anti-scatter grid is very difficult. This paper proposes a new imaging method that does not rely on anti-scatter grid. This method only makes small improvements on the existing imaging layout, and can easily and self-consistently determine the amount of scattering for scattering correction. A Monte Carlo simulation is given to show the performance of scatter estimation, and the relative difference between the estimated intensity of scatter and the real value (provided by the simulation) for an object is less than 2% when a known object with less areal density is applied for calibration.

Neutral beam injection systems based on RF negative ion sources are the best choice for high power long pulse (steady-state) operation of neutral beam injection systems. The negative ion source is the core component of the neutral beam injection system, which needs to achieve stable negative ion beam extraction and acceleration. During the operation of the negative ion source, the negative ion current will change, especially under long-pulse and high-energy operating conditions, so it cannot meet the requirements of stable operation. To achieve a stable extraction of the extracted beam, the research on beam feedback control was carried out and a beam feedback control system based on RF power adjustment was developed. The beam feedback control system has been applied in the RF negative ion source test platform. The test results show that the beam feedback control system can realize real-time feedback adjustment of the beam to obtain stable extraction of the beam. Beam feedback control based on RF power regulation is feasible, which provides support for the development of high-power RF negative ion sources.

Quench protection system plays an important role in superconducting magnet power supply system, as it can transfer and consume magnet energy rapidly. In the project of Comprehensive Research Facility for Fusion Technology (CRAFT), the technical requirement of breaking 100 kA direct current (DC) is proposed for the quench protection system, in which the vacuum circuit breaker is used as the switch of transfer branch. In this paper, a 100 kA DC vacuum circuit breaker with series structure is designed for the CRAFT project’s quench protection system, and the prototype is manufactured. In field test, with the cooperation of artificial zero-point circuit, the vacuum circuit has completed the 100 kA (DC) breaking test successfully.

It is necessary to monitor plutonium aerosol when doing experimental research with plutonium material. Plutonium material experiments are usually carried out in sealed containers, which guarantees that the plutonium aerosol will not leak out to the environment. The widely-used monitoring equipment are not suitable for plutonium aerosol monitoring in sealed containers because of its large volume. A new plutonium aerosol measurement system based on wavelength shift fiber and silicon photomultiplier (SiPM) is developed. In the new plutonium aerosol measurement system, ZnS(Ag) scintillator is used as detection material and wavelength shift fiber is used as photon transmission media . The new plutonium aerosol measurement system has the advantages of customizable detector size and shape, low power consumption, and relatively simple structure, which realizes remote measurement of plutonium aerosol in sealed containers. The measurement system can also discriminate α particles in n/γ-mixed radiation field.

To meet the beam intensity measurement requirements at the extraction section of the cyclotron in Heavy Ion Medical Machine (HIMM), the integral current transformer (ICT) and lock-in amplifier scheme was adopted to implement non-destructive and real-time beam intensity measurement on the medium energy beam transport line (MEBT). ICT acquires the relative beam intensity which can’t be monitored directly, as a result, the Faraday cup which is a kind of destructive detector was used to achieve calibration of ICT with beam. In this paper, the requirements and design scheme of beam intensity measurement at MEBT are firstly analyzed, and based on the design scheme, tests with this beam intensity measurement system are carried out in the laboratory and with beam. According to the test results, the beam intensity stability is about 90 nA, while the corresponding relative error is about 8%. Furthermore, the response time of ICT and Faraday cup system is less than 1 ms and 100 ms respectively, which meet the physical measurement requirements. Further study on the relationship between frequency-variation of cyclotron radio-frequency system and beam current measured by ICT will be carried on in the next step.

To improve the alignment accuracy of the magnet in high energy synchrotron radiation source, the magnetic center is used instead of the mechanical center to calibrate, the magnetic center position of the magnet is obtained by vibration wire or rotation wire technology, and the position of the wire is measured by capacitive sensor, so as to realize the association between the magnetic center and the collimation target. To achieve high precision measurement of wire position, it is necessary to calibrate capacitance sensor accurately. In this paper, a kind of capacitive sensor is introduced and its calibration method is studied, the data acquisition method of grid and data processing method of high order polynomial fitting are proposed, the sensor calibration platform is built and the corresponding calibration program is developed, and the automatic and high precision calibration of the sensor is realized. Analysis and comparison show that the calibrated capacitance sensor has achieved the displacement measurement accuracy of μm level, which provides a basis for high-precision collimation of magnets.

Taking the cross-linked ethylene tetrafluoroethylene copolymer (X-ETFE) cable used outside the spacecraft as the test object, the X-ETFE cable was irradiated with 8000 equivalent solar hours (ESH) vacuum ultraviolet (VUV) with a 5-fold acceleration factor. The electrical properties of the X-ETFE cable were analyzed through the limit voltage resistance and insulation material resistance tests. The molecular structure and micro morphology of the X-ETFE material were characterized by FTIR and SEM, The effects of different VUV irradiation time on X-ETFE cable have been studied.. The experimental results show that with the increase of VUV irradiation time, carbon accumulates on the material surface and darkens, and the appearance color of the cable gradually changes to dark brown; The ultimate withstand voltage and insulation resistance of X-ETFE cable show an overall downward trend, but the overall electrical performance level has no substantial change; The absorption peak of X-ETFE material at 1628 cm-1 gradually increases, indicating that the -C=C- free group in the molecular chain of X-ETFE material increases with irradiation time, resulting in microcracks on the surface of the material.

This paper introduces the trigger system of the middle energy X-ray device and experimental results. The trigger system controls triggering of laser-switches in six modules and the electric trigger part. Each switch is triggered by one laser. Seven lasers could be adjusted stand-alone. The test results show that the time jitter of laser was less than 0.3 ns and the time jitter of the laser-switch was less than 1.2 ns. For electric trigger system, the voltage of the matching load was 120 kV, the rising time was about 28 ns, the pulse width (FWHM) was 150 ns, and the time jitter was less than 0.5ns. The middle energy X-ray device could produce a maximum output voltage and current of about 4.2 MV /100 kA and the pulse width (FWHM) was 150 ns. The time jitter of X-ray was less than 3.4 ns in the middle energy X-ray device. The experimental results show that the trigger system was able to adjust the six modules accurately, which improves reliability of the middle energy X-ray device.

The cooling-down time limits the capability of repetitive operation of the pulsed magnet. A fast cooling method for the pulsed magnet based on heat transfer of flowing liquid nitrogen (LN2) in micro-channels formed inside the conductors of the pulsed magnet is presented. The large amount of heat produced during discharging of the pulsed magnet can be quickly dissipated by LN2 inside the micro-channels through the enlarged contact areas between LN2 and conductors, by single-phase LN2 flow and/or flow boiling. Furthermore, the impacts of the micro-channels on the performances (strengthening of the magnetic field, pulse duration and diameter of inner bore) can be tolerable. The principles of fast cooling method based on single-phase LN2 flow or flow boiling are elucidated. Numerical simulations and validation experiments of the fast cooling method indicate that pulsed magnet with inner bore diameter of 20 mm and magnetic field of 25 T can be cooled down in 30 s. The cooling speed of the pulsed magnet of the fast cooling method is increased by about 19 times compared with the conventional cooling method (600 s) where the pulsed magnet is simply immersed in LN2.

In the application of pulse power technology in tumor ablation, sewage treatment and other fields, it is found that bipolar electric pulse is often better than unipolar electric pulse, which greatly stimulates the research and development demand of bipolar high-voltage pulse power supply. A constant peak bipolar pulse generator is designed based on Boost closed-loop control. The generator perfectly combines the characteristics of Boost circuit and Marx generator to realize the generation of bipolar pulse with boost function. The peak detection circuit is used to sample the peak value of bipolar pulse and feed it back to DSP to realize closed-loop control, so as to realize the output of constant peak bipolar pulse. To verify the feasibility and stability of the proposed topology circuit, simulation and experiment of the developed 5-stage constant peak bipolar pulse generator are carried out. The experimental results show that when the input voltage is 100 V, a constant peak bipolar pulse waveform with repetition frequency of 5 kHz, pulse width of 5-10 μs and voltage amplitude of ±2.0 kV can be generated. The pulse power supply uses modular design, which is easy to cascade, compact, and can flexibly output bipolar or unipolar positive (negative) pulses with constant peak value.