Inverse synthetic aperture LiDAR (ISAL) is a radar system that combines synthetic aperture imaging and laser technology. However, due to high cross-range repetition frequency of the laser band, system complexity, and data redundancy, the echo signals of ISAL are usually nonuniform and incomplete. These data are called sparse aperture (SA) data, which lead to defocusing in ISAL images, and traditional imaging algorithms have limited capability to reconstruct target details under SA conditions. To address this problem, we propose a new framework called complex-valued primal-dual network (CPD-Net). Specifically, we unroll the iterations of the primal-dual (PD) hybrid gradient algorithm to a learnable deep network architecture, and gradually relax the constraints to reconstruct ISAL images from highly undersampled range profiles. To enable ISAL to image targets at different scales, we designed a complex-valued fusion inception module (CFIM) based on the Inception structure and CNNs. The Inception layers in CFIM employ multiple convolutional layers with different kernel sizes in parallel. The feature matrices obtained from different convolutional operations are concatenated along the depth dimension to form a deeper matrix, thereby enhancing the feature extraction capability of CPD-Net. In addition, the CPD-Net employs complex-valued operations to process the phase information of laser signals, and incorporates residual connections to improve CPD-Net's performance. Experiments on simulated data and measured data demonstrate that CPD-Net can obtain well-focused ISAL images with clear backgrounds under different signal-to-noise ratios (SNRs) and SA types. Moreover, compared with traditional optimization algorithms, CPD-Net significantly reduces computational time, making it more practical for real ISAL applications.

This study focuses on developing a dual-Helmholtz-coil magnetic field system for polarized He-3 to generate a uniform magnetic field. A theoretical analysis of the mechanism by which dual Helmholtz coils produce a uniform magnetic field was first conducted. Based on these findings, the magnetic device was designed with the finite element analysis software COMSOL, employing an optimization algorithm to efficiently set current densities and minimize the transverse gradient in the central region. The simulated transverse gradient achieved was 9.54 & times; 10(-5) cm(-1) in the 20 cm & times; 20 cm & times; 8 cm central region, and the experimental measurement yielded a value of 6.31 & times; 10(-4) cm(-1). Additionally, lifetime testing of two polarized He-3 cells revealed lifetimes of 151.0 +/- 1.8 h and 206.2 +/- 4.4 h, demonstrating the excellent magnetic field uniformity of the device. The current work integrates theoretical, simulated, and experimental results. It advances polarized He-3 magnetic technology, offers an upgrade solution for off situ polarized He-3 pumping stations to extend the volume of the uniform magnetic field without enlarging the coil size, and provides a method for using polarized He-3 on neutron beamlines requiring large acceptance angles.

The segmented solar telescope described in this study employs a simultaneous dual-wavelength measurement technique to achieve co-phase alignment.To meet the measurement requirements of a 20 μm range,5 nm root mean square precision,and edge jump rates of<10−6,this study focused on calibrating the dual-wavelength measurement system for the segmented-mirror solar telescope.Analysis of the relative error in the measurement system revealed that assembly-induced errors such as defocus,translation,scaling,and rotation markedly degrade measurement accuracy.To address these issues,we propose a defocus error compensation algorithm,based on the light intensity distribution of the point spread function(PSF)and an affine transformation model,to calibrate spatial pose deviations across the two measurement channels.A dual-wavelength measurement system was implemented on a segmented-mirror experimental platform for calibration.Experimental results demonstrated that the mean relative error decreased from −0.642 3 to −0.034 5 nm after calibration,reflecting improved reliability and stability of the co-phase measurements.

Objective Carrier frequency stability is one of the main factors affecting the imaging quality in inverse synthetic aperture LiDAR (ISAL) systems. Traditional optical and phase correction algorithms typically rely on hardware requirements or simplified models, making it difficult to obtain focused ISAL images under conditions of nonlinear frequency drift with noise. Therefore, we propose a diffusion model-based compensation method named Rdiffusion, which embeds carrier instability characteristics into the diffusion generation and inversion processes, enabling physics-based modeling and correction, thereby enhancing ISAL imaging quality and robustness. Methods Rdiffusion employs physical carrier jitter and recovery mechanisms as its forward and reverse diffusion processes. First, an ISAL signal model is established to describe the influence of carrier drift on range profiles. Then, the diffusion time step t is mapped to the carrier modulation frequency, ensuring the forward process of Rdiffusion corresponds to the ISAL signal model. The backward process of Rdiffusion predicts the carrier frequency diffusion mechanisms. Specifically, after learning results for different time scales via U-Net, Rdiffusion applies the inverse of the forward process to the defocused range profile to obtain a focused ISAL image. During training, Rdiffusion is optimized to predict the focused ISAL image x 0 under mean square error loss. Results and Discussions The imaging results from the range-Doppler (RD) algorithm exhibit pronounced defocusing and blurring, accompanied by significant speckle noise. As a supervised model, convolutional neural network (CNN) imaging quality often depends on the quality of labeled images, making such networks difficult to apply to non-cooperative target imaging. In contrast, the Rdiffusion model effectively suppresses defocusing effects caused by jitter and significantly reduces background noise. Numerical results of simulated data show that Rdiffusion achieves the highest peak signal-to-noise ratio (PSNR) and structure similarity index measure (SSIM), and the lowest image entropy (ENT), respectively. In Fig. 9, we selected measurement data from the MiG-25 to validate the performance and generalization capability of Rdiffusion. The imaging results from the RD algorithm exhibited severe defocusing effects, with target structures becoming progressively blurred and accompanied by significant background interference. The quality of CNN images also deteriorated substantially. In contrast, Rdiffusion maintained the primary morphology of the target effectively and demonstrated superior imaging stability compared to the RD algorithm. Additionally, we analyzed the computation times of different algorithms. The average computation time for Rdiffusion was approximately (0.078 +/- 0.005) s, while CNN's computation time was approximately (0.062 +/- 0.01) s. The computation time of the RD algorithm was about (4 +/- 0.7) s. This error range is based on statistics from 50 repeated tests. It is evident that neural networks demonstrate a significant computational efficiency advantage over traditional imaging algorithms. Furthermore, Rdiffusion ensures high imaging quality and stability under complex jitter conditions within a short timeframe, balancing imaging performance with real-time capability. This exhibits strong potential for engineering applications. Conclusions To address phase mismatch and imaging performance degradation caused by far-field carrier frequency jitter in ISAL systems, we propose an innovative Markov chain diffusion model named Rdiffusion, based on physical process embedding. Rdiffusion explicitly incorporates carrier frequency jitter mechanisms into the Markov chain structure of the diffusion process. This ensures that the generated chains during training align with the actual target degradation physical processes, thereby guaranteeing more robust image restoration under physical prior constraints during inference. The core advantage of Rdiffusion lies in achieving a deep integration of physical mechanisms and data-driven approaches. On one hand, physical constraints provide the diffusion model with a generation process highly consistent with imaging degradation mechanisms, effectively enhancing the generalization capability and interpretability of Rdiffusion. On the other hand, the data-driven training mechanism ensures the model can still achieve high-resolution imaging results in complex scenarios. In summary, the proposed Rdiffusion model balances computational efficiency with imaging quality, demonstrating robustness and scalability in complex interference environments.

Achromatic metalenses face stringent aperture and numerical aperture (NA) constraints, which have become a key bottleneck in metasurface imaging. To this end, a novel achromatic imaging method was first proposed, utilizing the unique wideband consistency of diffracted Bessel spots combined with non-blind image restoration techniques. To address the off-axis aberration issue, off-axis achromatic meta-axicons with eccentric conical phases were further designed to convert oblique plane waves into wideband uniform off-axis Bessel beams. Ultimately, a single metasurface integrating 9 meta-axicons with different design field angles was developed, and a meta-camera was constructed accordingly. After image restoration, the meta-camera achieves achromatic imaging within a 10° stitched field of view (FOV), and an angular resolution close to that of near-diffraction-limit lens with the same aperture throughout the entire FOV. The core idea of achieving achromatic imaging based on natural dispersion laws in this study enables the wideband minimalist optical system based on metasurface to completely circumvent the aperture limitation, providing a highly valuable solution for large-aperture meta-camera design that can simultaneously accommodate wideband and off-axis FOV.

Unstable ignition under low-load conditions remains a major challenge limiting the application of methanol in spark-ignition engines, despite its high-octane number and low-carbon content. Improving ignition stability and flame development is therefore critical for enabling efficient and clean methanol combustion. In this study, the aim is to investigate the effects of active pre-chamber methanol injection on turbulent jet flame development, equivalence ratio distribution, temperature fields, and OH radical formation. Optical diagnostics (schlieren imaging and flame chemiluminescence) combined with numerical simulations were employed to analyze ignition, jet penetration, and radical generation. Results show that the global equivalence ratio ($g) plays a critical role in jet flame morphology, ignition delay, and combustion duration. As $g increases, the jet flame transitions from a lean, weak combustion regime ($g =1.5) to an optimal condition characterized by strong penetration and sustained high-temperature/high-OH zones ($g = 2.0), and finally to an over-fueled state ($g = 2.5-4.0), resulting in oxygen deficiency and incomplete combustion. Probability density function (PDF) analysis indicates that $g = 2.0 ensures reliable ignition and enlarges the fraction of highly reactive zones, producing stronger flame kernels and more effective main-chamber ignition. Analyses of temperature and OH distributions further confirm that the moderate fueling case ($g = 2.0) yields the most concentrated and stable flame core, balancing radical buildup with heat release. These findings highlight that optimizing pre-chamber methanol injection is critical for achieving stable, efficient, and clean combustion, and provide guidance for advancing high-efficiency, low-carbon spark-ignition engines.

The in-situ aberration detection of optical systems is of great significance for the processing and alignment of optical systems, the development of lithography machines, and the on-orbit adjustment of space cameras. Traditional in-situ detection methods for optical systems, such as Phase Retrieval (PR) and Phase Diversity (PD), perform excellently under specific conditions. But they have limitations when facing complex conditions such as large numerical apertures or when the lower bound of the Nyquist frequency of the optical system is not satisfied. Therefore, a method is proposed to combine the extended Nijboer-Zernike diffraction physical model with a deep neural network. Firstly, a deep residual network with the squeeze-and-excitation (SE) attention mechanism is constructed. Secondly, a mapping relationship from the intensity image point spread function (PSF) to the phase distribution is established to achieve feature extraction of the diffracted light intensity and prediction of the coefficients for phase description. Finally, the predicted coefficients are combined with the ENZ diffraction model to obtain the predicted PSF image, so as to realize the wavefront detection of the optical system. Experimental results show that when the numerical aperture (NA) of the optical system is large and the Nyquist sampling is not satisfied, the residual wavefront RMS between the real wavefront image and the reconstructed wavefront image is about 0. 02 lambda, which is better than other methods. In comparison with other deep learning methods, this method is an unsupervised method, which not only reduces the dependence on a large amount of training data but also improves the accuracy of wavefront detection.

The Chinese Giant Solar Telescope is the next-generation infrared and optical solar telescope of China, with a current design featuring an 8-meter aperture and a 2-meter-wide ring primary mirror structure. This ring-segmented structure effectively addresses the thermal control and high-precision magnetic field measurement challenges. For the Chinese Giant Solar Telescope, the segmented alignment process directly influences the co-focus and co-phase of the primary mirror. During this process, two critical technical issues must still be resolved: first, for the three out-of-plane degrees of freedom (Piston: translation along the Z-axis; Tip and Tilt: rotations about the X-axis and Y-axis) that require active adjustment, the initial mechanical alignment errors on the millimeter scale must be minimized to within the capturing range of subsequent optical measurements (such as the co-phase measurement system and wide-field camera). Second, for the in-plane degrees of freedom (Dx and Dy: translation along the X and Y axes; Clocking: rotation about the Z-axis), which do not undergo active adjustment but still impact the imaging quality of the segmented, the adjustments must be within the permissible error range. To address this, based on the actual needs of segmented alignment for The Chinese Giant Solar Telescope, it presents the corresponding alignment requirements for both out-of-plane degrees of freedom and in-plane degrees of freedom, considering their adjustment characteristics and influencing factors. The alignment requirements for the out-of-plane degrees of freedom are as follows: Tip and Tilt must be adjusted to within +/- 1', and the piston to within +/- 100 mu m. For the in-plane degrees of freedom, the required adjustments are as follows: Dx must be adjusted to +/- 67 mu m, Dy to +/- 25 mu m, and clocking to +/- 0.07', in order to meet the image quality requirements of the primary mirror. Currently, three-dimensional coordinate measuring instruments are commonly used for segmented alignment, and the laser tracker has become the most widely used tool due to its high precision, long working distance, and strong real-time measurement capabilities. This paper focuses on exploring the feasibility of using a laser tracker for the initial adjustment of the Chinese Giant Solar Telescope, particularly its applicability in adjusting both in-plane and out-of-plane degrees of freedom. Given the ring primary mirror structure of the Chinese Giant Solar Telescope, a measurement scheme based on the laser tracker is proposed: Reflectors are placed on surfaces other than the optical surface of the segment, such as the side or bottom surfaces, and the spatial position of the segment is described using its local coordinate system (with the position of the reflectors known relative to the segmented local coordinate system). The laser tracker is then employed to measure the position of the reflector on the segment. By combining the measurement results from the laser tracker with the position of the fixed points relative to the local coordinate system, a least-squares fitting method is used to calculate the six coordinate transformation relationships between the segment (or between the segment and a specific spatial reference position), which include three rotational degrees of freedom and three translational degrees of freedom. Through these six transformation parameters, the in-plane and out-of-plane degrees of freedom of the segment can be extracted and compensated for via the adjustment mechanism, thereby achieving precise alignment of the segment. During the actual alignment process, the primary factor affecting the alignment precision is the measurement accuracy of the in-plane and out-of-plane degrees of freedom, as the adjustment mechanism provides high precision. Measurement accuracy is influenced by two main factors: first, the measurement errors of the laser tracker, which are related to the working distance (the further the distance, the greater the error); second, systematic errors caused by the installation of fixed point positions, which are related to installation errors and the roughness of the installation surface, leading to systematic biases in the measurement results. To verify the feasibility of the proposed method, preliminary experimental validation was conducted in a ring-segmented experimental system, demonstrating the viability of the out-of-plane degrees of freedom measurement approach. For the Chinese Giant Solar Telescope, the estimated results show that when the working distance of the laser tracker is less than 8 m and the systematic error at the fixed points is less than 10 mu m, the measurable accuracies are as follows: Tip is +/- 0.57', tilt is +/- 0.33', piston is +/- 64.58 mu m, Dx is +/- 56.89 mu m, Dy is +/- 52.78 mu m, and clocking is +/- 0.57'. These results show that using a laser tracker for adjusting the out-of-plane degrees of freedom can effectively meet the accuracy requirements for integration with the subsequent optical measurement system. However, due to the higher precision required for the in-plane degrees of freedom, especially for Dy and Clocking, it is currently difficult to further improve the alignment precision of these degrees of freedom using only the laser tracker. Therefore, it is necessary to combine other measurement methods to explore alignment solutions with even higher precision. This research provides valuable reference for the alignment work of the Chinese Giant Solar Telescope and offers technical support for the subsequent implementation of co-focus and co-phase, as well as for improving the effectiveness of co-phase control.

The challenge of observing faint space objects during daytime has long hindered the full-time utilization of ground-based telescopes. Interference detection technology serves as a novel approach for suppressing skylight background noise, which can be employed either independently to mitigate skylight background or combined with conventional methods to achieve enhanced suppression capabilities. This study focuses on the observational performance of an interference detection system for actual faint space targets in daylight conditions. Through observations of three stellar targets (Arcturus, Vega, and Polaris) during sunset, we present, to our knowledge, the first demonstration of daytime stellar detection using an interference detection system. By comparing observational results between the shearing interference detection system and the conventional detection system for these stellar targets, this work reveals that the shearing interference method significantly outperforms traditional approaches-both in terms of visual image quality and target signal-to-noise ratio (SNR). Specifically, the shearing interference system achieved SNRs of 15.4 for Arcturus, 23.3 for Vega, and 23.3 for Polaris, whereas conventional methods attained maximum SNRs of only 5.0 (Arcturus), 10.7 (Vega), and 3.8 (Polaris).

Segmented mirror technology represents a crucial direction in the development of large-aperture telescopes,with co-phase technology being the key challenge for achieving diffraction-limited imaging.Compared to monolithic telescopes,segmented telescopes exhibit significant differences in secondary mirror alignment.This paper systematically reviews the secondary mirror alignment techniques and segment co-phase adjustment methods employed in existing segmented telescopes.Additionally,we analyze the co-phase detection technologies applied in several domestic segmented experimental systems.The study evaluates the advantages,limitations,application domains,and future trends of current co-phase detection methods.This research aims to provide technical insights and references for future studies on alignment and co-phase adjustment schemes in segmented telescopes.

We theoretically and experimentally demonstrate the generation of high-topological charge, extreme-ultraviolet (EUV) spatiotemporal optical vortices (STOV) from high-order harmonic generation. EUV-STOVs are unique structured light tools for exploring ultrafast topological laser-matter interactions.

In the field of coherent diffraction imaging, phase retrieval is essential for correcting the aberration of an optic system. For estimating aberration from intensity, conventional methods rely on neural networks whose performance is limited by training datasets. In this Letter, we propose an untrained physics-driven aberration retrieval network (uPD-ARNet). It only uses one intensity image and iterates in a self-supervised way. This model consists of two parts: an untrained neural network and a forward physical model for the diffraction of the light field. This physical model can adjust the output of the untrained neural network, which can characterize the inverse process from the intensity to the aberration. The experiments support that our method is superior to other conventional methods for aberration retrieval.

This article presents a 6.78 MHz single-stage wireless power transfer receiver (RX) using 0X/1X regulating rectifier with improved voltage mode (VM) delay compensation for implantable medical devices (IMDs). The regulation of 0X/1X modes is achieved through pulse width modulation (PWM), with the frequency set at 1/16 of the carrier clock to ensure circuit stability and enhance power conversion efficiency (PCE) under light loads. VM delay compensation has the advantage of power consumption. However, the traditional VM on-delay compensation may become ineffective because a large voltage pulse will occur on the RX side input when the circuit delays increase. Therefore, an improved VM on-delay compensation method is proposed. Coarse compensation for on-delays can be achieved by reusing the control signal generated from off-delay compensation loop, as off-delay compensation processes will not be influenced by on-delay compensation processes. It allows the on-delay compensation loop to function normally and finely compensate for on-delays. The proposed rectifier is designed in 0.18μm CMOS process. It can achieve output voltage management of 1.8V. Comparing with other regulating rectifier, the proposed rectifier has a fast transient response of 14 μs and a small chip area of 0.91 mm 2 . The simulation results indicate that this rectifier can attain high PCE of 84.6% in light load (500Ω) and peak PCE of 93.5% in heavy load (50Ω). The high voltage conversion ratio (VCR) of over 93% is also achieved under the large circuit delays.

With the development of space detection technology, the detection of long-range dark and weak space targets has become an important issue in space detection. Cross-strip anode photon imaging detectors can detect weak light signals with extremely low dark count rates and are well suited to applications in long-range target detection systems. Since cross-strip anode detectors are expensive to develop and fabricate, a theoretical analysis of the detection process is necessary before fabrication. During the detection process, due to the dead time of the detector, some photon-generated signals are aliased, and the true arrival position of the photon cannot be obtained. These aliased signals are usually removed directly in the conventional research. But in this work, we find that these aliased signals are not meaningless and can be applied to center of mass detection. Specifically, we model the probabilistic mechanisms of the detection data, compute the average photon positions using aliased and non-aliased data and prove that our method provides a lower variance compared to the conventional method, which only uses non-aliased data. Simulation experiments are designed to further verify the effectiveness of the aliasing data for detecting the center of mass. The simulation results support that our method of utilizing the aliasing data provides more accurate detection results than that of removing the aliasing data.

Soft x-ray coherent diffractive imaging is used to probe the fundamental scales of topological magnetic textures. 3D vector ptycho-tomography enables high resolution 3D static imaging of nanoscale spin textures in dipole-stabilized skyrmions.

Spatiotemporal orbital angular momentum (ST-OAM) of light is an emergent, spatiotemporally sculptured light. Such spatiotemporal optical vortices carry transverse OAM and exhibit novel properties. However, the lack of a simple and straightforward characterization method substantially slows its progress and potential adaptions for future applications. Here we demonstrated a simple, stationary, single-frame method to quantitatively characterize ST-OAM pulses. Our new method can measure the presence of ST-OAM, space-time topological charge numbers, OAM helicity, pulse dispersion, and beam divergence. We also investigated the nonlinear properties of ST-OAM pulses, uncovering the conservation of space-time topological charges in a second-harmonic generation process.

Methods to probe and understand the dynamic response of materials following impulsive excitation are important for many fields, from materials and energy sciences to chemical and neuroscience. To design more efficient nano, energy, and quantum devices, new methods are needed to uncover the dominant excitations and reaction pathways. In this work, we implement a newly-developed superlet transform-a super-resolution time-frequency analytical method-to analyze and extract phonon dynamics in a laser-excited two-dimensional (2D) quantum material. This quasi-2D system, 1T-TaSe2, supports both equilibrium and metastable light-induced charge density wave (CDW) phases mediated by strongly coupled phonons. We compare the effectiveness of the superlet transform to standard time-frequency techniques. We find that the superlet transform is superior in both time and frequency resolution, and use it to observe and validate novel physics. In particular, we show fluence-dependent changes in the coupled dynamics of three phonon modes that are similar in frequency, including the CDW amplitude mode, that clearly demonstrate a change in the dominant charge-phonon couplings. More interestingly, the frequencies of the three phonon modes, including the strongly-coupled CDW amplitude mode, remain time- and fluence-independent, which is unusual compared to previously investigated materials. Our study opens a new avenue for capturing the coherent evolution and couplings of strongly-coupled materials and quantum systems.

The Latin hypercube design (LHD), because of its one-dimensional projection uniformity, is commonly used in computer experiment. The randomly generated LHD may have too many concentrated design points, and factors may be highly correlated. In this article, we suggested a local greedy strategy for searching optimal LHDs. Our strategy consists of two parts. One is a swap process for doing a local greedy search in a polynomial time. The other is a simulated annealing process for jumping out of the possible local optima. Our strategy is flexible and adapts to various space-filling criteria of LHDs. The simulated experiments illustrated that our proposed algorithm can produce LHDs with well space-filling property and orthogonality. Compared to other classical design algorithms, our algorithm performed better on the criteria related to the point distance and the column correlation. Moreover, for the response surface approximation, the Kriging model using our produced optimal LHD performed more robust on the surface prediction.