Divergence Compensation Research Articles

JFLN: Joint Feature Learning Network for 2D sketch based 3D shape retrieval

Cross-modal retrieval attracts much research attention due to its wide applications in numerous search systems. Sketch based 3D shape retrieval is a typical challenging cross-modal retrieval task for the huge divergence between sketch modality and 3D shape view modality. Existing approaches project the sketches and shapes into a common space for feature update and data alignment. However, these methods contain several disadvantages: Firstly, the majority approaches ignore the modality-shared information for divergence compensation in descriptor generation process. Secondly, traditional fusion method of multi-view features introduces much redundancy, which decreases the discrimination of shape descriptors. Finally, most approaches only focus on the cross-modal alignment, which omits the modality-specific data relevance. To address these limitations, we propose a Joint Feature Learning Network (JFLN). Firstly, we design a novel modality-shared feature extraction network to exploit both modality-specific characteristics and modality-shared information for descriptor generation. Subsequently, we introduce a hierarchical view attention module to gradually focus on the effective information for multiview feature updating and aggregation. Finally, we propose a novel cross-modal feature learning network, which can simultaneously contribute to modality-specific data distribution and cross-modal data alignment. We conduct exhaustive experiments on three public databases. The experimental results validate the superiority of the proposed method. Full Codes are available at https://github.com/dlmuyy/JFLN.

Hybrid Smith predictor and phase lead based divergence compensation for hardware-in-the-loop contact simulation with measurement delay

On the ground the hardware-in-the-loop (HIL) simulation is a good approach to test the contact dynamics of spacecraft docking process in space. Unfortunately, due to the time delay in the system the HIL contact simulation becomes divergent. However, the traditional first-order phase lead compensation approach still result in a small divergence for the pure time delay. The serial Smith predictor and phase lead compensation approach proposed by the authors recently will lead to an over-compensation and an obvious convergence. In this study, a hybrid Smith predictor and phase lead compensation approach is proposed. The hybrid Smith predictor and phase lead compensation can achieve a higher simulation fidelity with a little convergence. The phase angle of the compensator is analyzed and the stability condition of the HIL simulation system is given. The effectiveness of the proposed compensation approach is tested by simulations on an undamped elastic contact process.

Mid-infrared ZGP OPO with divergence compensation and high beam quality.

Divergence compensation, optimization of the optical-to-optical efficiency, and high beam quality of signal and idler beams of a high-energy mid-infrared ZnGeP2 (ZGP) optical parametric oscillator (OPO) have been demonstrated by use of a Galilean telescope inside the nonplanar fractional-image-rotation enhancement (FIRE) ring resonator. With a small variation of the distance between the lenses of the telescope, the divergences of signal and idler beams could be adjusted. Up to 36 mJ of mid-infrared pulse energy in the 3-5 µm wavelength range is obtained with 92 mJ of pump energy on crystal. The beam quality factors M2 are < 1.5 for the resonant signal beam and the non-resonant idler beam, respectively. Actually, this is an improvement of the beam quality by a factor 3 for the signal and ~2.7 for the idler beam compared without using a telescope inside the FIRE ring resonator.

Low-Order Model based Divergence Compensation for Hardware-In-The-Loop Simulation of Space Discrete Contact

The divergence problem of the hardware-in-the-loop (HIL) simulation for flying objects contact in space is investigated. The HIL simulation divergence is due to time delays of the motion simulator and the force measurement system. In this study, a low-order model based divergence compensation method is proposed. The force measurement delay is modeled as a pure delay and compensated by a first-order phase lead force compensator after truncating high-order terms of the Taylor expansion. For the motion simulator, its dynamic response delay is modeled as a three-order model and compensated by a second-order phase lead position compensator. The low-order compensators are used because high-order derivatives of the signal are inaccurate and sensitive to noises. Simulations and experiments are performed to verify that the proposed approach can compensate the simulation divergence effectively and satisfactorily.

Divergence compensation for hardware-in-the-loop simulation of stiffness-varying discrete contact in space

The hardware-in-the-loop (HIL) contact simulation for flying objects in space is challenging due to the divergence caused by the time delay. In this study, a divergence compensation approach is proposed for the stiffness-varying discrete contact. The dynamic response delay of the motion simulator and the force measurement delay are considered. For the force measurement delay, a phase lead based force compensation approach is used. For the dynamic response delay of the motion simulator, a response error based force compensation approach is used, where the compensation force is obtained from the real-time identified contact stiffness and real-time measured position response error. The dynamic response model of the motion simulator is not required. The simulations and experiments show that the simulation divergence can be compensated effectively and satisfactorily by using the proposed approach.

Fabrication of periodic nanostructures by single-point diamond turning with focused ion beam built tool tips

Periodic nanostructures have been widely used on emerging nano-products such as plasmonic solar cell and nano-optics. However, lack of cost-effective fabrication techniques has become the bottleneck for commercialization of these nano-products. In this work, we develop a scale up approach to fabricate high-precision nanostructures in large area. In this method, a nano-scale single crystal diamond (SCD) tool is produced by focused ion beam (FIB) machining. The nano SCD tool is then further applied to cut periodic nanostructures using single-point diamond turning (SPDT). A divergence compensation method and surface topography generation model forms a deterministic FIB fabrication approach. It has been used to generate four periods of the required periodic nano-grating structures (with a minimal dimension of 150 nm) on a normal SCD tool tip and achieves 10 nm form accuracy. The contribution of the beam tail effect has also been evaluated by using the surface topography simulation method. The fabricated diamond tool is then applied to obtain nano-grating on an electroless nickel substrate in a total area of 5 × 2 mm2 through SPDT. The whole SPDT machine process only takes 2 min (with a material removal rate up to 1.8 × 104 μm3 s−1). Due to the elastic recovery that occurred upon the workpiece material, the practical cutting width is 13 nm smaller than the tool tip. The machining trial shows it is very promising to apply this scale up nanofabrication approach for commercialization of nano-products which possess period nanostructures.

Fabrication of Freeform Micro Optics by Focused Ion Beam

In this work, two kinds of freeform micro optics were successfully fabricated by using focused ion beam machining. A divergence compensation method was applied to optimize the machining process. Both dynamic variation of the sputter yield and the extra ion flux contributed by the beam tail were taken into consideration. Measurement results on the surface topography indicated that 3-fold improvement of the relative divergence was achieved for both optics when compared with conventional focused ion beam milling without any corrections. Furthermore, investigations on the influences of scanning strategies, including raster scan, serpentine scan and contour scan, were carried out. The serpentine scan is recommended for the fabrication of freeform optics by focused ion beam technology owing to the minimal beam travelling distance over the pattern area.

A predictive divergence compensation approach for the fabrication of three-dimensional microstructures using focused ion beam machining

In this paper, the major factors which cause fabrication divergence in the focused ion beam (FIB) milling process are discussed. A divergence compensation approach is outlined which calculates the corrected dwell time in order to allow for the divergence caused by the FIB milling process; this is due to overlap effects and the angular-dependent sputter yield. A multi-pass scanning method is used to reduce the fabrication divergence precipitated by atom redeposition. Microstructures, such as parabolic, hemispherical and sinusoidal shapes and a nano hemispherical structure have been produced during the FIB milling experiments using conventional bitmap milling and proposed divergence compensation approaches. A flat top/bottom surface is obtained in convex/concave structures when using the conventional bitmap FIB milling approach. Further research shows that the reasons for this phenomenon are mainly related to both the aspect ratio of the structures and the existence of an oxide layer on the substrate surface. The flat top/bottom phenomenon can be removed by using the combination of a novel divergence compensation approach and the removal of the oxide layer from the substrate surface prior to FIB machining. The experimental results show that the surface form accuracy has been greatly improved by using this method and the overlap effect can be suppressed by carefully choosing the normalized pixel spacing.

DIVERGENCE EFFECTS IN A LAYERED EARTH

Of the various factors which influence reflection amplitudes in a seismic recording, divergence effects are possibly of least direct interest to the interpreter. Nevertheless, proper compensation for these effects is mandatory if reflection amplitudes are to be of diagnostic value. For an earth model consisting of horizontal, isotropic layers, and assuming a point source, we apply ray theory to determine an expression for amplitude correction factors in terms of initial incidence, source‐receiver offset, and reflector depth. The special case of zero offset yields an expression in terms of two‐way traveltime, velocity in the initial layer, and the time‐weighted rms velocity which characterizes reflections. For this model it follows that information which is needed for divergence compensation in the region of normal incidence is available from the customary analysis of normal moveout (NMO). It is hardly surprising that NMO and divergence effects are intimately related when one considers the expanding wavefront situation which is responsible for both phenomena. However, it is evident that an amplitude correction which is appropriate for the primary reflection sequence cannot in general be appropriate for the multiples. At short offset distances the disparity in displayed amplitude varies as the square of the ratio of primary to multiple rms velocities, and favors the multiples. These observations are relevant to a number of concepts which are founded upon plane‐wave theory, notably multiple attenuation processes and record synthesis inclusive of multiples.