This paper investigates the fluidic leak rate through sealing contact surfaces by comparison between model calculation and experiment measurement. The focus is on an experimental device designed to measure the leak rate of the static seals with a simpler structure, smaller errors, stronger stability, and more functions. Using the device, experiments were carried out to four test pieces with different surface characteristics, whose leak rates were measured separately. Compared with the calculation results obtained from the fractal surface leak rate prediction model, the correctness and the application range of the model were verified, and the effects of different surface topographies and material properties of the four test pieces on the leak rate were analyzed as well. The experimental device was also used to perform single-factor comparison experiments, which were then combined with the theoretical prediction model to analyze the effects of the sealing surface contact load, fluid pressure, and surface apparent size on the leak rate, so that theoretical support and experimental evidence for selecting the parameters of sealing device was provided.

This paper investigates the fluidic leak rate through metal sealing surfaces by developing fractal models for the contact process and leakage process. An improved model is established to describe the seal-contact interface of two metal rough surface. The contact model divides the deformed regions by classifying the asperities of different characteristic lengths into the elastic, elastic-plastic and plastic regimes. Using the improved contact model, the leakage channel under the contact surface is mathematically modeled based on the fractal theory. The leakage model obtains the leak rate using the fluid transport theory in porous media, considering that the pores-forming percolation channels can be treated as a combination of filled tortuous capillaries. The effects of fractal structure, surface material and gasket size on the contact process and leakage process are analyzed through numerical simulations for sealed ring gaskets.

Depth completion of forward objects is critical to safe and effective autonomous driving as it can estimate dense-depth from sparse LiDAR (i.e., Light Detection and Ranging) data and RGB images. The study proposes a self-supervised depth completion method to address the deficiency in depth sensing images caused by the uneven distribution of point-clouds generated by long-range LiDAR. The proposed method uses a stacked-hourglass-network structure as the backbone, which allows to perform multi-level deep fusion to extract features from uneven point-clouds. In addition, a fusion module is introduced to fuse global-local features from different stages of the network output, and a pooling guidance mechanism is designed to guide network convergence through adaptive pooling guidance depth construction. Field experiments were conducted on the Tsuen Wan Line in Hong Kong to verify the performance of the proposed network. The results indicate that the proposed network outperforms the benchmark self-supervised depth completion methods, with a root mean squared error (RMSE) of 8165.673mm and a mean absolute error (MAE) of 2596.418mm. Specifically, the RMSE and MAE were improved by 4.7% and 2.4%, respectively, compared to the benchmark methods.

Moving target indication (MTI) based on space–time adaptive processing (STAP) has been widely used in airborne radar due to its ability for clutter suppression performance. However, the existing MTI methods suffer from the problems of insufficient training samples and low detection probability in a non-homogeneous clutter environment. To address these issues, this paper proposes a novel deep learning framework to improve target indication capability. First, combined with the problems of target indication caused by the non-homogeneous clutter, the clutter-plus-target training dataset was modeled by simulation, where various non-ideal factors, such as aircraft crabbing, array errors and internal clutter motion (ICM), were considered. The dataset considers various realistic situations, making the proposed method more robust. Then, a five-layer two-dimensional convolutional neural network (D²CNN) was designed and applied to learn the clutter and target characteristics distribution. The proposed D²CNN can predict the target with a high resolution to implement an end-to-end moving target indication (ETE-MTI) with a higher detection accuracy. In this D²CNN, the input was obtained by the clutter-plus-target angle-Doppler spectrum with a low-resolution estimated only by a few samples. The label was given by the target angle-Doppler spectrum with a high-resolution obtained by the target’s exact angle and Doppler. Thirdly, the proposed method used a few samples to improve the target indication and detection probability, which solved the problem of insufficient samples in the non-homogeneous clutter environments. To elaborate, the proposed method directly implements ETE-MTI without the support of the conventional STAP algorithm to suppress the clutter. The results verify the validity and the robustness of the proposed ETE-MTI with a few samples in the non-homogeneous and low signal-to-clutter ratio (SCR) environments.

Accurate detection of human motion states and prediction of intentions are crucial for the natural walking assistance provided by lower limb exoskeletons. Several methods for gait task recognition, phase estimation, and motion trajectory prediction have been proposed, but few studies have considered all of these simultaneously. This paper proposes a novel scheme for gait recognition and trajectory prediction, capable of detecting gait events, recognizing multiple gait tasks, estimating gait phases, and predicting trajectories. Firstly, this scheme proposes a gait task recognition method based on gait events and a multilayer perceptron neural network (MLPNN), which reduces computational cost and improves recognition stability. Next, a gait phase estimation method based on resettable symmetric adaptive oscillators (RSAOs) is introduced, which offers faster convergence speed and smaller error compared to traditional adaptive oscillators. Subsequently, a trajectory prediction method with adaptive dynamic motion primitives (ADMP) and feedback regulation is proposed. Finally, gait task recognition, phase estimation, trajectory prediction, and comprehensive online experiments are carried out respectively. The online experimental results show that the recognition accuracy of gait tasks can reach 99.2%, the phase estimation error is less than 0.3652rad, and the trajectory estimation error is less than 1.1814°.

With the development of society, the aging population and the number of stroke patients are increasing year by year. Rehabilitation exoskeletons can help patients carry out rehabilitation training and improve their activities of daily living. First, this paper designs a reconfigurable exoskeleton for upper limb rehabilitation. Second, the working space and singular configuration of the exoskeleton are analyzed. Then, Dynamic Movement Primitives (DMP) and sliding mode control are combined to form a new control strategy. Additionally, by changing the working mode of the gravity compensation device and different control methods, the control experiment of the exoskeleton is carried out. The advantages of sliding mode control under combinational reaching law (CRL-SMC) are verified. The influence of the gravity compensation device on motor driving torque and energy consumption is also analyzed. Finally, experimental results show that compared with sliding mode control under power reaching law (PRL-SMC) and PID control, CRL-SMC has better control performance in single joint trajectory tracking and end trajectory tracking, improving control performance by at least 60 In the best case, the gravity compensation device can reduce the energy consumption by 81.90% and the maximum motor current by 69.25% of the driving element.