This article introduces a wall-climbing robot that uses the reverse thrust of the propeller as the adsorption force. The robot is symmetrically distributed in structure and the adsorption force is symmetrically distributed before and after so that it can adapt to the surface of a variety of different media materials and achieve stable adsorption and movement of a variety of wall surfaces. The robot mainly uses the reverse thrust of the aircraft propeller as the adsorption force to achieve wall adsorption. The robot relies on four wheels to move forward. The forward power mainly comes from the combined action of the propeller reverse thrust component and the front wheel driving force. During the movement of the robot, the steering is realized by the front wheel differential control. In this paper, we design the structure of a dual-propeller dynamic adsorption wall mobile robot, plan the movement process of the robot from the ground to the wall, analyze the stable adsorption conditions of the robot wall, and carry out the robot’s motion performance and adaptability test under different ground/wall environments to verify that the robot is stable and feasible.

Traditional computational fluid dynamics (CFD) techniques deduce the dynamic variations in flow fields by using finite elements or finite differences to solve partial differential equations. CFD usually involves several tens of thousands of grid nodes, which entail long computation times and significant computational resources. Fluid data are usually irregular data, and there will be turbulence in the flow field where the physical quantities between adjacent grid nodes are extremely nonequilibrium. We use a graph attention neural network to build a fluid simulation model (GAFM). GAFM assigns weights to adjacent node-pairs through a graph attention mechanism. In this way, it is not only possible to directly calculate the fluid data but also to adjust for nonequilibrium in vortices, especially turbulent flows. The GAFM deductively predicts the dynamic variations in flow fields by using spatiotemporally continuous sample data. A validation of the proposed GAFM against the two-dimensional (2D) flow around a cylinder confirms its high prediction accuracy. In addition, the GAFM achieves faster computation speeds than traditional CFD solvers by two to three orders of magnitude. The GAFM provides a new idea for the rapid optimization and design of fluid mechanics models and the real-time control of intelligent fluid mechanisms.

In response to the rehabilitation needs of stroke patients who are unable to benefit from conventional rehabilitation due to the COVID-19 epidemic, this paper designs a robot that combines on-site and telerehabilitation. The objective is to assist the patient in walking. We design the electromechanical system with a gantry mechanism, body-weight support system, information feedback system, and man-machine interactive control system. The proposed rehabilitation robot remote system is based on the client/server (C/S) network framework to realize the remote control of the robot state logic and the transmission of patient training data. Based on the proposed system, doctors can set or adjust the training modes and control the parameters of the robot and guide remote patient rehabilitation training through video communication. The robotic system can further store and manage the rehabilitation data of the patient during training. Experiments show the human-computer interaction system of the lower limb rehabilitation robot has good performance, can accurately recognize the information of human motion posture, and achieve the goal of actively the following motion. Experiments confirm the feasibility of the proposed design, the information management of stroke patients, and the efficiency of rehabilitation training. The proposed system can reduce the workload of the doctors in practical training.

Mode decomposition methods, such as proper orthogonal decomposition and dynamic mode decomposition (DMD), have introduced a novel data-driven approach for flow prediction. These methods aim to identify a collection of modes that capture the essential flow features. Subsequently, the flow field data are projected onto these modes to reconstruct and predict the evolution of the flow field. However, due to their inherent linearity, mode decomposition methods are limited in effectively handling unsteady and nonlinear flow exhibiting significant nonlinearities. In this study, we propose a spectral graph convolutional deep neural network (SGC-DNN). It employs the eigenvectors of the Laplacian matrix as modes to fully utilize the adjacency information within the graph structure to solve flow on an unstructured grid better. Additionally, we employ a DNN (deep neural network) to model the temporal evolution of each mode, thereby enhancing the model's adaptability to nonlinear flow fields. To evaluate the performance of our proposed SGC-DNN, we compare its prediction results with those obtained using DMD and DNN for the flow around a cylinder on unstructured grids at various Reynolds numbers (ranging from 1000 to 500 000). We also compared the predictive results of these three models for flow with complex geometries, such as the Da Vinci pipeline flow and intracranial aneurysm blood flow. The comparative analysis demonstrates that SGC-DNN outperformed the other models, yielding lower L2 relative errors and higher R2 values. These outcomes highlight the superiority of SGC-DNN in accurately predicting unsteady and nonlinear flow characterized by graph structures.