Research on control strategies for rehabilitation robots has gradually shifted from providing therapies with fixed, relatively stiff assistance to compelling alternatives with assistance or challenge strategies to maximize subject participation. These alternative control strategies can promote neural plasticity and, in turn, increase the potential for recovery of motor coordination. In this paper, we propose a control strategy that dynamically switches between assistance and challenge modes based on the user's performance by amplifying or reducing the deviation between the user and the rehabilitation robot. For a seamless cognitive and physical interaction between the robot and the patient, we propose a multisensor fusion system to provide accurate activity and motor capability recognition, fall detection and physical fitness assessment in the rehabilitation training process. Moreover, an adaptive radial basis function (RBF) neural sliding-mode (ARNNSM) controller that dominates a mobile chaperonage lower limb rehabilitation training robot is proposed. The controller employs asynchronous deviation and functional assessment to determine the subject's capabilities and computes a corresponding assistance torque with a challenging factor for the desired locomotive function. Some sufficient conditions are derived based on algebraic graph theory and Lyapunov theory to ensure the asymptotic stability of the systems. Simulation examples illustrate the effectiveness of the proposed controllers. The ARNNSM controller and accompanying algorithm are demonstrated experimentally with healthy subjects in a new type of mobile chaperonage lower limb rehabilitation robot.

Robot-aided training strategies that allow functional, assist-as-needed, or challenging training have been widely explored. Accurate activity recognition is the basis for implementing alternative training strategies. However, some obstacles to accurate recognition exist. First, scientists do not yet fully understand some rehabilitation activities, such as abnormal gaits and falls; thus, there is no standardized feature for identifying such activities. Second, during the activity identification process, it is difficult to reasonably balance sensitivity and specificity when setting the threshold. Therefore, we proposed a multisensor fusion system and a two-stage activity recognition classifier. This multisensor system integrates explicit information such as kinematics and spatial distribution information along with implicit information such as kinetics and pulse information. Both the explicit and implicit information are analyzed in one discriminant function to obtain a detailed and accurate recognition result. Then, alternative training strategies can be implemented on this basis. Finally, we conducted experiments to verify the feasibility and efficiency of the multisensor fusion system. The experimental results show that the proposed fusion system achieves an accuracy of 99.37%, and the time required to prejudge a fall is approximately 205 milliseconds faster than the response time of single-sensor systems. Moreover, the proposed system also identifies fall directions and abnormal gait types.

Currently, the most widely adopted fall prediction methods use kinematic sensors for data collection and set thresholds to identify a fall based on artificial experience or neural network learning algorithms. These methods require classification and calibration in advance. Striking a balance between sensitivity and specificity is difficult when setting thresholds. Backpropagation (BP), support vector machines (SVMs) and other neural network algorithm models require professional expertise and are too complicated, difficult to understand, and computationally expensive, which renders embedded development with these models difficult. In response to these problems, this paper builds a fuzzy logic-enhanced adaptive neural network classifier (FLEANN); uses the prior knowledge of experts to construct fuzzy rules; employs neural networks to drive the underlying data to confirm and optimize logical rules; and utilizes an adaptive fuzzy neural network. The parameters of the algorithm are modified and optimized using a large amount of data training to improve the heuristics, transparency, and robustness of the neural network. The enhanced adaptive fuzzy neural network is applied in rehabilitation training for the classification of imprecise, uncertain, and nonlinear falls; abnormal gait; and normal gait. The experimental results confirm that the accuracy of this classifier is better than that of conventional neural network classifiers and that its logical structure is clear, which enables its application to embedded development. The reduction in computational demand enables the elimination of high-configuration PCs and reliance on a single chip microcomputer.

Converting customer needs into specific forms and providing consumers with services are crucial in product design. Currently, conversion is no longer difficult due to the development of modern technology, and various measures can be applied for product realization, thus increasing the complexity of analysis and evaluation in the design process. The focus of the design process has thus shifted from problem solving to minimizing the total amount of information content. This paper presents a New Hybrid Axiomatic Design (AD) Methodology based on iteratively matching and merging design parameters that meet the independence axiom and attribute constraints by applying trimming technology, the ideal final results, and technology evolution theory. The proposed method minimizes the total amount of information content and improves the design quality. Finally, a case study of a rehabilitation robot design for hemiplegic patients is presented. The results indicate that the iterative matching and merging of related attributes can minimize the total amount of information content, reduce the cost, and improve design efficiency. Additionally, evolutionary technology prediction can ensure product novelty and improve market competitiveness. The methodology provides an excellent way to design a new (or improved) product.

Fault detection of UAVs is one of the main technical means to ensure the normal flight of UAVs. Traditional methods are based on a single model for anomaly detection, but a single model cannot learn the features of UAVs from multiple perspectives. Multi-model Neural Networks can learn features of the data from multiple perspectives. Inspired by this, a Multimodal Model Fusion (MMF) based UAV fault detection model is proposed in this paper. The MMF model consists of three parts: firstly, a Multilayer Perceptron (MLP) and a Convolutional Neural Network (CNN) are used to obtain nonlinear features and key features through parallel computation of homologous UAV data; secondly, key features and nonlinear features are merged to obtain a fusion vector; finally, a fully connected layer is used to compute the fusion vector, and then the results are output by a Softmax classifier. The experimental results show that the accuracy of the MMF model is higher than that of the single-model by about 2% to 7% on the ALFA dataset.