One of the challenges in motor imagery (MI) classification tasks is finding an easy-handled electroencephalogram (EEG) representation method which can preserve not only temporal features but also spatial ones. To fully utilize the features on various dimensions of EEG, a novel MI classification framework is first introduced in this paper, including a new 3D representation of EEG, a multi-branch 3D convolutional neural network (3D CNN) and the corresponding classification strategy. The 3D representation is generated by transforming EEG signals into a sequence of 2D array which preserves spatial distribution of sampling electrodes. The multibranch 3D CNN and classification strategy are designed accordingly for the 3D representation. Experimental evaluation reveals that the proposed framework reaches state-of-the-art classification kappa value level and significantly outperforms other algorithms by 50% decrease in standard deviation of different subjects, which shows good performance and excellent robustness on different subjects. The framework also shows great performance with only nine sampling electrodes, which can significantly enhance its practicality. Moreover, the multi-branch structure exhibits its low latency and a strong ability in mitigating overfitting issues which often occur in MI classification because of the small training dataset.

In chest and abdomen robotic radiosurgery, due to the motion delay of the robotic manipulator, the tumor position tracking process has a period of delay. This delay ultimately affects the accuracy of radiosurgery treatment. To address the influence of the delay in robotic radiosurgery, a Long-and-Short-Term Memory (LSTM) network as a deep Recurrent Neural Network (RNN) has been applied in a prediction network model for respiratory motion tracking in recent years. However, patients' respiratory state may change in the process of treatment, which may influence the accuracy of prediction. Therefore, it is necessary to update the prediction network through additional data, such as the actual position of the tumor obtained by X-ray imaging. However, the LSTM network has a long update time, and it may not be able to complete the prediction model update in a cycle of X-ray acquisition. To solve this problem, a fast prediction model based on Bidirectional Gated Recurrent Unit (Bi-GRU), is proposed in this paper. This method can reduce the average updating time of the network model by 30%.

Topology optimization is a widely used lightweight design method for structural design of the collaborative robot. In this article, a topology optimization method for the robot lightweight design is proposed based on finite element analysis of the assembly so as to get the minimized weight and to avoid the stress analysis distortion phenomenon that compared the conventional topology optimization method by adding equivalent confining forces at the analyzed part’s boundary. For this method, the stress and deformation of the robot’s parts are calculated based on the finite element analysis of the assembly model. Then, the structure of the parts is redesigned with the goal of minimized mass and the constraint of maximum displacement of the robot’s end by topology optimization. The proposed method has the advantages of a better lightweight effect compared with the conventional one, which is demonstrated by a simple two-linkage robot lightweight design. Finally, the method is applied on a 5 degree of freedom upper-limb exoskeleton robot for lightweight design. Results show that there is a 10.4% reduction of the mass compared with the conventional method.

As an emerging multifunctional material, Gallium based room temperature liquid metal has attracted a lot of attention for a variety of applications due to its mobility and deformability. However, controlling the motion of a liquid metal droplet accurately still remains unrevealed, which restricts its application in many fields. In this paper we propose a hybrid framework that would control the motion of a liquid metal droplet in a one-dimensional fluidic channel. A dynamic model of a liquid metal droplet immersed in the electrolyte when an electrical field is applied to each end of the channel is discussed first, followed by a setpoint controller designed to calculate the current input needed to drive the liquid metal droplet to its destination with vision feedback. To obtain the desired high-resolution current output, a fast and high-resolution current output power supply will be established by integrating a fast PID controller and a simple programmable DC power supply. The effectiveness of this proposed approach will be verified by controlling a liquid metal droplet so that it reaches its destination inside the PMMA channel. This proposed approach may lead to the development of tiny soft robots, or micro-fluidic systems that can be driven accurately by liquid metal droplets.