Search

Natural control methods based on surface electromyography (sEMG) and pattern recognization are promising for hand prosthetics. However, the control robustness offered by scientific research is still not sufficient for many activities of daily living (ADLs). Difficulty results from limited sEMG signals susceptible to interference in clinical practice, it needs to synthesize hand movement and sEMG to improve classification robustness. Human hand ADLs are made of complex sequences of finger joint movements, and capturing the temporal dynamics is fundamental for successful hand prosthetics control. Current research suggests that recurrent neural networks (RNN) are suited to automate feature extraction for time series domains, and dynamic movement primitives (DMP) can provide representation of hand kinematic primitives. We design a multimodal deep framework for inter-subject ADLs recognization, which: (i) implements heterogeneous sensors fusion; (ii) does not require hand-crafted features; and (iii) contains the dynamics model of the hand ADLs task control. We evaluate our framework on Ninapro datasets. Results show that our framework outperforms competing deep networks with single modal and some of the previous reported results.

When carrying out special services of rescue and firefighting in buildings, underground works, tunnels and mine, where has no GPS signal, it's important to get location information of special staff, which used by guiding rescue, evacuation and force dispatch. In order to meet the above requirements, this system uses the inertial measurement unit constructed by 3-axis acceleration, 3-axis gyroscope and 3-axis geomagnetic sensor to complete the measurement of pedestrian motion information, calculate the personnel attitude through the cosine matrix, and use the extended Kalman filter the calculation of the speed and position of the personnel is carried out, and the zero speed detection is realized by the gyroscope to reduce the accumulated error of the calculation, thus realizing the key technology of inertial navigation. Using ZigBee and WIFI modules, the development of portable wireless network nodes, to build a wireless self-organizing network. Using embedded microcontrollers and MEMS inertial sensors to build a wearable wireless navigation and positioning system. Using wireless ad hoc network, implement real-time online monitoring of the staff positions and movement trajectory. The wearable wireless navigation and positioning system can accurately obtain the position and movement trajectory of the rescue personnel. Using the wireless ad hoc network, the position and the trajectory information of the rescue personnel can be transmitted to the control center in real time to provide information for the disaster relief command and decision support service.

A millimeter-wave beam multiplexing method using a hybrid beamforming is proposed. We theoretically show that inter-subarray coding at the subarray type interleaved configuration can reduce interference and multiple beams of the same high gain as the full connection type. The condition that a hybrid beamforming of the subarray type can create the multiple high gain beams is clarified. For the case of beam multiplexing, the interleaved configuration with inter-subarray coding is confirmed to be robust for interference increasing by errors of beam direction estimation. Therefore, we think that the interleaved configuration is suitable for millimeter-wave beam multiplexing.

The problem of group decision making toward expressing preference relation in linguistic labels in fuzzy decision system, constructs deviation function by transform function , make use of the idea of optimization, builds optimal model, an optimal algorithm of group decision making based on linguistic preference information is proposed, called "least variance method". Calculate simply and manipulate easily are features of this method, and finally, an example is given to show the efficiency and practicality of this method.