No abstract

Rich and accurate tactile perceptive capability is important for both humans and robots. Soft materials exhibit unique characteristics to construct high‐performance tactile sensors such as high sensitivity and high resistance to overload. In this work, a multiaxis tactile sensor based on a soft anisotropic waveguide that can distinguish normal force and shear force, which can greatly expand its potential uses in practice, is reported. First, the anisotropy of the waveguide sensor's response to vector forces is validated numerically and experimentally, and then two of those anisotropic units are embedded into one cladding with a crossed‐over layout, to form a multiaxis sensor. Then, a calibration algorithm is implemented on this sensor and the reconstruction of vector forces is achieved at an average accuracy of 28.0 mN for normal forces and 81.1 mN for shear forces, both with the sensing range of 1 N. Using this device, three demonstrations are shown to give outlooks of its potential application in human and robotic grasping tasks: a wearable tactile sensor for collecting human's haptic data during operation, a uniaxial gyroscope, and a robotic gripper's tactile sensor for assisting an unlocking task with a key. This work is a step toward more functional soft waveguide‐based force sensors.

No abstract

No abstract

As electromagnetic environments are increasingly complex, there are more kinds of radar jamming signals. Active jamming recognition has problems of the low recognition accuracy and the high computational complexity, especially under a low jamming‐to‐noise ratio (JNR). Herein, a deep learning network based on bilinear EfficientNet and attention mechanism is proposed to recognise and classify eight kinds of jamming signals automatically. Firstly, the one‐dimensional interference signal is transformed into a two‐dimensional time–frequency image by the time–frequency analysis. Based on the transfer learning of EfficientNet‐B3, the effective features of the time–frequency image are automatically extracted by the two‐way network with attention mechanism, and the active interference classification is realised. The experimental results show that the method's overall recognition rate for eight kinds of interference signals is more than 97.5% when the JNR is −8 dB and is close to 100% at −2 dB.