Biomimetic and porous nanofiber-based hybrid sensor for multifunctional pressure sensing and human gesture identification via deep learning method

Near-field electrospinning (NFES) is a site addressable microfabrication process and is utilized to deposit the micro/nano polyvinylidene fluoride (PVDF) fibers arrays on printed circuit board nanofiber-based based piezoelectric sensor architectures. In addition, a biomimetic and flexible hybrid self-powered sensors (BHSS) was created by hybridizing both Cu - biomimetic Polydimethylsiloxane triboelectric sensors to enhance the energy-harvesting characteristic. The optimized BHSS had open-circuit voltage (V_OC) of 15 V and 115 nA of short-circuit current (I_SC) and a maximum average power density is 675 μW m⁻² with a load of 10 MΩ. Furthermore, an intelligent glove and the force sensor with are successively confirmed that the developed BHSS has promising applications in wearable self-power sensor technology. The machine learning algorithm of Long Short-Term Memory (LSTM) in the context of gesture recognition was used and effectively distinguish five human actions satisfactorily. LSTM based real-time electrical signals of five gestures dataset with varying duration and complexity can achieve an overall classification rate of 82.3%.