Model and design of magnetostrictive flexible sensor for gesture recognition based on curvature and magnetism fusion

By employing traditional single-mode sensors, such as bending strain and magnetic fields, it is challenging to fully perceive the attitude and position of the human hand, thus reducing the accuracy of gesture recognition and hindering human–computer interaction. Based on force-magnetic sensitivity of electrodeposited magnetostrictive films, we present bi-perceptive flexible sensor which can detect the changes in magnetization state under the curvature radius/magnetic field simultaneously. The second harmonic frequency signal model of sensor is established under non-adiabatic conditions based on the Euler-Bernoulli beam theory, the Jiles-Atherton model, the Boltzmann statistics and the electromagnetic theory. With the guidance of the experimentally verified model, we optimize the excitation parameters, structural measurements and material properties and design a local signal processing system for measuring the curvature radius and magnetic field signals under low power consumption. The maximal sensitivity of the sensor is 22 mV/mm in the curvature radius of 10–65 mm. The maximal sensitivity of the sensor is 1.78 mV/(kA/m) in the magnetic field of 1–11 kA/m. The human–computer interaction platform for collecting gesture signals is built based on new bi-perceptive flexible sensors. The 12 hand gestures, confused with traditional stress/strain sensors, are classified and recognized with 90.1 % accuracy with the ANFIS classifier.