Fatigue recognition of crane operators based on sEMG signals

Malfunctional operation resulted from fatigue of operators is the primary reason of cranes' accidents, therefore fatigue recognition of crane operators is becoming an urgent request in present research. Based on biomechanical theory, fatigue growth principle of crane operators was studied to ascertain characterization parameters of fatigue. Originated at human-machine-environment multidisciplinary terrain, surface electromyogram (sEMG) was integrated to execute online measurement of crane operators. Wavelet package-Sample Entropy based energy ratio was adopted to extract features of sEMG signals, since they changed in a stochastic process with nonlinear, non-stable and chaotic characteristics. Spiking neural network (SNN) was proposed to classify features into fatigue and non-fatigue classes. Therefore, fatigue recognition of crane operators was established, which could be applied widely in intelligent prediction and monitoring of operators' fatigue in practical engineering.