带触觉的肌电假手握力模糊控制方法

Fuzzy logic based grip force control method for myoelectric prosthesis with tactile feedback

从注重残疾人本体感出发,提出了一种新的用于肌电假手握力控制的方法。该方法利用短时傅里叶变换思想对信号进行时间片分割,并对时间片内的信号分析选取了几个具有代表性的肌电信号参数,并用训练后的神经网络对其进行分析处理,进而估计出相应的肌肉用力大小。同时将肌电假手指端的力触觉传感器所得的力触觉信号和肌肉用力大小信号输入模糊控制器,根据设计好的模糊控制规则调节肌电假手驱动电机的转动速度从而控制握力大小。该方法可以有效减小个人差异性和肌电信号传感器检测位置不同对控制效果的影响,可以减小肌电传感器安装难度并可供不同的残疾人使用。实验结果证明了该握力控制方法的有效性。

In light of the feeling and experience of the disabled, this paper proposes a new grip force control method for myoelectric prosthesis. According to the theory of STFT, the whole myoelectric signal is divided into several time segments. Some representative parameters of the surface electromyography （SEMG） signal in these time segments are selected, which are analyzed and processed with a trained neural network, and then the muscle force strength is esti- mated. The tactile information obtained from the tactile sensor at the prosthesis finger end and the muscle force strength signal are inputted into the fuzzy logic controller, then the motor speed of the myoelectric prosthesis is adjus- ted according to the designed fuzzy logic rules so that the grip force can be adjusted. This method could significantly minimize the influences of individual difference of the SEMG signal and the detection position of the SEMG sensors on the control effects, and decrease the installation difficulty of the myoelectric sensor, which can be used by non-specific disabled. Experiment results prove the effectiveness of the proposed grip force control method.