Voice Control of a Robotic Arm for Hysterectomy and Its Optimal Pivot Selection

This paper presents a method to recognize the voice command which is using for control a rbototic arm for hysterectomy. We extract MFCCs (Mel Frequency Cepstrum Coefficients) characteristic parameters as the original input, then put it into the CNNs (Convolutional Neural Networks) model after specific processing. After obtain the speech recognition model, we input the voice of command generate by a operator and then it would predicted a voice command and take corresponding action on robot. The plantform we used to verify our model is a 6-DOF manipulator. In order to promote maneuverability of this robot, we adopt a method to optimize the selection of Remote Center of Motion (RCM). Experiments show that this speech recognition meodel based on CNNs is fulfill the requirment of surgery and controling robot by its command is feasible.