Research And Design Of The Forearm Prosthesis Control System Based On Android Mobile Phone Voice

This thesis design a forearm prosthesis control system through the voice recognition of Android mobile. Through phone, the users can control the forearm prosthesis to finish lots of actions by voice, for example, keeping the balance of the hand, closing the balance, turning up and down, turning inside and outside, grabbing objects, opening the hand, suspending rotation. This system allows the users to finish eating bread or drinking water and other basic necessities.The system design an offline voice APP for Android mobile phone, which can identify the voice of the users in the environment without network and pass the identification results to the control panel of the forearm prosthesis through the Bluetooth of the mobile phone. Deeply probe into the theory and key technology of the speech recognition and the construction of the Android development environment and the data communication between the Android mobile phone with the control panel of the forearm prosthesis, which need to the development of offline voice APP of the mobile phone. Among them, the offline voice APP of the mobile phone is developed based on a speech recognition engine called PocketSphinx. The related circuits of the control panel of the forearm prosthesis are designed independently.The core chip of the control panel chose STM32F103C8T6.In order to detect whether there is a slip between the hand and the grab target and achieve the capture of eggs and other fragile items, the pressure sensor and the sliding sensor are installed in the hand and the corresponding signal conditioning circuits are designed respectively, so as to realize the prosthetic hand soft fetching function. Among them, the sliding sensor chose PVDF piezoelectric film and the pressure sensor chose A3 01 force sensitive film sensor.In addition, in order to prevent the water outside the cup when the user complete drinking water and other types of action, the jaws of death of the prosthesis hand need to keep parallel with the ground plane. To achieve this function, the accurate posture estimation should be got by the senor called MPU6050. Among them, the quaternion method and the complementary filtering algorithm are used to solve the attitude of the forearm prosthesis.