Research On Dual-coil-based Wireless Torque Sensor And Automatic Testing System

In recent years, intelligent bicycles are becoming more and more popular. They make riding more efficient by a core function called ride-state feedback. Torque sensor is the key component to detect the torque on the mechanical parts. Compared with speed sensors and acceleration sensors, the torque sensor is better on the feedback of ride-state due to its high precision, fast frequency response, good reliability and long service life. And it is widely used in motor-assisted bicycles which are exported to Europe in recent years.Among many the torque measurement methods, the torque sensor depended on metal resistance strain gauge has got ripe development with advantages such as excellent measurement accuracy and wide range of application in the field of torque measurement. In order to have better work results, in this thesis, torque sensor is designed by dual-coil for power supply and signal transmission. It applies electromagnetically coupled resonant wireless power transmission mode to solve the problem that the torque sensor rotates continuously around axis while it can’t be connected by power supply cable. It uses the mutual inductance coil which transmits radio frequency signal wirelessly at the same time. Experiments showed that the sensor achieved a good performance to complete the signal detection.To ensure the qualified output performance of the torque sensor, an automatic testing platform based on torque sensor’s working environment is designed independently in this thesis. The platform simulates the environment of the torque sensor in mid-assisted bicycle system. It also analyzes and calibrates the driving force of assisting bicycles’, and takes them over equivalently in the test system instead of the actual test. In the testing process, servo motor simulates the force exerted by driving gear and the chain. The magnetic powder brake provides the resistance, and the standard torque speed sensor is used to provide tested speed information for comparison. All slave computer devices are integrated in the LabVIEW platform software on a master PC, which collects the signal and communicates via PCI-6229 cards and VISA I/O bus. The test system realizes the durability test and heat-sensitive test, which provides a new method for the torque sensor’s test.In addition, the thesis also builds up a torque sensor model, i.e. performance assessment model(PAM). This model can evaluate accurately the output performance of the torque sensor, and it can predict the test results of torque sensor. The model can simulate and predict performance curves of the torque sensors under different temperature and time duration. Tested by experiments, it has been verified that the prediction model has high prediction accuracy and prediction stability, and its test error is between 2% and 5%. The system helps to reduce the test cost and provides further benefits in optimizing the design of the torque sensor. It has a very important reference for the test of the torque sensor.