Design And Research Of A Novel High-precision Six-axis Force Sensor

Traditional single-axis sensors have failed to meet the needs of modern production and technology.The six-axis force sensor has attracted more and more scholars' attention.The six-axis force sensor can measure the force and torque of six directions simultaneously and is widely used in robot arms,aerospace,biomedical and other fields.However,there are also many problems in the six-axis force sensors at the current stage,and the high-precision and high-sensitivity sensors are desperately need of further research.Combining the theory and methods of calculation and experimental study,this paper proposes a new high precision structure six-axis force sensor and analyzes the structure of the sensor optimization on the basis of the original sensor research.The main research content includes:1.Based on the analysis of the sensors' performance,a new structure of composite beam six-axis force sensor is proposed.The sensor is simple,and the deformation is obvious,the patch position is easy to operate and determine.The optimal size parameters of the sensor are determined by orthogonal test and finite element simulation.2.Abaqus analysis software was used to build the sensor model,and the static analysis and dynamic analysis were carried out on the sensor.The strain variation trend of the sensor under the influence of external load and the natural frequency of the sensor are obtained.The results show that the sensor has a good linearity.3.Based on the principle of Wheatstone bridge,the paper decide how to arrange the strain gauges.With the finite element analysis as the theoretical guidance,the maximum deformation zone of elastomer can be determined.Based on the symmetry of the sensor structure,the reasonable method of arranging strain gauges is determined and the output is increased appropriately.The dimensional coupling problem of the six-axis force sensor is reduced through arranging the strain gauges.4.Development of special calibration device for sensor.According to the structure and shape of the sensor,a special set of sensor calibration device is designed.According to the results of simulation experiments,studying the decoupling of the sensor.Using the least square method and BP neural network method to calculate the relationship between input and output.The research shows that BP neural network method has better decoupling effect.