Automatic Mining Method of Medical Three-Dimensional Force Sensor Data Under Multi-Stage Redundant Strong Interference
Targeting at the problem that medical 3D force sensor is easily affected by external environment such as electromagnetic field,which produces a large number of similar characteristic data,resulting in the output of disordered signals and reducing the control accuracy and measurement speed of the sensor,a data mining method of 3D force sensor under multi-level redundancy and strong inter-ference is proposed.Redundant data of the three-dimensional force sensor are acquired according to the angle calibration theory.The similarity index function is introduced to calculate the redundancy factor,and the activity of redundant data of three-dimensional force sensor is obtained to complete the data redundancy classification.High performance of filtering redundant data of three-dimensional force sensor is realized by using difference denoising algorithm.Spectral clustering algorithm is used to construct Laplacian matrix,elim-inating redundant data,and realizing automatic data mining of three-dimensional force sensor.The simulation results show that the con-trol accuracy of the proposed method is 96.54％,the measurement speed is 0.61 ms,and the energy consumption is 0.26 kcal.It is proved that the proposed method has high control precision,fast measurement speed and excellent transmission effect,and can meet the demand of force feedback control in robot-assisted surgery.
three-dimensional force sensorredundant datadata miningangle calibrationexponential functiondifference denoisingspectral clustering algorithm
万方数据
维普
