Research On The Detection Method Of Photovoltaic Silicon Wafer Thickness And Line Mark Defect

Solar energy green environmental protection,is one of the new energy with very broad application prospects.China’s photovoltaic cell industry is developing rapidly,and under the market competition,the requirements of photovoltaic cell production enterprises for product quality are gradually improving.On the photovoltaic cell production line,the quality of the original silicon wafer directly affects the processing and manufacturing of the subsequent processes.There are many defects of the original silicon wafer,and the thickness and line trace are the important factors affecting the quality of the silicon wafer.Because the precision unit of the line trace is microns,it is difficult to identify manually.It is important to study the thickness of photovoltaic silicon wafer and improve the product quality.First,the detection requirements of silicon wafer thickness and line trace defects are analyzed,and the overall detection scheme is designed.In terms of hardware,a hardware selection test was carried out,and a testing system was set up to ensure the smooth progress of testing.In terms of software algorithm,the surface morphology data of the surface of silicon wafer was designed.Secondly,to improve the accuracy of the data,a quantifiable compensation model is established for the inclination of the measured object.To solve the problem of poor sensor consistency,using the least squares method improves the consistency of the six sensors and further improves the detection accuracy of the system.In addition,other factors affecting the detection accuracy are analyzed and corresponding solutions are given.The data processing method adapted to the detection project is studied to facilitate the subsequent algorithm calculation.Then,the detection algorithm of thickness and line trace defects is proposed.In the thickness measurement part,the thickness measurement formula with corrective variables is established according to the principle of radiation thickness measurement,which improves the accuracy of thickness measurement.In the detection of line trace defects,aiming at the data fluctuation caused by interference factors,the line trace preprocessing algorithm based on Gaussian filter is put forward to eliminate the interference factors,and the real morphology data of the silicon wafer surface is obtained,and the real height and width of the line trace are calculated combined with the zero point theorem to complete the detection of the line trace defect.Finally,to verify the feasibility and accuracy of the present detection method,the detection software was developed on the vs.platform.According to the experimental results,the detection method of photovoltaic silicon wafer thickness detection accuracy reaches3.8um,the repeat accuracy reaches 2um,the detection accuracy of line trace defect reaches96.3%,and the average detection time of each silicon wafer is 0.9s.The experimental data prove that the proposed detection method is feasible,efficient and accurate,and can be used for the industrial detection of photovoltaic silicon wafers.