Research On The Analysis Method About Subdivision Error Of Photoelectric Encoder Based On Non-uniform Sampling

With the rapid development of industrial products, military equipment and aerospace technology, the photoelectric encoder which is applicable to different work environment had become the research hotspot. Angle measuring accuracy and resolution of the encoder were put forward higher requirements. The subdivision error of moire fringe signal is the main factor that affect accuracy of the photoelectric encoder.Therefore, developing to deeply analyse the precise signal of encoder with high precision, exploring the source of subdivision error, achieving the dynamic accuracy of real-time monitoring, on-line correcting signal according to the result of detection accuracy that improve the encoder performance index and test accuracy and promote the wider application of photoelectric encoder in different field. It has positive significance.Based on reference of domestic and foreign literature, the principle of moire fringe signal of photoelectric encoder was firstly researched. Precision moire fringe signal equation is established according to the actual work of the encoder. The affection of the encoder subdivision error about the quality of moire fringe was analyzed through equation. Using the Fourier algorithm derived parameter calculation formula of moire fringe signal and solved the dynamic subdivision error of encoder in working. Thus the moire fringe signal correction method is proposed.To improve measurement precision of the moire fringe photoelectric signal of encoder, the analysis method was proposed based on the non-uniform sampling moire fringe photoelectric signal in view of the work state of non-uniform rotation.The fixed sampling frequency was unlimited. Based on photoelectric signal of encoder, the precise code signal of encoder with different rotational speed was reconstructed in using non-uniform sampling point by curve fitting least square method. Reconstruction error was analyzed. This method overcame the problem of the uniform sampling method in the strict requirements of the sampling frequency. It also reduced the amount of calculation and got real photoelectric encoder of the original signal.Real moire fringe signal was got after reconstruction. Parameters calculation method of moire fringe signal based on discrete Fourier transform algorithm was proposed by the ideal sampling point data points of uniform sampling. The mathematical model was established based on the actual signal. The equation of signal parameters such as frequency, amplitude and phase was deduced in using of Fourier transform method. The dynamic subdivision error of encoder in working was calculated by the relation between moire fringe signal parameters and subdivision error. The precision of real-time measurement was accomplished at work. The dynamic detection accuracy of subdivision error of encoder with high precision was improved.The signal correction error method based on genetic algorithm was presented in allusion to real moire fringe signal of encoder. To improve the algorithm optimization, shorten the operation time and enhance the global search ability, the basic genetic algorithm was improved. Using the improved genetic algorithm corrected error of the actual signal. The moire fringe signal was close to ideal. The precision and resolution of the encoder was improved. The foundation was laid for the signal precision real-time correction of high-precision encoder.The analysis system of subdivision error in high precision encoder was established according to theoretic algorithm. Optical polyhedron with 17 surface andCCD biaxial autocollimator in accuracy of 0.1" were detection device. The pretreatment circuit and core processing circuit were set. The graphical user interface and data processing algorithm were designed in the computer and microprocessor. It accomplished the static and dynamic subdivision error of encoder with high precision to be measured and analyzed.The signal of 21-bit absolute photoelectric encoder was analyzed and processed in using of the method in this paper. The proposed algorithm about sampling frequency in this paper is not strictly required. The sampling data points could really reflect the original signal refer to different speed of encoder. The sampling frequency was set about 100 KHz. The maximum relative error of reconstructing signal in precise code was in the scope of(0.2%, 0.2%). The standard deviation of dynamic subdivision error in the encoder which was actual testing was ±0.7 ".Then the standard deviation of dynamic subdivision error in the encoder was ±0.45" in revising moire fringe signal by the improved genetic algorithm. The experimental results show that the non-uniform sampling analysis method in this paper could effectively improve measurement accuracy of the subdivision error in the photoelectric encoder. The correction of signal with the improved genetic algorithm was feasibility and accuracy. Design of the subdivision error analysis system could be used as accuracy detection device about the encoder at the actual job. It had great significance for developing the photoelectric encoder with high precision and resolution.