The Application Of Fuzzy Technology In Photoelectric Detection

Based on the theory of fuzzy techniques and adaptive neuro-fuzzy inference system (ANFIS) technique, this paper analyzed the main factors that what influenced the precision of photoelectric detection and error compensation method. According to complex, time-varying and uncertain control systems, the paper introduced a kind technique of adaptive neural-fuzzy inference systems, and with testing the light illuminance of photodiodes as compensation object, through the data acquisition and data-processing, studied the key technologies of this kind of error compensation method, and a good result is obtained.Through the analysis of the photoelectric detection circuit's structure and the affecting factors, set the influence factors that photo-current, and temperature as inputs, and set illumination as output, established a double input and single output of two-dimensional adaptive neuro-fuzzy inference system, and taking the data of collection of photoelectric detection experiments as sample data, based on the training sample data, the system will automatically generate fuzzy control rules and automatically adjust the membership function parameters, achieve optimal system structure. The final output data can be good approximation sample data, it proves that the feasibility of this method.In research common error compensation method based on the experimental data, dealt with the data with common error compensation method, comparatively analyzed the processing result with results of adaptive neuro-fuzzy inference system (ANFIS) technique, it is concluded that the adaptive neuro-fuzzy inference system method possessed more advantage, through experiment of illumination detection to test the effectiveness of the method, and the results show that the RMS error of this method is very small, it can approach sample data. Therefore, this method is feasible and effective that used in the photoelectric detection circuit of error compensation.