Weak Signal Detection of Photoelectric Sensor Based on Extreme Learning Machine
Download as PDF
Li Jianfang, Zhou Yanmin
In order to improve the accuracy of automatic detection of weak signal of photoelectric sensor, an automatic detection method of weak signal of photoelectric sensor based on neural network is proposed. Firstly, the weak signal sequence of photoelectric sensor is collected and reconstructed in phase space by mutual information method and false neighbor point method. Then, the weak signal sequence of photoelectric sensor is modeled and detected by limit learning machine. Finally, the weak signal data of photoelectric sensor is used for simulation experiment. Compared with other methods, the results show that this method can accurately reflect the chaotic characteristics of weak signal of photoelectric sensor, restrain the interference and influence of noise on useful signal, and greatly improve the detection accuracy. It can be applied to the automatic detection of weak signal of photoelectric sensor.
Photoelectric Sensor, Weak Signal Detection, Extreme Learning Machine, Chaos Analysis