In order to more effectively improve the optical current transformer (OCT) signal noise ratio

with focus on the noise characteristics of OCT

a method of OCT noise analysis based on Allan variance is proposed. With this method and by referring to the structure of noise and noise properties of OCT

noise can be divided into six types: quantized noise

white Gaussian noise

1/f noise

index related noise

temperature noise and sinusoidal noise. Thus an OCT random noise model containing six type noise terms is obtained. Calculate and identify the noise factor of all the main noise sources in the OCT random noise model from the measured data of OCT for laboratory. Experiments show that this method is able to effectively identify each noise coefficient of OCT

laying a foundation for OCT using the real-time filtering algorithm to improve measuring precision. This work is supported by National Natural Science Foundation of China (No. 51277066).