With the acceleration of intelligent substation construction, large-scale intelligent substation construction and expansion projects have been put into operation. In this context, due to the differences in experience and habits between designers and operators, it is difficult to standardize the design of optical fiber circuit in smart substation. The design and correctness check of optical fiber circuit are completed manually, the design efficiency is low and the correctness check of drawings is difficult. To solve the above problems, an automatic design scheme of optical fiber circuit in intelligent substation based on neural network algorithm is proposed. By collecting a large number of physical configuration description files (SPCD files) of substations and learning from artificial intelligence, the decision tree of neural network is established to automatically output the design drawings of optical fiber circuit according to the requirements of secondary system design. This scheme improves the efficiency and quality of the design of optical fiber circuit in intelligent substation by using artificial intelligence technology, and then realizes the standardization and intellectualization of the design of optical fiber circuit. This work is supported by Youth Fund of National Natural Science Foundation of China (No. 61602251) and Research Project of State Grid Jibei Electric Power Co., Ltd. (No. 52010117000G).