An Intelligent Digital Secretary for Design of Electric Power Engineering
Yiqi Lu, Jinghai Xie, Songhe Lu, Chuye Hu, Ying Xu, Shaorong Wang · 2020 IEEE 1st China International Youth Conference on Electrical Engineering (CIYCEE) · 2020
In order to improve the digital level of design of electric power engineering and ensure the accuracy and reliability, this paper proposes an intelligent digital secretary. It clarifies the construction ideas and applications of this secretary from three aspects: the architecture, functions and intelligent business services. In intelligent business services, this paper proposes for the first time the application of feature learning based on Auto-encoder for electric power engineering design. Then, the feature parameters learned by Auto-encoder are trained through BP neural network to get the values that meet design standards. Furthermore, this paper constructs the loss function based on the optimal bias to get the results more in line with actual design requirements. Automated design algorithm based on the error model is also provided in this paper. The proposed intelligent digital secretary provides a feasible idea for realizing the leap from current manual design mode to digital design mode.