Quantitative Analysis Model of Transmission Line Anti-icing Technical Transformation Benefit Based on Deep Learning

To accurately analyze the benefit output of transmission line anti-icing technical transformation projects and ensure their continuous and effective implementation, this study proposes a quantitative analysis model based on deep learning. Taking the robot anti-icing technical transformation project of transmission lines as an example, we select benefit output indices including electricity price contribution, social benefits, power outage reduction benefits, and fault handling benefits. Using the project's investment data as input and each benefit output index as output, we construct a quantitative analysis model for transmission line anti-icing technical transformation benefits based on a deep belief network. This model realizes the quantitative analysis of benefit outputs from robot anti-icing technical transformation projects on transmission lines. The experimental results show that when the model has 4 hidden layers and 128 neural structures within each hidden layer, the M value of the proposed model training result is the lowest, only 8.98%, with high quantitative analysis accuracy. Moreover, the quantified indicators of the various benefits output of the A region transmission line robot anti icing technical renovation project from 2015 to 2021 are very close to the historical actual values, with a maximum error of only 0.06 million yuan, which can provide scientific basis for the continuous and effective promotion of actual transmission line robot anti icing technical renovation projects in various regions.