Condition Calibration of Hydraulic Power Station Secondary Equipment Pressure Plate Based on Faster-RCNN Model

To address a series of problems that exist in the manual calibration of the secondary equipment platen of hydropower stations, this paper proposes the state calibration of the secondary equipment platen of hydropower stations based on the Faster-RCNN model. The efficiency and accuracy of the model in small target detection are improved by introducing the short connecting layer and feature map pyramid network; Transfer learning is introduced, where the weight parameters obtained by pre-training the model using publicly available datasets are used as the initial values for formal training, reducing the model's dependence on the original data. The experimental results show that the accuracy of the target detection model used in this paper is 90%, which is obviously better than other detection models. High-precision and high-efficiency state identification of secondary equipment platen can be realised in practical engineering inspection.