The Design Of Automatic Calibration System For Oil-Surface Thermometers For Transformers

The oil-surface thermometers for transformers(referred to as oil-surface thermometers)is mainly used to monitor the oil temperature of transformer oil.The calibration of traditional oil-surface thermometers needs to be separately calibrated for its temperature probe and temperature switch.During calibration,the temperature of the thermostatic bath needs manual point-by-point setting,and the reading of the operating temperature of temperature switch and pointer dial of oil-surface thermometers is also done manually.The entire process is less efficient.So to improve the calibration efficiency,an automatic calibration system was designed to realize the calibration of its temperature probe and temperature switch simultaneously in a single heating process.The automatic recognition of its pointer dial is realized by a neural network algorithm.The main research contents and innovations are as follows:1.The standard temperature field is generated by programming the thermostatic bath to meet the constant temperature of the temperature probe calibration and the fixed slope temperature rise temperature range for calibrating the temperature switch.Which changed the traditional way of separate calibration.2.Real-time acquisition of the temperature probe and the temperature switch signal through the acquisition card to automatically record the calibration data.In the calibration process,the influence of the time constant on the temperature switch calibration result is taken into account and corrected,and the calibration accuracy is improved.3.A new recognition method based on convolutional neural network is proposed.The recognition of the pointer value is realized by the combination of multiple neural networks.The network analysis shows that the fitting error of the pointer value can reach within ±0.1.