Temperature sensor for intelligent monitoring system of dry-type transformer

The substation has undergone further transformation of the power system. During the construction of the substation, it was decided to apply artificial intelligence technology to strengthen the monitoring of transformers. The transformation work revolves around two aspects: real-time monitoring and long-term data collection and accumulation. The monitoring objectives include three aspects: transformer function, electricity safety, and power supply stability. Long term data collection includes line loss collection and changes in electricity consumption in the target area.

In terms of real-time monitoring, to ensure the effectiveness of intelligent monitoring. Based on data information, the standard parameters of the observed object should be collected first, and the intelligent device should be trained on the machine. On this basis, establish several models. During the specific work, sensors are used to match information data with the model, understand the working situation of the system, and achieve effective detection. Taking the monitoring of transformers as an example, under normal working conditions, the working temperature of transformers in substations is 50-70 ℃. If there is inner core corrosion or other faults, the resistance will significantly increase. The core indicators of intelligent monitoring are mainly transformer temperature. During the experiment, first, it is necessary to collect information on the transformer within 30 working days, and use debugging experiments to understand the temperature situation of the voltage regulator at different load levels. Four standard models are constructed: low load model (20% -30%), medium load model (40% -60%), high load model (70% -80%), and overload model (over 90%). On this basis, additional temperature values of transformers under different fault conditions, including overvoltage, overcurrent, and core melting, were collected. All data sets were input into the microcontroller, and a total of 1041 samples were obtained. Random forest method was used to quickly complete machine training and achieve memory.

The main components of an intelligent system include sensors, communication structures, microcontrollers, and ancillary parts. The intelligent system is externally located near the main transformer for real-time monitoring around the clock. The monitoring work lasted for 30 days, and the results show that the transformer is working normally. The intelligent system can also accurately provide the temperature value of the transformer during operation, indicating that its working situation is good. In addition, in order to understand whether the intelligent system can identify problems and issue alarms in abnormal situations, relevant parameters are input into the computer for virtual experiments. The experiment is conducted 30 times, and the intelligent system can accurately identify and issue alarms in a short period of time. Based on the results of simulation experiments and conventional experiments, the intelligent system performs well in transformer monitoring work. The above research results demonstrate the application value of artificial intelligence technology in electrical automation control.

• Prev：Top 10 Best Transformer and Temperature Controller Suppliers in China
• Next：How many degrees Celsius should the temperature of a dry-type transformer not exceed