Research Of Dry-Type Transformer Temperature Control System Based On Gray Prediction Theory

Dry-type transformers are widely used in various aspects during the production of people’s lives. Dry-type transformers temperature control system is used to insure that the dry-type transformers work within the normal temperature range. The performance of the dry-type transformers is related to the normal work and life of the dry-type transformers. Therefore, the study of dry-type transformers temperature control system has important practical significance.The paper summarizes the research status of dry-type transformers temperature control system and the application in various fields of the gray prediction theory. The paper describes the modeling process of the GM(1,1) prediction model which is commonly used in gray prediction theory. The paper analyzes the applicable conditions of GM(1,1) prediction model and calculates its relevant parameters. The paper discusses the prediction accuracy of the common gray forecasting model is not high in dealing with oscillation sequence and in order to solve the problem a smooth operator is introduced to the forecasting model. The paper then analyzes the mechanism of the smooth operator and improves the prediction model so that the improved gray prediction model has a higher prediction accuracy in dealing with the temperature sequence which changes oscillatorily. The improved gray prediction model is applied to forecast the three-phase current and temperature of the dry-type transformers in the temperature control system. A control strategy based on gray prediction theory of the dry-type transformer temperature control system is given. In the control strategy the real-time temperature and forecast temperature are used as decisive factors to control when to turn on the fans, and the real-time current and forecast current are used as logical factors to control how many fans to turn on.The circuit of the dry-type transformer temperature control system is designed though simulation. The paper selects PIC18F4520 chip as the core processor and selects PT100 platinum resistance as the temperature sensor. The temperature signal acquisition circuit, current signal acquisition circuit, signal processing circuit, display circuit, key circuit, serial communication circuit and alarm circuit are designed. A software control program of the temperature control system and a flow chart of the software program are given in the paper. The detailed analysis to the microcontroller initialization, digital filtering, gray prediction data modeling, A / D converter, serial communications, key debounce, the watchdog and so on are carried out. At last the paper uses Proteus software to simulate the designed hardware and software and gives the simulation results when the temperatures of the transformer at 0 ℃, 110 ℃ and 130 ℃, which prove the correctness of the design.