Research on Improving the Measurement Accuracy of the Hall Current Sensor Based on the Extreme Learning Machine

For the hall current sensor is affected by temperature requires temperature compensation, this paper adopts the method of multi-sensor technology, the method of current sensor as the main sensor, temperature sensor as auxiliary sensor, the measured data as data fusion input, different test current in hall current sensor, the temperature sensor to monitor its working temperature, according to the monitoring results using the Extreme Learning Machine set up the function relationship between the measured current, hall current sensor output voltage and its working temperature and data fusion processing to weaken the temperature interference to the current sensor. Finally, it is shown that through simulation, the measurement accuracy of current sensor is 1~2 orders of magnitude higher than before compensation, which can better meet the requirements of actual measurement.