Application of SF6 Gas Concentration Temperature Compensation Based on Improved PSO-RBF

For the temperature drift phenomenon of SF6 gas concentration sensor in high-voltage electrical equipment, the compensation accuracy of the sensor is improved by establishing the temperature compensation model of gas concentration sensor with RBF neural network, and combining the particle swarm optimization algorithm for the optimization of parameters of RBF temperature compensation model. The experimental results show that the method can better compensate the temperature of the gas concentration sensor, and the output of the compensated sensor is less affected by the ambient temperature, with a maximum error of 20 ppm over the full scale range.