Non-contact voltage measurement is of great significance for intelligent and digital transmission lines of new power systems and real-time characterization of line health. At present, there are problems in non-contact voltage measurement, such as poor determination of measurement point position, low calculation accuracy and great influence by environmental noise, which seriously restricts the engineering application of non-contact voltage measurement methods. Here, a measurement point optimization and voltage inversion calculation method for transmission line sensors combining genetic algorithm and particle swarm algorithm is proposed. Firstly, in order to reduce the influence of electromagnetic interference on non-contact measurement, according to the principle of electric field coupling measurement, the number of observation matrix conditions is taken as the fitness function, and the observation point of the electric field sensor is optimized. Secondly, taking the difference between the offline electric field strength theory and the measured value as the objective function, the genetic particle swarm algorithm is used to calculate the inversion voltage. Finally, the test was carried out under the three-phase test platform built. The results show that under the application of 10% electromagnetic noise, the voltage measurement error is less than 5%, and the environmental noise interference of the voltage inverse calculation result is reduced by 12.9%.