Sigma-Delta modulator calibration method based on neural network

The invention discloses a Sigma-Delta modulator calibration method based on a neural network. The Sigma-Delta modulator calibration method comprises the following steps of: obtaining a Sigma-Delta modulator; the method comprises the following steps of: 1, establishing an actual Sigma-Delta modulator and a reference ideal Sigma-Delta modulator; 2, coherent sampling is adopted, the same input signal is input into a modulator, and FFT simulation is carried out respectively; sampling as a sample data set and a reference data set; step 3, constructing a calibration scheme of the Sigma-Delta modulator; step 4, constructing a convolutional neural network model; step 5, inputting the sample data set to train the convolutional neural network, obtaining a trained neural network calibration model, and iteratively optimizing model parameters and hyper-parameters of the convolutional neural network in the training process until the neural network converges; 6, the trained neural network calibration model is applied to the