Assembly line ADC calibration method based on GABP neural network

The invention requests to protect an assembly line ADC calibration method based on a GABP neural network. The method is mainly composed of three stages of pipeline ADC model building, GABP neural network model building and model training and calibration performance testing. The method comprises the following steps: firstly, constructing a pipeline ADC model, generating ADC digital output with errors and reference ADC digital output by adjusting error parameters in the ADC model, and constructing a neural network data set by using the two groups of data; then, building a neural network model, optimizing network parameters by using a genetic algorithm, and training the model by using a training set; and carrying out calibration performance test on the GABP neural network model by using a test set in the data set, carrying out fast Fourier transform (FFT) on the output of the GABP neural network to obtain a spectrogram and a performance index, and carrying out calibration performance evaluation. The problems tha