The invention discloses a phase angle amplitude PID adaptive method based on three-dimensional magnetic characteristic measurement of BP neural network. In the process of signal processing, adopting afrequency domain approach, compared with the time domain method, the harmonics are decomposed into frequency domain, the amplitude and phase angles are independently closed-loop controlled, as the amplitude is traversed to minimize the error between the output voltage and the desired voltage, then the phase angle is found to minimize the output error, PID can quickly find the appropriate amplitude and phase angle, so that the output waveform vector in three directions is a standard spherical or ellipsoidal after synthesis, so that the actual waveform can quickly and accurately approximate thedesired waveform. When the excitation frequency and amplitude change, the weights of the hidden layer and the output layer of the neural network change correspondingly, so that the PID parameters canbe adaptively changed with the feedback adjustment process of the waveform, so that the rapidity and accuracy of the magnetic measurement process are greatly improved, and the response time is greatly reduced.