<p>The Hindmarsh-Rose model, as a nonlinear neuronal oscillator, has a high sensitivity to weak signals. Aiming at the problem that the initial surge characteristics of the compressor are weak and difficult to identify, this paper proposes an initial surge identification method based on the Hindmarsh-Rose model. Firstly, we introduced an implicit mapping method to study the motion response of the Hindmarsh-Rose model to the simulated weak signal. Then, combined with the deep surge signal of the compressor, it is found that the Hindmarsh-Rose model could suppress the interference of noise and thereby enhance the surge characteristics of the signal through scale transformation analysis and SNR measurement. Further, we studied the influence of two key factors on the SNR of the signal optimized by the Hindmarsh-Rose model: the control parameters of the neuronal system and the scale transformation frequency. Based on this, we obtained the optimal parameters. Finally, through experimental verification, compared with the traditional signal processing methods, this method can effectively improve the SNR of the vibration characteristics during the initial stage of surge and achieve the purpose of early detection of surge.</p>