Study on high-frequency noise optimisation in analogue circuits under stochastic signal fluctuations

By optimising the high-frequency noise of analogue circuits, the anti-interference ability of analogue circuits can be effectively improved, ensuring the purity of signals during transmission and processing.Therefore, an optimisation method of high-frequency noise in analogue circuits considering the stochastic fluctuation characteristics of signals is proposed.Collect analogue circuit signals, extract wavelet packet energy features, composite multi-scale entropy features and stochastic fluctuation features, and use self-attention mechanism to enhance the features, thereby achieving high-frequency noise recognition in analogue circuits.After determining the presence of high-frequency noise in the signal, CEEMDAN is used to decompose the analogue circuit signal and optimise the high-frequency noise of the analogue circuit through signal filtering and reconstruction operations.The experimental results show that the high-frequency noise recognition accuracy of the proposed method remains stable at over 95%, with an average noise coefficient of about 0.15 and an optimisation time of less than 2 s.