双Duffing振子逆向联合信号检测模型

Three efficient methods are proposed to solve three key problems in Duffing oscillator detection system, including the system perturbation by input noise, the synchronization of phase position of driving force and target signal, and the judgment of system output state. First, a reverse phase change method based on the critical threshold dictionary is proposed, where the dictionary is built on the monotonic property of the critical threshold with the change of noise intensity, so that the input signal perturbation is reduced efficiently under low Signal to Noise Ratio (SNR). Second, a double system joint detection method is proposed to solve the problem of phase position synchronization of driving force and target signal. Third, a state of system motion judgment method based on output phase space graph representation is proposed and its feasibility is proved mathematically. The embedded representation of the detection results is fulfilled with low computation and high precision. Finally, the Duffing oscillator detection model suitable to engineering practice is built under low SNR, by which the lowest detection SNR is reduced significantly with constant detection probability.