A Study On The Chaos Control And The Application Of The Signal Detection Based On The Modified ADVP Model

Signal processing is an important branch of information science, whose basic theory and technology has been rapidly developed and widely applied in recent decades. Detecting the weak signal with strong noise is a part of its applications. As it can measure the weak signal which the traditional way can't measure, making the signal measurement accuracy is greatly improved, so it is getting the fast development and the universal attention.The chaotic system is sensitive to the small signal and immune to the noise, due to this, it has good prospect in the field of weak signal detection. With the modified ADVP model as the research object, the weak signal detection based on the chaotic theory is studied in this paper. The detail work has been done as follow:(1) The modified ADVP model is analyzed and the ranges of the parameter when the model is in the various periodic state and chaotic state are obtained. Because real-time, convenient operation, strong robustness and applicable to different types of chaotic system are the characteristics of the method of adaptive pulse perturbation, the model which is in the chaotic state is stabilized to the periodic orbits by using this method. The simulation results show that the model was controllable. To the shortcoming of parameter periodic perturbation method, the paper put forward the improvement strategy. It needn't calculate the exact solution of the parameter, so the calculation process is greatly simplified and the calculation efficiency is increased. It can stabilize the modified ADVP model to the periodic orbits by choosing the parameters within the specified range. The simulation results demonstrated the effectiveness and expeditiousness of the improvement strategy. These works provide the theoretical and practical basis for applying the modified ADVP model to weak signal detection.(2) According to the principle of weak signal detection using chaotic oscillator, a new weak signal detection system based on the modified ADVP model is established, and the weak sinusoidal signal with strong noise has been detected successfully by the system, which proves that the new system is sensitive to the weak periodic signal and immune to the strong noise. In the research of detecting the frequency of the weak sinusoidal signal, according to the frequency measurement method, which searchs the variance based on the optimization theory, considering the shortcoming of constraints nonlinear programming method and general traversal search method, a way to measure the weak sinusoidal signal frequency is put forward, which uses the genetic algorithm to search the biggest variance value of the system output signal. The simulation research proves the effectiveness and efficiency of this method. Finally, according to that method, the detection of the multifrequency signal whose frequency is either close or non-close is realized(3) Using the new system, three typical engineering signals are detected. Firstly, the cycle rectangular pulse signal is detected by two different strategies, whose effectiveness was proved by the result of simulation. Secondly, two value of non-periodic with strong noise is detected. By sampling and smoothing the envelope of output signal time diagram, the 0, 1 sequence which the two value of non-periodic signal stands for is restored successfully. This work provides a new idea for the detection and restoration of the two value of non-periodic signal. Finally, the research of detecting the breathing signal, which is belong to the biological signal, has been done, it is a significant exploration in the field of weak biological signal detection based on the chaotic oscillator.