| The topic of this research is establishment and application of the new model of drag components of memory. The memristor, proposed by the Chinese American scientist and researcher Chua Shaotang in 1971, Professor Chua defines the memristor as: a two terminal nonlinear circuit element determined by both the quantity of electric charge and magnetic flux. But the development of semiconductor technology is not mature, Professor Chua’s research on memristor does not capture much attention at that time. But in 2008, the research of the memristor has made rapid development, and the first physical memristive devices with nano materials is produced by Hewlett-Packard Co’s laboratory in this year.According to the literature, the memristor has nonlinear characteristics, which could produce chaotic behavior on certain resistance circuit. As the chaotic circuit, which is same to ordinary chaos circuit, generated by the memristor is sensitive to the circuit parameters; it also extremely depends on the initial value. Due to its memory function, and nano scale characteristics, memristors can be applied in the neural network, analog circuit areas.The current commonly memristor model include HP model, Joglekar, Biolek, Prodomakis window function model and Pickett model. Through the studies of the basic characteristics and shortages of these memristor models, this article designs a threshold memristor model based on Pickett model, and analyzes its memristor properties by simulation and analysis of the voltage current curves and time-domain waveform of the improved memristor. On the basis of the improvement of memristor model, a resistor chaotic circuit of three orders has established, and a dynamic analysis on this chaotic circuit is made. In order to further verify the rationality of the model, a circuit simulation based on mathematical model is set up, and the results are basically consistent with the numerical simulation.In the part of application, a preliminary study of the secure communication of memristor chaotic circuit system is established. Through the numerical simulation analysis, the signal can be encrypted by the design and the encrypted signal has a chaotic and pseudo-random characteristics. Experimental results show that the decrypted signal is almost the same to the source signal. The use of memristor chaotic circuit method for voice encryption is feasible in theory. |