摘要
由于传感器网络的自身特征和节点的资源受限,使得对观测信号的处理必须考虑量化和能耗等因素,而引入的量化噪声同时增加了系统整体噪声的复杂性.针对传感器网络中整体噪声统计特性难以准确数学建模的特点,提出了一种基于代价参考粒子滤波的混沌信号重构算法.算法采用容积点变换以获得相对准确的更新粒子,并将局部重构信号的代价增量构建为广义似然比函数,用来选择传感器网络中的有效工作节点.仿真结果表明:所提算法可实现混沌信号的有效重构,且在噪声统计特性未知时性能要优于容积卡尔曼粒子滤波算法;算法同时能够通过选择不同的广义似然比阈值,实现网络能耗和重构精度的折中.
A memristor is a nonlinear nanoscale-sized element with memory function, and it has an italic type “8” voltage-current relation curve that looks like a pinched hysteresis loop characteristic. The memristor is utilized to construct chaotic circuit, which has attracted the attention of the researchers. At present, most of studies focus on applying one or two memristors to the chaotic circuit. In order to study the multi memristor chaotic circuit, in this work we propose a six-order chaotic circuit with two flux-controlled memristors and a charge-controlled memristor. A corresponding six-order nonlinear dynamic differential equation of the circuit state variables is established. The dynamic properties of the circuit are demonstrated in detail. The analyses of equilibria and equilibrium stability show that the circuit has an equilibrium located in the three-dimensional space which is constituted by memristor internal state variables, and it is found that the equilibrium stability is determined by the circuit parameters and the initial states of three memristors. The Lyapunov exponent spectra and bifurcation diagrams of the circuit imply that it can produce two bifurcation behaviors by adjusting its parameters, which are Hopf bifurcation and anti-period doubling bifurcation. The hyperchaos, transient chaos and intermittency cycle phenomena are found in the same system. The dynamical behavior of this circuit is dependent on the initial state of memristor, showing different orbits such as chaotic oscillation, periodic oscillation and stable sink under different initial states. Finally, the simulation results indicate that some strange attractors like lotus type and superposition type are observed when voltage and electricity signal in observing chaotic attractors are generalized to power and energy signal, respectively. And the attractor production between the energy signals of the memristors are studied. Specially, when different initial conditions of three memristors are used to simulate the circuit, we can find t
出处
《物理学报》
SCIE
EI
CAS
CSCD
北大核心
2017年第4期36-44,共9页
Acta Physica Sinica
基金
国家自然科学基金(批准号:60872123)
国家自然科学基金委员会广东省自然科学基金联合基金(批准号:U0835001)
广东省高校科技创新计划基金(2013KJCX0178)资助的课题~~
关键词
代价参考
混沌信号重构
能耗
广义似然比
cost reference, chaotic signal reconstruction, energy consumption, generalized likelihood ratio
