CNN超混沌特性及在空间多目标测控系统中的应用研究

发布时间：2018-04-16 00:06

本文选题：细胞神经网络 + 超混沌　；参考：《哈尔滨工业大学》2015年硕士论文

【摘要】：CNN(Cellular Neural Network,细胞神经网络)是基于Hopfield神经网络和细胞自动机提出的一种反馈型神经网络。CNN具有多个可供设置的参数且当CNN的参数设置满足一定的条件时,就会产生混沌及超混沌现象。CNN超混沌系统具有初值敏感性,由CNN超混沌系统生成的超混沌扩频序列具有良好的伪随机特性,能够代替传统扩频序列应用于CDMA(Code Division Multiple Access,码分多址)系统中。此外,CNN优良的并行信号处理能力也可以用来解决CDMA系统的多用户检测问题。目前,空间多目标测控系统在通信与数据传输层面多采用CDMA体制,即用不同的扩频序列来区分不同的卫星。因此,对CNN超混沌系统的研究完全可以应用于空间多目标测控系统中。本文针对CNN超混沌特性开展理论和应用研究具有重要意义。本文首先对CNN超混沌系统的动力学特性展开研究,建立CNN全互连拓扑结构及状态方程,分析CNN超混沌系统的初值敏感性、有界性和遍历性、Lyapunov指数、奇怪吸引子等特性,并基于上述特性,提出了CNN超混沌系统判定及筛选算法,该算法可以从大量的CNN中筛选出能够产生超混沌现象的CNN系统,这为以后的研究工作奠定了基础。然后本文对CNN超混沌扩频序列的生成、量化及筛选展开了研究。通过对超混沌扩频序列的性能进行分析,提出了CNN超混沌扩频序列筛选算法,以筛选出具有良好特性的超混沌扩频序列。将筛选出的CNN超混沌扩频序列应用于CDMA系统之中,可提高系统的容量、误码性能和保密性能。然后本文对CDMA系统中的多用户检测问题展开了研究。首先分析了几种常见的多用户检测算法,然后结合最优多用户检测算法和CNN的Lyapunov函数,提出了CNN多用户检测算法,并针对其容易陷入Lyapunov函数局部极小点的问题进行了简化降阶。简化的CNN多用户检测算法能够快速有效地减弱用户间的多址干扰。最后,本文研究了CNN超混沌系统在空间多目标测控系统通信与数据传输层面的应用。通过搭建空间多目标测控系统的通信与数据传输模型,设计具体通信方案,将筛选出的CNN超混沌扩频序列及简化的CNN多用户检测算法应用于空间多目标测控系统的收发两端,使系统获得了良好的性能。
[Abstract]:CNN(Cellular Neural Network (CNN) is a feedback neural network based on Hopfield neural network and cellular automata.The chaotic and hyperchaotic systems are sensitive to initial values, and the hyperchaotic spread spectrum sequences generated by CNN hyperchaotic systems have good pseudorandom characteristics.It can be used in CDMA(Code Division Multiple access (CDMA) system instead of traditional spread spectrum sequence.In addition, the excellent parallel signal processing ability of CDMA can also be used to solve the problem of multiuser detection in CDMA system.At present, CDMA system is widely used in the communication and data transmission of space multi-target TT & C system, that is, different spread spectrum sequences are used to distinguish different satellites.Therefore, the study of CNN hyperchaotic system can be applied to space multi-target measurement and control system.In this paper, it is of great significance to study the hyperchaos characteristics of CNN in theory and application.In this paper, the dynamical characteristics of CNN hyperchaotic system are studied, and the topological structure and state equation of CNN complete interconnection are established. The initial value sensitivity, boundedness and ergodicity of CNN hyperchaotic system are analyzed.Based on the above characteristics, a CNN hyperchaotic system decision and screening algorithm is proposed. The algorithm can screen out the hyperchaotic CNN system from a large number of CNN, which lays a foundation for future research.Then, the generation, quantization and screening of CNN hyperchaotic spread spectrum sequences are studied.Based on the analysis of the performance of the hyperchaotic spread spectrum sequence, the CNN hyperchaotic spread spectrum sequence screening algorithm is proposed to screen the hyperchaotic spread spectrum sequence with good characteristics.The CNN hyperchaotic spread spectrum sequence is applied to the CDMA system, which can improve the capacity, error performance and security performance of the system.Then, the problem of multi-user detection in CDMA system is studied in this paper.In this paper, several common multiuser detection algorithms are analyzed, and then the CNN multiuser detection algorithm is proposed by combining the optimal multiuser detection algorithm with the Lyapunov function of CNN, and the order reduction of the CNN multiuser detection algorithm is simplified to the problem that it is easy to fall into the local minima of Lyapunov function.The simplified CNN multiuser detection algorithm can reduce the multiple access interference between users quickly and effectively.Finally, this paper studies the application of CNN hyperchaotic system in the communication and data transmission of space multi-target measurement and control system.By building the communication and data transmission model of the spatial multi-target measurement and control system, the specific communication scheme is designed. The selected CNN hyperchaotic spread spectrum sequence and the simplified CNN multi-user detection algorithm are applied to the two ends of the spatial multi-target measurement and control system.The system has achieved good performance.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：V556;O415.5

【参考文献】