针对已有混沌直接序列扩频通信系统的破译算法在低扩频因子和较大多径衰落下无法有效破译信息码的问题,提出了一种基于双模型无迹卡尔曼滤波混沌拟合的破译算法.所提算法联合多模型滤波原理,利用信息符号为有限集的特点,针对其不同取值分别建立对应的滤波模型.各模型下滤波器并行工作,利用广义同步系统同时拟合原混沌系统并估计混沌直扩信号,通过估计误差确定最佳匹配滤波模型,从而得到信息符号的估计.进一步通过引入误差控制因子,增大了不同模型下估计误差的距离,不仅有利于信息符号判定,并且减小了噪声和多径衰落对破译结果的影响.理论推导和仿真结果均证明提出的算法优于已有破译算法. Aiming at the problem that the deciphering algorithm of the chaotic direct sequence spread spectrum communication system can not effectively decipher the information code under the low spreading factor and the large multipath fading, a decipher algorithm based on chaotic fitting of the double model unscented Kalman filter is proposed The proposed algorithm, combined with the principle of multi-model filtering, uses the information symbol as a finite set, and establishes corresponding filter models according to their different values.The filters under each model work in parallel, and simultaneously the original chaotic system is fitted by using the generalized synchronous system The chaotic DS signal is estimated, and the best matching filter model is determined by estimating the error, so as to get the information symbol estimation.Furthermore, by introducing the error control factor, the distance of estimation error under different models is increased, which is not only beneficial to information symbol decision, The effects of noise and multipath fading on the deciphering results are small.The theoretical derivation and simulation results show that the proposed algorithm is superior to the existing deciphering algorithm.