【摘要】 为在不影响检测概率的情况下有效降低计算量,在一种适用于机动目标且对目标的观测噪声未加任何限制的模型检测与跟踪算法中,采用了改进自适应Viterbi算法。针对Viterbi算法计算量大,实时性较差的缺点,采用一种改进的自适应算法,即结合Viterbi-Beam和Viterbi-Window算法以减少原Viterbi算法中的冗余状态,同时保证差错率无明显提高。我们通过对此算法的仿真表明,在相同的多目标跟踪仿真环境下,改进自适应Viterbi算法可以更有效地提高计算速度,且算法性能与原Viterbi算法相似。更多还原

【Abstract】 Viterbi algorithm has been applied in a examination and tracing algori thm, which is based on hided-Morkov model and suits for the motorized target model and with no restrict ion on the recorded noise of the target. In the light that Viterbi algorithm has a large calculation and has a defect in being on-time, the improved self-adapta tion algorithm reduces the redundancy in Viterbi algorithm and guarantees no evi dent rise in error rate. Through the emulation of this algorithm, it is proved t hat the improved self-adaptation algorithm can speed up the calculation effectiv ely in the same emulation environment of multi-target tracing and its function i s similar to that of the original Viterbi algorithm.更多还原