本文提出了基于一种更严格最大熵谱估计的快速递推算法。讨论了当数据长度有限时，为满足最大熵谱估计的四个近似等价条件，所提出的严格目标函数。在此基础上，通过使目标函数达到极小，用递推法估计出各阶反射系数，然后由Levinson递推得到最大熵谱参数的估计值。该算法算法稳定，快速。计算机仿真表明，此估计方法具有良好的估计质量。 In this paper, a fast recursive algorithm based on a more rigorous maximum entropy spectrum estimation is proposed. We discuss the strict objective function proposed to satisfy the four approximate equivalent conditions of the maximum entropy spectrum estimation when the data length is limited. On this basis, the objective function is minimized, and the recursion method is used to estimate the reflection coefficient of each order, then the estimated maximum entropy spectral parameters are recursed by Levinson. The algorithm algorithm is stable and fast. Computer simulation shows that this estimation method has good estimation quality.