现有脉冲星轮廓恢复方法多基于等封装段（bin）间隔分布策略折合观测轮廓，导致信号处理结果会受到封装段的限制，难以兼顾恢复轮廓的信噪比（SNR）以及脉冲到达时间（TOA）的计算精度。针对这一问题，结合脉冲轮廓波形不同区域对信号处理结果的影响，根据完整观测信息，对基于等光子分布策略、变封装段的轮廓恢复方法进行了研究。同时，为能直接构建最佳的脉冲观测轮廓，同时考虑了变封装段间隔及封装段数的优化策略，利用信息准则对轮廓恢复的最佳折叠bin数进行了分析，摆脱人为经验选取问题。仿真结果表明：与传统方法相比，应用提出的变封装段轮廓恢复方法能准确恢复出与模板轮廓高度相似的脉冲星观测轮廓，在观测数据充足的情况下，信号处理过程相位差计算精度能提高25%以上；观测数据不充足时或对于毫秒脉冲星，方法具有明显优势，恢复轮廓的波形特征更加突出，与模板的相似度指标提升3%~7%左右，相位差计算精度的提升能达到接近50%。因此，研究的基于等光子分布策略的变封装段脉冲星轮廓恢复方法原理及实现简单，适用于实际工程应用，能有效提高脉冲星信号处理精度。

For most existing methods for pulsar profile recovery， the observation profiles are folded based on the distribution of equal bin interval. With this strategy， the signal processing results will be limited by bin， making it difficult to give consideration to both the Signal-to-Noise Ratio （SNR） of profile and the estimation accuracy of pulse Time of Arrival （TOA）. To solve this problem， considering the influence of different regions of pulse profile on signal processing， a profile recovery method based on distribution strategy of equal photon number is studied， and bin changes according to the complete observation information. In addition， to be able to build the best observation pulse profile directly， optimization of variable （varying？） bin interval and number is considered simultaneously， and the optimal bin number for profile recovery is analyzed using the information criterion. Human experience is thus eliminated in determining the number of bin. Simulation results show that the proposed method can accurately recover the observation profiles similar to the pulsar template. Compared to traditional methods， the proposed method can improve the calculation accuracy of phase difference during signal processing by more than 25% with sufficient observation data. In case of insufficient observation data or for millisecond pulsars， the advantages of the method are more obvious， with more prominent features of recovered profile， 3%-7% increase of the index of similarity with the template， and nearly 50% improvement of the corresponding calculation accuracy of phase difference. With simple principle and implementation， the proposed method can effectively improve the accuracy of pulsar signal processing， and is applicable for engineering applications.