传统双模超声成像系统中，B模式与多普勒模式通常采用交替发射的工作机制以实现结构成像与血流速度估计。然而，该分时复用策略显著限制了多普勒采样率与时间分辨率，导致血流动力学信号的连续性不足，并在一定程度上增加了系统的硬件调度复杂度与实时处理负担。为突破这一瓶颈并实现波形一体化设计，本文提出了一种基于差分频率（Differential Frequency, DF）处理的血流速度估计方法。所提方法利用高频宽带发射信号在接收端进行差频运算，提取两个高频分量间的频率差以生成等效低频信号，从而在无需交替发射的条件下实现血流速度估计。该差频策略在保持高分辨率特性的同时，显著降低了信号处理复杂度与硬件带宽需求，实现了结构成像与血流检测的波形资源共享。此外，为抑制差频运算过程中可能引入的伪影，本文提出基于协方差矩阵平均的稳健处理策略，有效改善了估计结果的稳定性与精确度。本论文主要聚焦于如何复用高频宽带B模式发射资源进行血流速度估计，旨在实现成像模式的统一化与实时性提升。通过理论推导与仿真实验验证，结果表明所提方法在低信噪比条件下仍能保持较高的血流流速谱估计精度与时间连续性，与传统低频发射方法及稀疏估计算法相比，计算复杂度显著降低且伪影抑制效果更优。综合而言，本文提出的基于差分频率的血流速度估计方法为便携式超声系统的波形一体化设计提供了一种可行且高效的信号处理方法，在保持成像质量与估计精度的同时，大幅提升了系统运行效率，特别适用于未来小型化、便携化场景，如：可穿戴式超声脉搏血流监测仪，支持长期实时血管成像和血流动态监测，为未来高帧率、实时化可穿戴式超声成像系统的实现奠定了技术基础。

In conventional dual-mode ultrasound imaging systems, B-mode and Doppler mode typically operate in an alternating transmission manner to achieve structural imaging and blood flow velocity estimation. However, this time-division multiplexing strategy significantly limits the Doppler sampling rate and temporal resolution, resulting in discontinuous hemodynamic signals and increased complexity in hardware scheduling and real-time processing. To overcome this limitation and realize waveform integration, this paper proposes a blood flow velocity estimation method based on Differential Frequency (DF) processing. The proposed approach exploits high-frequency broadband transmission signals and performs frequency-difference operations at the receiver to extract the frequency difference between two high-frequency components, thereby generating an equivalent low-frequency signal for velocity estimation without alternating transmissions. This DF strategy maintains high-resolution characteristics while substantially reducing signal processing complexity and hardware bandwidth requirements, achieving waveform resource sharing between structural imaging and flow detection. In addition, a covariance-matrix-averaging-based robust processing strategy is introduced to suppress artifacts arising from differential operations, improving estimation stability and accuracy. This research mainly focuses on reusing high-frequency broadband B-mode transmission resources for blood flow velocity estimation, aiming to unify imaging modes and enhance real-time performance. Theoretical analysis and simulation results demonstrate that the proposed method maintains high estimation accuracy and temporal continuity even under low SNR conditions. Compared with conventional low-frequency and sparse estimation methods, it exhibits lower computational complexity and stronger artifact suppression. In summary, the proposed differential-frequency-based blood-flow velocity estimation method offers a feasible and efficient signal processing approach for unified ultrasound waveform design. It maintains imaging quality and estimation accuracy while significantly improving system efficiency, making it particularly suitable for future miniaturized and portable applications, such as wearable home-use ultrasound devices and handheld blood-flow monitors that support long-term real-time monitoring. This work establishes a solid technical foundation for high-frame-rate, real-time, and miniaturized ultrasound imaging systems.