This paper proposes a real-time economic scheduling algorithm for distributed microgrids based on bionics, combining the Sparrow Search Algorithm (SSA) with the multi-cluster reservoir Deep Echo State Network (MRDESN), to address the problems of slow response speed and insufficient accuracy of traditional iterative algorithms in complex microgrids. The algorithm simulates the behavior of sparrow colonies, optimizes network parameters through the discoverer-participant mechanism, and simultaneously adopts a hierarchical reserve pool structure to enhance the nonlinear fitting ability, achieving real-time balance of power load and power generation demand. The research constructs the MRDESN model, and utilizes SSA to optimize key parameters such as the pool size, sparsity, and spectral radius, thereby enhancing training efficiency and global search capability. The experiment takes six distributed generating units as the objects and uses the load data of North American cities to verify the performance of the algorithm. The results show that the absolute errors of the optimized SSA-MRDESN network in the two sets of test data are controlled within ±0.25kW and ±0.2kW respectively, the relative errors reach the order of 10u207b³, and the mean absolute error (MAE) and root mean square error (RMSE) are significantly reduced compared with the traditional methods. This algorithm, by integrating biological bionic mechanisms with deep networks, effectively enhances the real-time performance and accuracy of microgrid dispatching, providing a new solution for economic operation in high-volatility scenarios.