Distributed economic scheduling has strong load balancing and scalability, and is suitable for scenarios where power sources are distributed. As a distributed peer-to-peer network, blockchain has the characteristics of decentralization, high security and high data sharing. This meets the needs of distributed economic scheduling. To solve the problems of difficult global information sharing, individual deception and insufficient security, this paper proposes a strategy based on a blockchain and gradient correction method. The strategy considers various constraints, and adopts the gradient correction method of unified slightly increasing variables to iteratively calculate the power of each power source, and achieve the optimal power allocation under the minimum economic cost to the power grid. The Ethereum technology is adjusted and improved to reduce its computing power consumption and deployment cost, and the scheduling algorithm is contracted to optimize the information sharing between nodes, prevent individual deception and improve the security of the system in combination with the blockchain storage, consensus and encryption mechanisms. Simulation experiments and analysis show that the distributed economic scheduling strategy based on blockchain and gradient modification method has fast convergence speed, good robustness, security and scalability. This work is supported by the National Natural Science Foundation of China (No. 62171185).