In order to improve the smoothness of wind power consumption and grid-connected power, for the single investment in new energy hydrogen production mode faces the dilemma of low equipment utilization and operating losses. In this research paper, we have devised a novel operation model for a multi-principal united enterprise. This model depends on the classical Shapley value method to deal with the calculation of operation costs, the allocation of investment costs, and the distribution of benefits. On this basis, an optimal scheduling model of new energy hydrogen production system based on the objective function of economy and power supply reliability is proposed, and the model is solved by the improved slime mold algorithm, which improves the searching efficiency by initializing the population with Sobol’s low-discrepancy sequences, and introduces the searching strategy obeying the Levy distribution to enhance the optimization ability of the algorithm. Simulation analysis results show that the proposed strategy can effectively utilize the time-sharing tariff mechanism and obtain significant economic benefits while maximizing the utilization of wind and light. By sharing the capacity of multiple subjects, the overall economy of the hydrogen production plant is greatly improved, and a win-win cooperation among multiple parties is realized. The research results of this study provide a new angle for the building and running management of large-scale new energy use facilities that have whole-network public-benefit attributes.