A weather causal traceability wind-light-hydrogen collaborative prediction control method

Abstract

The application provides a weather causal tracing wind-solar-hydrogen collaborative prediction control method, comprising: accurately characterizing the cross-scale dynamic behavior of the fan under extreme working conditions through historical memory fractional calculus and space-time graph convection diffusion operator; using a deep Koopman operator to mine the transient laws of the sea-atmosphere wave multi-physical field, and combining transfer entropy and Jacobian information flow to construct a causal network to realize physical tracing from weather anomalies to power fluctuations. The weather foresight deduction and causal tracing information are used as a feedforward signal, based on a model predictive control framework and an energy storage state adaptive mechanism, to dynamically adjust the power distribution of the electrolytic cell, the fuel cell and the energy storage battery, forming a "perception-tracing-decision-execution" closed loop. The method breaks through the bottleneck of traditional steady-state models in representing transient loads, significantly improving the grid stability of deep-sea renewable energy systems and the service life of hydrogen energy equipment.

Images