Multi-stage stochastic programming energy scheduling method and system considering unpredictability

The invention discloses a multi-stage stochastic programming energy scheduling method and system considering unpredictability, which take the minimum daily operation expected cost as a target, consider the uncertainty of distributed renewable energy power generation, establish a multi-stage stochastic programming model, and can determine a pre-scheduling decision in each stage according to the implementation of previous uncertain information, thereby improving the scheduling efficiency. The method is not influenced by future uncertain information, the actual operation rule of the system is met, and the unexpectability is met; in order to avoid curse of dimensionality of solving a multi-stage random programming problem, an optimal scheduling decision tree obtained by solving multi-stage random optimization by adopting an SDDP algorithm has a wider decision space than a single decision scheme obtained by deterministic optimization, and a scheduling decision can be updated based on implementation and decision of