Many-objective optimization scheduling of cascade reservoirs in small watersheds based on an evolutionary multitasking framework.

With the intensification of global climate change and increasing anthropogenic pressures, effective water resource management has become a critical challenge for sustainable development. Small watershed cascade reservoirs must balance multiple competing objectives, including flood control, hydropower generation, and water supply for ecological, agricultural, and industrial uses. This study develops a many-objective optimization scheduling model for cascade reservoirs in the Lushui River Basin and proposes a constrained many-objective evolutionary multitasking optimization algorithm (EMCMOA) to solve it. The algorithm incorporates a dual-task structure with dynamic knowledge transfer to improve search efficiency and solution quality. Experimental results show that EMCMOA outperforms several state-of-the-art algorithms on benchmarks and real-world scenarios, achieving up to 15.7% improvement in IGD and 12.6% increase in HV. Furthermore, EMCMOA demonstrates strong adaptability to varying hydrological conditions, providing reliable and adaptive scheduling strategies. These results highlight its capability to support flexible, many-objective trade-offs in real-world water resource management.