Study on reservoir optimal operation based on coupled adaptive ε constraint and multi strategy improved Pelican...

Area of Science:

• Optimization algorithms
• Water resource management
• Flood control engineering

Background:

• Reservoir group operation is a complex, multi-stage optimization challenge.
• Existing algorithms struggle with high-dimensional, constrained problems in flood control.

Purpose of the Study:

• To develop an enhanced optimization algorithm for optimal reservoir operation.
• To improve flood control efficiency in cascade reservoirs.

Main Methods:

• Coupling the Pelican optimization algorithm with adaptive ε constraint methods.
• Incorporating population initialization, reverse differential evolution, and t-distribution perturbation.
• Applying the improved algorithm (ε-IPOA) to a peak-cutting model for Sanmenxia and Xiaolangdi reservoirs.

Main Results:

• The ε-IPOA algorithm demonstrated strong optimization ability and stable performance on test functions.
• Applied to cascade reservoirs, ε-IPOA achieved a 44% peak shaving rate at Huayuankou control point.
• Effectively reduced peak flow to 12,319 m³/s, well below the safe overflow limit of 22,000 m³/s.
• Outperformed ε-POA and ε-DE algorithms, which failed to find valid solutions.

Conclusions:

• The proposed ε-IPOA algorithm offers a novel and effective approach for optimizing cascade reservoir flood control.
• This method provides a significant advancement in managing complex water resource systems for enhanced safety.