<p>Reservoir operation must balance multiple competing objectives, yet increasing system complexity makes the construction of effective multi-objective models challenging. Existing approaches often include all objectives without considering their interactions, leading to high-dimensional problems and reduced decision efficiency. This paper proposes a framework for constructing a multi-objective reservoir operation model based on competitive feature recognition. The framework introduces quantification indicators (CEI and TCEI) to measure pairwise competition between objectives and applies them to the Nieerji Reservoir, considering hydropower, water supply, environment, agriculture, and wetlands. The classification of competition strengths among objectives is shown to change with variations in water demand (e.g., low vs. high scenarios). Based on this identification, multiple objective function reconstruction schemes are designed. The results reveal that removing the strongest competing objective from the function significantly improves the values of other objectives, while removing the weakest one has minimal impact, particularly under high water demand. Furthermore, optimizing a single strongly competing objective as the sole function achieves the best performance for that objective, albeit at the cost of poorer values for others with which it strongly competes. By identifying objectives with different competition strengths, decision-makers can strategically reconstruct the objective function—such as omitting the weakest objective to simplify the model or focusing on the strongest to boost a specific target. This approach enables a balanced allocation among sectors, minimizes conflicts, and promotes more sustainable and efficient water resource management.</p>