Urban river not only has the important function in urban hydrological environment, but also is an area for entertainment. Water quality assessment is the core technique in water resource management. As the typical urban river, water samples were collected at 5 sampling points in Xi’an moat from January 2018 to December 2020, and 10 physicochemical parameters were analyzed. In this paper, a comprehensive water quality index (WQI) is designed based on the criterion of water quality classes and entropy weight method firstly. Secondly, the crucial water quality parameters is determined by using mutual information, coefficient of variation and the water quality classes. Finally, an enhanced beetle antennae search algorithm is proposed to optimize the weight values of the crucial parameters in the range 0 to 1, which represent the ratio of the crucial parameter in the minimum WQI (WQImin) model. The WQImin models with the various crucial water quality parameters are implemented for water quality assessment. The effectiveness and superiority of the proposed enhanced beetle antennae search algorithm are validated in comparison with other evolutionary algorithms. The results show that the proposed WQImin model can assess the water quality accurately.