Main content area

The Cascade Reservoirs Multi-Objective Ecological Operation Optimization Considering Different Ecological Flow Demand

Yang, Zhe, Yang, Kan, Hu, Hu, Su, Lyuwen
Water resources management 2019 v.33 no.1 pp. 207-228
algorithms, models, power generation
In order to coordinate the power generation and downstream ecological protection benefits,the optimal ecological scheduling of cascade reservoirs is pivotal. In current paper, an improved chaotic normal cloud shuffling frog leaping algorithm (CNSFLA) based on chaotic initialization, cloud model evolution strategy and heuristic frog activation mechanism is proposed to overcome defects of conventional SFLA. Moreover, the multi-objective ecological scheduling model for Qingjiang cascade reservoirs is established with consideration of basic, suitable and ideal ecological flow requirements in Geheyan and Gaobazhou control sections. Afterwards, the model established is applied to cascade reservoirs ecological scheduling in Qingjiang. The water level corridor and penalty function are used to handle constraints. The scheduling results for long series indicate that benefits of power generation and basic ecology flow requirement in downstream control sections are fulfilled completely. While for suitable and ideal ecology flow requirements, 98.33%, 99.17 and 86.33%, 88.17% guarantee rate corresponding to Geheyan and Gaobazhou control sections can be achieved during whole scheduling periods. In terms of typical dry year, the less inflow makes it hard to reach the ideal ecology flow requirement in control sections during several periods even though the scheduling by CNSFLA. The guarantee rates and mean monthly ecological flow shortage are 66.67%, 66.67% and 71, 36 m³/s, respectively. Finally, performance analysis of CNSFLA verifies its effective search ability with high quality & stability results. The cascade power generation obtained by CNSFLA in long series scheduling is 75.45(10⁸ kW·h), corresponding guarantee rates of ideal ecological flow requirement are 86.33 and 88.17%.