Abstract
Information about future streamflow is important for hydropower production planning, especially for damless hydro-power plants. The river flow is a reflection of various hydrological, hydrogeological, and meteorological factors, which increases the direct modeling difficulty, and favors the use of data-driven methods. In this paper, we propose the use of one-dimensional convolutional neural networks (1d-CNN) for multi-day ahead river flow forecasting and we present a multi-input model using correlated-input time-series. The proposed model was applied at the Madeira River, the Amazon’s largest and most important tributary, near the Santo Antonio damless hydro-power plant. We compared the proposed correlated-input 1d-CNN to a single-input 1d-CNN model and some baseline models. Furthermore, we conclude that 1d-CNN performed better than all baseline models and that the correlated-input forecasting model is 5 times smaller than the single-input equivalent with accuracy improvements.
Highlights
Streamflow forecasting is an important activity that can help in hydro-power scheduling, reservoir planning, and flood management
We present a unidimensional convolutional neural network (1d-CNN) river flow forecasting model using correlated-input time series
A section of the northern region of Brazil can be seen in Fig. 1, the main rivers of this region are shown in blue and the Madeira River is highlighted in red, the site where data was collected is highlighted by a cross near the city of Porto Velho
Summary
Streamflow forecasting is an important activity that can help in hydro-power scheduling, reservoir planning, and flood management. Streamflow forecasting plays an important role in the environmental and economic sciences. In what concerns the Brazilian electric system, river flow forecasting is extremely relevant, considering 66.6% of domestic electrical energy was supplied by water sources in 2018 [1]. Future streamflow data could be used to model the amount of energy that would be possible to generate from water sources and use this model to optimize the use of resources, especially in systems that use damless hydroelectric plants. The river flow forecast depends on a wide number of variables such as precipitation, evapotranspiration, ambient temperature, soil characteristics of the catchment, VOLUME XX, 2017 The runoff is the most important input for the hydropower system and one of the most decisive parameters in hydropower production planning and decision making [2], [3].
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have