The inflow prediction in the reservoir is a very important and challenging task for the reservoir operation. However, in a large basin, there are many sub-catchments. So initially, one needs to forecast inflow for individual sub-catchments and then sequentially progressing downstream to forecast inflow into reservoir. Which data driven or artificial intelligence technique will be more appropriate for such application?
depends on what data you have for your area. if no or fragmented data, probably a simple estimation technique can be used. with more data more complex modeling can be used.
For a short term prediction (e.g. during a flood event), then I would recommend using the Data-based Mechanistic Modelling approach of Peter Young (https://www.researchgate.net/profile/Peter_Young6)
available as a Matlab toolbox called Captain
http//captaintoolbox.co.uk
(assuming that you have access to Matlab). For longer term predictions, you will need a forecast of rainfall, and use a rainfall runoff model to convert this into an impact on the streamflow. This can be done using the Captain toolbox, or a number of other rainfall-runoff models. If using a rainfall-runoff model, you will need to combing this with a routing model (e.g. Muskingum-Cunge or Lag-Route - I prefer Lag-Route). For example:
https://www.researchgate.net/publication/229357444_Development_of_a_distributed_flow_model_for_underpinning_assessment_of_water_allocation_options_in_the_Namoi_River_Basin_Australia
Article Development of a distributed flow model for underpinning ass...
- Badges
- Science topic