River water volume clustering and water quality evaluating method based on water-power, water quality and SOM

Abstract

The invention discloses a river water volume clustering and water quality evaluating method based on water-power, water quality and SOM. Firstly, collecting all data of a river and inputting the data to a water-power model and a water quality model to obtain water-power data of each node of the river and pollution concentration time sequence of each node; then inputting river stream flow data at each moment to a SOM nerve network to obtain a reasonably classified n*n modes, wherein the n*n modes are combined together to obtain a topological graph of the river stream flow; then according to the clustering results of flow, performing feature extraction on pollution concentration time sequence results of each node by means of the SOM nerve network to obtain a topological graph of each pollution source for river pollution distribution and obtaining the pollution mode classification of the pollution source in the one-dimensional river through the topological. By means of the method of the invention, the water volume and water quality of a one-dimensional river can be assessed and the results can be expressed visually.

Description

technical field [0001] The invention relates to the technical field of one-dimensional river flow and water quality evaluation, in particular to a river water volume clustering and water quality evaluation method based on hydrodynamics, water quality and SOM. Background technique [0002] As people pay more and more attention to water resources and environment, there are various evaluation methods of water quality, such as traditional fuzzy comprehensive index method, gray evaluation method, and BP neural network with strong self-learning and self-adaptive ability. The traditional fuzzy comprehensive index method and gray evaluation method require precise mathematical models to describe, and it is difficult to truly simulate the nonlinear change process of the water environment system in view of the complex nonlinear relationship and uncertain factors in the water system; BP neural network The network method has the ability to represent any nonlinear relationship and has str...

AI Technical Summary

Problems solved by technology

The traditional fuzzy comprehensive index method and gray evaluation method require precise mathematical models to describe, and it is difficult to truly simulate the nonlinear change process of the water environment system in view of the complex nonlinear relationship and uncertain factors in the water system; BP neural network The network method has the ability to represent any nonlinear relationship and has strong self-learning and self-adaptive capabilities. It can map the combination of simple nonlinear function functions. It can solve the nonlinear relationship and non-linear relationship in the water system when evaluating water quality However, it is not suitable for the evaluation and classification of complex and changeable water environment systems, and often does not take into account the location of river pollution discharge sources and the impact of discharge on river water quality

Due to the inability to provide very accurate water quality information at nodes, the efficiency of river governance is not high enough

Images