A method for retrieving chlorophyll-a content based on sparrow search optimization BP neural network and readable storage medium thereof

Abstract

This invention discloses a method for chlorophyll a content inversion based on a Sparrow Search Algorithm (SSA) optimized BP neural network, relating to the field of machine learning applied to data inversion technology. This method uses a drone to acquire hyperspectral data of the study water area and simultaneously collects measured chlorophyll a data from the surface water. After preprocessing, a dataset is constructed and divided into training and validation sets. The initial weights and thresholds of the BP neural network are optimized using the Sparrow Search Algorithm (SSA) to construct an SSA-BP model. After training and parameter calibration, the model's reliability is verified using the validation set. Finally, the model is applied to invert the spatiotemporal distribution characteristics of chlorophyll a concentration in the study water area. This invention solves the problems of insufficient adaptation to nonlinear relationships and the BP neural network's tendency to get trapped in local optima in traditional inversion methods, significantly improving the accuracy and stability of chlorophyll a content inversion. Furthermore, the model has strong applicability, providing efficient and reliable technical support for water nutrient status monitoring, algal bloom early warning, and ecological environment assessment.

Images