A new hybrid model based on relevance vector machine with flower pollination algorithm for phycocyanin pigment...

Area of Science:

• Environmental Science
• Water Resource Management
• Computational Intelligence

Background:

• Cyanobacteria blooms pose risks to water resources, necessitating precise monitoring.
• Traditional methods for cyanobacteria assessment, like chlorophyll-a and cell counts, are inefficient and inaccurate.
• Phycocyanin (PC) concentration is a more reliable indicator for cyanobacteria monitoring, but traditional estimation methods are costly and time-consuming.

Purpose of the Study:

• To develop a novel hybrid artificial intelligence (AI) model for accurate prediction of phycocyanin (PC) concentration in water resources.
• To address the limitations of standalone AI models in handling nonlinear systems and environmental data uncertainties.
• To evaluate the performance of the proposed hybrid AI model against traditional standalone models for water resource management.

Main Methods:

• A hybrid AI model integrating Relevance Vector Machine (RVM) and Flower Pollination Algorithm (FPA) was developed (RVM-FPA).
• The RVM-FPA model was applied to predict PC concentration in water resources.
• Model performance was assessed by comparing the hybrid RVM-FPA with a standalone RVM model at two distinct monitoring stations (508 and 478) using statistical and graphical evaluations.

Main Results:

• The hybrid RVM-FPA model demonstrated superior performance in predicting PC concentration at both tested stations.
• The proposed hybrid model significantly outperformed the standalone RVM model.
• The RVM-FPA model proved to be a reliable tool for estimating PC concentration in water resources.

Conclusions:

• The hybrid RVM-FPA model offers a significant advancement in monitoring cyanobacteria concentration in water resources.
• This AI-driven approach provides a more accurate and efficient alternative to traditional monitoring methods.
• The RVM-FPA model can be reliably employed for effective water resource management and early detection of cyanobacteria blooms.