Abstract:Coagulation-based dissolved air flotation is generally recognized as an efficient and economical method for removal of algae from eutrophicated water, whose efficiency is directly affected by the structural characteristics of flocs. However, it is not clear which floc structures are suitable for microalgal floatation, whether large and compact flocs or small and loose flocs have good rising performance. In this study, a novel technology for dynamically tracking the growth and rising behaviors of flocs was developed based on Python-OpenCV to visually uncover the optimal structural characteristics of flocs for Microcysts aeruginosa coagulation-flotation. Different coagulation flow fields (velocity gradient, G ) were adopted to artificially create flocs with various structural characteristics and floatation kinetics. Results showed that G value had significant effects on the growth and rising behaviors of flocs, thus achieving intuitively different flotation kinetics. An idea rising velocity of 91.5 m h −1 was achieved by flocs with optimal structural characteristics generated at a suitable G of 34.1 s −1 , which was 29–75% higher than that of flocs at other G . The rising velocity of floc was synergistically determined by the size and fractal dimension of floc. The structural characteristics (size and fractal dimension) of flocs were introduced into the calculation model of rising velocity, which guided the principle of optimal floc structure characteristics (large and compact) suitable for Microcysts aeruginosa removal using DAF. These may provide new technology for the studies of the growth and settling/rising behaviors of flocs, and lay a theoretical foundation for the design of DAF system with high load rate and the selection of optimal operating conditions applied to cyanobacteria blooms prevention and control.

Constructing a visual detection model for floc settling velocity using machine learning

Constructing a visual detection method for coagulation effect based on image feature machine learning

Application of In-Situ Micro Laser Transmission on Real-Time Monitoring of Flocculation Process

The coagulation-precipitation turbidity prediction model for precision drug delivery system based on deep learning and machine vision

Deep learning-based flocculation sensor for automatic control of flocculant dose in sludge dewatering processes during wastewater treatment

Real-time monitoring of activated sludge flocs via enhanced mask region-based Convolutional Neural networks

Evaluation of Activated Sludge Settling Characteristics from Microscopy Images with Deep Convolutional Neural Networks and Transfer Learning

Image‐based characterization of flocculation processes through PLS inspired representation learning in convolutional neural networks

Research on Water-Level Recognition Method Based on Image Processing and Convolutional Neural Networks

Water Quality Detection Based on FCN and Embedded System

Convolutional Neural Network-Based Machine Vision for Non-Destructive Detection of Flooding in Packed Columns

Data-driven modeling of sluice gate flows using a convolutional neural network

Application of deep learning in predicting suspended sediment concentration: A case study in Jiaozhou Bay, China

Optimizing coagulant dosage using deep learning models with large-scale data

Prediction of submicron particle dynamics in fibrous filter using deep convolutional neural networks

Uncovering the optimal structural characteristics of flocs for microalgae flotation using Python-OpenCV

Numerical investigation of carbon dioxide capture using nanofluids via machine learning

Improving the Precision of Solids Velocity Measurement in Gas-Solid Fluidized Beds with a Hybrid Machine Learning Model

A Novel CNN Modeling Algorithm for the Instantaneous Flow Rate Measurement of Gas-liquid Multiphase Flow

Segmentation and deep learning to digitalize the kinematics of flow-type landslides

Debris flow detection and velocity estimation using deep convolutional neural network and image processing