Optimization of Liquid−Liquid Mixing in a Novel Mixer Based on Hybrid SVR-DE Model

Abstract

To solve the problem of evenly mixing flocculant and sewage, a new type of two-chamber mechanical pipe mixer was numerically calculated and its working principle was studied by means of the internal flow field. The single factor numerical simulation and analysis of some of the structural parameters in the mixer were carried out to determine the influence of different parameters on the results. Latin hypercube sampling was used to design 100 sets of test tables for the four variables of the branch pipe diameter, sewage flow rate, the installation height of the impeller, and the angle of the deflector. The results were optimized using the SVR-DE algorithm. After optimization, the variation coefficient of export flocculant mixing uniformity was 16.02%, which was increased by 74.94% compared with the initial 63.921%. The power consumption of the impeller was reduced by 8.30%. The concentration curves of the flocculant at different positions of the outlet tube could quickly converge to the target value.

Numerical Study on Cavitating Flow-Induced Pressure Fluctuations in a Gerotor Pump

“…Notably, there is no direct contact between the internal and external rotors and tip clearances. relevant works have focused on optimization design methods and machine learning algorithms applied to fluid machinery [25][26][27]. For instance, Huang et al [28] proposed a hybrid neural network approach for predicting the energy performance of centrifugal pumps.…”

“…Ivanović et al [24] established a mathematical prediction model and studied the factors affecting gerotor pump flow and volumetric efficiency using a factor design and response surface methodology. In addition to these studies, other relevant works have focused on optimization design methods and machine learning algorithms applied to fluid machinery [25][26][27]. For instance, Huang et al [28] proposed a hybrid neural network approach for predicting the energy performance of centrifugal pumps.…”

The regular study of the influence of rotational speed and height on the mixing performance ofArc-Blade stirrer at different blade diameters

“…Guo et al [7] investigated the effect of impeller radius and impeller mounting height on gas-liquid mixing, showing that increasing impeller mounting height increases the gas content rate in a delayed axial direction. Hao Wang et al [8] optimised design parameters such as impeller mounting height and deflector angle and showed that the optimised impeller power was reduced by 8.3%. Yuan Tan et al [9] investigated the effect of lateral angle, number of paddles and speed on mixing performance of near-nozzle continuous mixers and concluded that four-paddle agitators provided better mixing performance.…”