Machine Learning-Based Inversion Algorithm for Particle Size Distribution of Non-Spherical Particle System

Jiaxing Xu; Min Xia; Kecheng Yang; Yinan Wu; Wei Li

doi:10.3788/AOS221901

Acta Optica Sinica, Volume. 43, Issue 9, 0929002(2023)

Machine Learning-Based Inversion Algorithm for Particle Size Distribution of Non-Spherical Particle System

Jiaxing Xu, Min Xia, Kecheng Yang, Yinan Wu, and Wei Li^*

School of Optical and Electronic Information, Huazhong University of Science and Technology, Wuhan 430074, Hubei , China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(12)

Fig. 1. Schematic diagram of GRNN-based particle size distribution inversion model

Download full size

Fig. 2. Flowchart of finding optimal smoothing factor for GRNN-based particle size distribution inversion model

Download full size

Fig. 3. Simulation of biconcave-disk red blood cell

Download full size

Fig. 4. Comparison of neural network inversion result and theoretical particle size distribution of biconcave-disk red blood cell particle system. (a) Optimal result of particle size distribution inversed by GRNN; (b) worst result of particle size distribution inversed by GRNN

Download full size

Fig. 5. Comparison of regularized inversion result and theoretical particle size distribution of biconcave-disk red blood cell particle system

Download full size

Fig. 6. Simulation of ellipsoidal blood red cell

Download full size

Fig. 7. Comparison of neural network inversion result and theoretical particle size distribution of ellipsoidal red blood cell particle system. (a) Optimal result of particle size distribution inversed by GRNN; (b) worst result of particle size distribution inversed by GRNN

Download full size

Fig. 8. Comparison of regularized inversion result and theoretical particle size distribution of ellipsoidal red blood cell particle system

Download full size

Table 1. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of training matrices on biconcave-disk red blood cell particle system
View table
Table 1. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of training matrices on biconcave-disk red blood cell particle system
Number of training matrices 1 2 5 10 20
Mean value of $J_{Σ}^{2}$ 2.2183 1.7541 1.2656 1.0291 1.0027

Table 2. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of scattering angles on biconcave-disk red blood cell particle system
View table
Table 2. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of scattering angles on biconcave-disk red blood cell particle system
Number of scattering angles 9 5 3 2 1
Mean value of $J_{Σ}^{2}$ 1.0027 1.6128 4.1401 5.9012 116.8775

Table 3. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of training matrices on ellipsoidal red blood cell particle system
View table
Table 3. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of training matrices on ellipsoidal red blood cell particle system
Number of training matrices 1 2 5 10 20
Mean value of $J_{Σ}^{2}$ 1.5880 1.1942 0.7800 0.7091 0.6568

Table 4. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of scattering angles on ellipsoidal red blood cell particle system
View table
Table 4. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of scattering angles on ellipsoidal red blood cell particle system
Number of scattering angles 9 5 3 2 1
Mean value of $J_{Σ}^{2}$ 0.6568 1.5453 4.6411 6.1484 51.0278

Tools

Get Citation

Copy Citation Text

Jiaxing Xu, Min Xia, Kecheng Yang, Yinan Wu, Wei Li. Machine Learning-Based Inversion Algorithm for Particle Size Distribution of Non-Spherical Particle System[J]. Acta Optica Sinica, 2023, 43(9): 0929002

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category: Scattering

Received: Oct. 31, 2022

Accepted: Dec. 12, 2022

Published Online: May. 9, 2023

The Author Email: Li Wei (weili@hust.edu.cn)

DOI:10.3788/AOS221901

Topics

laser devices and laser physics

Lasers and Laser Optics

Laser physics

laser manufacturing

Instrumentation, Measurement and Metrology

Table 1. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of training matrices on biconcave-disk red blood cell particle system

Table 1. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of training matrices on biconcave-disk red blood cell particle system

Table 2. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of scattering angles on biconcave-disk red blood cell particle system

Table 2. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of scattering angles on biconcave-disk red blood cell particle system

Table 3. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of training matrices on ellipsoidal red blood cell particle system

Table 3. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of training matrices on ellipsoidal red blood cell particle system

Table 4. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of scattering angles on ellipsoidal red blood cell particle system

Table 4. Mean values of evaluation indexes of inversion effect for neural networks trained with different numbers of scattering angles on ellipsoidal red blood cell particle system