Classification of Wood Chips Using Electrical Impedance Spectroscopy and Machine Learning.

Area of Science:

• Materials Science
• Biotechnology
• Data Science

Background:

• Wood chips are crucial for pulp and biofuel industries.
• Advanced material analysis can optimize wood chip utilization processes.

Purpose of the Study:

• To analyze heartwood content in pine chips and bark content in birch chips.
• To develop and validate a novel electrode system for material analysis.
• To apply machine learning for accurate classification of wood chip components.

Main Methods:

• Electrical Impedance Spectroscopy (EIS) with a novel electrode system (42 Hz-5 MHz).
• Three-directional measurements (x, y, z) using three electrode pairs.
• Machine learning algorithms: K-nearest neighbor (KNN), Decision Tree (DT), Support Vector Machines (SVM).

Main Results:

• 91% accuracy in classifying pure pine heartwood and birch bark using EIS and KNN.
• 73% accuracy for pine heartwood in mixed materials (four groups).
• 64% accuracy for birch bark in mixed materials (five groups).

Conclusions:

• EIS combined with KNN is effective for analyzing wood chip composition.
• The developed method shows potential for improving industrial wood processing.
• Further refinement may enhance classification accuracy for complex mixtures.