Raw Material Functionality Prediction Using PCA Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting the functionality of raw materials in dairy and plant-based products are time-consuming and unreliable, as they require extensive trial and error with various end products, and existing prediction methods are costly or rely on subjective sensory assessments.
Innovation Solution
A method using Principal Component Analysis (PCA) to evaluate physicochemical properties of raw materials, creating a score plot to identify clusters that predict the functionality of these materials in dairy and plant-based products, allowing for rapid and reliable selection of suitable raw materials.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional testing methods are used to evaluate raw material functionality in end products, then reliable product properties can be achieved, but the pre-selection process becomes extremely time-consuming
Solution Approach 1:
The patent applies preliminary action by performing Principal Component_analysis on raw material data before actual product manufacturing. The method pre-evaluates raw materials from multiple suppliers using statistical models and predefined test parameters, creating a predicted functionality ranking before any end product is made. This allows the most suitable raw materials to be selected in advance, avoiding time-consuming trial-and-error testing while ensuring reliable product properties.
2Stability of the object's composition
If multiple raw material candidates from different suppliers are evaluated to account for quality variations, then product consistency can be maintained, but the number of tests required increases significantly
Solution Approach 1:
The patent applies parameter changes by transforming multiple raw material parameters into a reduced set of principal components through statistical analysis. Instead of testing each raw material parameter individually across multiple suppliers, the method converts data such as protein content, fat content, moisture, and other characteristics into principal components that capture the essential variations. This dimensional reduction maintains the ability to detect quality differences while dramatically reducing the number of tests required.
Solution Approach 2:
The patent applies universality by creating a multi-functional evaluation system that simultaneously assesses multiple suppliers, multiple raw material parameters, and multiple product requirements using a single integrated statistical framework. The Principal Component_analysis model serves multiple functions: it compares suppliers, identifies quality variations, predicts product functionality, and ranks raw materials all in one process, thereby maintaining product consistency without proportionally increasing testing effort.
3Measurement precision
If comprehensive analytical data collection is performed on raw materials, then accurate prediction of end product properties is possible, but the complexity of data processing increases
Solution Approach 1:
The patent applies the extraction principle by isolating and focusing on the most critical sources of variation in raw material data through Principal Component_analysis. Instead of processing all analytical data equally, the method extracts the principal components that explain the majority of variance in the data, effectively filtering out noise and less important variations. This extraction process maintains high prediction accuracy by focusing on the key factors that truly influence end product properties while simplifying the data processing burden.
Data Source
Figure 1
Figure 2
AI summary
A method is proposed for predicting how the incorporation of unknown raw materials affects the predictable and implementable properties of end products.