An online detection method for the double-effect concentration process of Danhong injection
A Danhong injection, double-effect concentration technology, applied in measuring devices, weighing by removing certain components, and conducting material analysis by optical means, etc., to achieve the effect of improving the level of quality control and stable product quality
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0048] Example 1 (Establishing a global model using all concentrated liquid samples)
[0049] 1. Design a near-infrared online detection device
[0050] See figure 1 , Near-infrared online detection device includes first valve 1, second valve 2, third valve 3, fourth valve 4, concentration equipment first-effect heating chamber 5, concentration equipment first-effect evaporation chamber 6, concentration equipment second-effect heating chamber 5 ', concentration equipment two-effect evaporation chamber 6', circulation manifold bypass 7, miniature magnetic drive gear pump 8, filter 9, flow cell 10, fiber optic probe 11, sampling tube 12, near-infrared spectrometer 13, one-effect circulation manifold 14 , Two-effect circulation main pipe 14' and steam pipeline; the first-effect heating chamber 5 of the concentration equipment is connected through the first-effect circulation main pipe 14 to the first-effect evaporation chamber 6 of the concentration equipment, and the second-effect h...
Embodiment 2
[0076] According to the method of Example 1, the difference is that the principal component analysis method is used to classify the concentrated liquid samples into high and low concentrations, and local models of each quality control index are established according to the types.
[0077] 1. Spectral principal component analysis
[0078] Principal component analysis was performed on the collected 176 concentrated liquid samples. As shown in the figure, the first principal component score of high-density samples is higher, and the first principal component value of low-density samples is lower. The value 0.3 is the boundary, and high and low concentration (density) samples are divided into two categories. The above results indicate that for samples with a large concentration span, segmented modeling may obtain better predictive power than global modeling.
[0079] 2. Partial least squares regression was used to establish a local quantitative correction model for each quality contro...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More