Adiabatic acceleration calorimetric method based on machine learning

A technology of accelerated calorimetry and machine learning, applied in the direction of thermal development of materials, etc., can solve the problems of temperature tracking process hysteresis and other problems

Active Publication Date: 2020-10-23
NANJING UNIV OF TECH
View PDF8 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] The present invention proposes a machine learning-based adiabatic acceleration calorimetry method, and its purpose is to solve the problem of hysteresis in the temperature tracking process of the existing adiabatic acceleration calorimeter in the stage of adiabatically tracking the thermal runaway of the sample to be measured

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Adiabatic acceleration calorimetric method based on machine learning
  • Adiabatic acceleration calorimetric method based on machine learning
  • Adiabatic acceleration calorimetric method based on machine learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0040] A machine learning-based adiabatic accelerated calorimetry method, which is based on a mathematical nonlinear exponential growth fitting formula without any physical and chemical parameters, combined with the initially known data of the exothermic stage of the adiabatic process of the measured sample, the fitting is obtained to describe the adiabatic The fitting curve and fitting formula of exothermic reaction; and then carry out repeated experiments, so that the temperature control system of the instrument can be programmed to increase the temperature according to the fitting formula to obtain an adiabatic reaction curve that is more in line with the actual situation. The temperature difference between the experimental curves obtained by carrying out repeated experiments according to the formula, when the difference is small enough (ΔT≤1°C), it is judged that the obtained experimental curve is the closest to the actual adiabatic heat release process of the sample, which ...

Embodiment 2

[0054] This embodiment provides a method of adiabatic acceleration calorimetry based on machine learning, which is mainly applied to the calorimetric test of the adiabatic heat release stage of the sample in the heating-waiting-search mode. The heating-waiting-searching mode is the main function of the adiabatic acceleration calorimeter. The operation mode is the most widely used, so this embodiment can meet the requirements of different security detection and risk assessment.

[0055] A method of adiabatic accelerated calorimetry based on machine learning. The method takes the selected mathematical function model as the core. Since the chemical reaction system is exothermic during the thermal runaway process and the exothermic rate gradually accelerates with the increase of the reaction temperature, it shows Exponential growth trend, and the chemical reaction system has a certain complexity, so this embodiment chooses a second-order exponential growth fitting function to descr...

Embodiment 3

[0074] The decomposition process of organic peroxides is a typical exothermic reaction, and the decomposition rate is from slow to fast. When the temperature rises to a certain level, autothermal decomposition will occur, which may eventually lead to thermal runaway or more serious hazards; this example uses Taking tert-butyl hydroperoxide (TBHP) standard solution as an example, the adiabatic exothermic process of TBHP was explored according to the adiabatic acceleration calorimeter experiment, so as to achieve better thermal and kinetic analysis and evaluation.

[0075] The specific analysis process is as follows: the adiabatic acceleration calorimeter is selected as the heating-waiting-search (H-W-S) mode, the temperature in the start-up interval is set to 70°C, the constant temperature time in the start-up interval is 40min, the step heating rate is 10°C / min, and the step length of the step temperature rise is 5°C, the temperature detection threshold is 0.02°C / min, and the s...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides an adiabatic acceleration calorimetric method based on machine learning. The method comprises the following steps: obtaining a mathematical nonlinear exponential growth fittingformula for accurately describing the adiabatic heat release process through fitting of a mathematical nonlinear modelaccording to actually measured temperature rise data of sample in heat release stage, and enabling the temperature control system of the adiabatic acceleration calorimeter to perform programmed temperature rise according to the obtained fitting formula, and obtaining an experimental curve closer to the actual adiabatic temperature rise process of the sample so that more accurate adiabatic calorimetric measurement of the thermal runaway process is realized. According to the method, measurement errors caused by temperature tracking lag are overcome, and later thermal and dynamic analysis and evaluation can be carried out more fully and accurately; the finally obtained adiabatic thermal experiment curve can effectively improve the accuracy of subsequent thermodynamic calculation of the sample, and has important guiding significance for guiding implementation of reaction safety risk evaluation and chemical safety evaluation.

Description

technical field [0001] The invention proposes a machine learning-based adiabatic accelerated calorimetry method, which belongs to the technical field of accelerated calorimetric equipment for thermal safety of chemicals and chemical processes. Background technique [0002] In the chemical industry, chemical production is characterized by flammability and explosion, high temperature and high pressure, poisonous, harmful and corrosive, as well as a wide variety of hazardous chemicals, complex process and harsh process conditions, and most of the chemical reactions are exothermic reactions. Potential The importance of chemical safety production is self-evident because of the high risk. The thermal runaway of chemical reaction is one of the main causes of accidents. Therefore, the research and understanding of the thermal hazard of chemical reaction process and chemical materials is particularly critical. [0003] By testing the thermodynamic, kinetic and thermal hazard paramete...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G01N25/20
CPCG01N25/20
Inventor 蒋军成姚航倪磊卞海涛
Owner NANJING UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap