Material pyrolysis behavior simulation method and device based on chemical and element composition, electronic equipment and storage medium
By establishing a composition-parameter mapping model through deep learning, the pyrolysis behavior of materials can be predicted based on chemical and elemental composition, and simulated thermogravimetric curves can be generated. This solves the problem of time-consuming and material-intensive processes in existing technologies and achieves efficient acquisition of thermogravimetric curves.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 中国石油大学(北京)克拉玛依校区
- Filing Date
- 2026-04-27
- Publication Date
- 2026-05-29
AI Technical Summary
Existing thermogravimetric analysis methods require actual samples and thermogravimetric instruments for experiments, which is time-consuming and material-intensive, making them unsuitable for high-throughput material screening and early concept design, and they are also sensitive to experimental conditions.
By establishing a composition-parameter mapping model through deep learning, the pyrolysis behavior of materials can be predicted using chemical and elemental composition, generating simulated thermogravimetric curves and differential thermogravimetric curves, thus avoiding the need for actual experiments.
It reduces costs and time, improves the efficiency of obtaining thermogravimetric curves, and provides rapid data support for material design and performance evaluation without requiring actual thermogravimetric experiments.
Smart Images
Figure CN122117182A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of thermal analysis and material characterization technology, and is a method, apparatus, electronic device and storage medium for simulating the pyrolysis behavior of materials based on chemical and elemental composition. Background Technology
[0002] Thermogravimetric (TGA) curves and micro-TGA curves obtained from thermogravimetric analysis are core data for studying the thermal behavior of materials. They can reflect stages such as water loss, thermal decomposition, carbonization, and burnout. Specific TGA curves show when and how much weight is lost, while micro-TGA curves reveal when and at what rate weight loss occurs. They are important means of evaluating the thermal stability, combustion characteristics, and reaction kinetic parameters of materials.
[0003] Currently, thermogravimetric analysis (TGA) is widely used in the study of various material systems, including coal, wood-based combustibles, polymers, and lubricating oils. TGA curves and micro-thermogravimetric curves can extract key characteristics such as volatile matter temperature, maximum weight loss rate temperature, and burnout temperature, and further calculate combustion characteristic indices or pyrolysis kinetic parameters. For example, in coal combustion characteristic studies, TGA curves are often converted into mass loss rate curves to assess ignitability and combustion intensity. In the pyrolysis kinetic analysis of wood-based panels and other combustible materials, mechanistic functions are typically selected based on TGA curves and micro-thermogravimetric curves, and integral or differential methods are used to fit the reaction mechanism and kinetic parameters.
[0004] Currently, commonly used methods for obtaining thermogravimetric curves and micro-thermogravimetric curves include: I. Direct experimental data (measured by instruments) By using a thermogravimetric analyzer (TGA) under programmed temperature control (linear heating or isothermal) and a set atmosphere, the change in sample mass with temperature or time is continuously measured, and the raw temperature-mass (or mass percentage) data is directly recorded. The thermogravimetric curve (TG curve) is then plotted. The micro-thermogravimetric curve (DTG curve) is obtained by taking the first derivative of the thermogravimetric curve with respect to temperature or time.
[0005] II. Other methods The data processing and conversion method utilizes existing thermogravimetric curve data (such as ASCII tables) to perform data smoothing, differentiation, and unit conversion to obtain micro-thermogravimetric curves.
[0006] For example: Existing patent document CN113945479B discloses a thermogravimetric analysis (TGA) test method based on a high-temperature furnace. This method includes four steps: mixing and preparing initial reactants, conducting a control group test, conducting an experimental group test, and performing error correction to obtain the actual mass residual rate (MR) of the high-temperature furnace TGA test. The present invention's high-temperature furnace-based TGA test method, in calculating the actual mass residual rate (MR), introduces an error correction value EC by setting a control group. This error correction corrects the uncorrected mass residual rate (MR0) of the experimental group, making the final test data closer to the true value. This overcomes the limitation of high-temperature furnaces not being able to weigh reactants in real time, thereby improving the accuracy of high-temperature furnace-based TGA tests.
[0007] Existing patent document two, publication number CN115876635A, discloses a method for analyzing industrial components of biomass based on thermogravimetric analysis (TGA). The analytical method includes the following steps: S1: preparing a biomass sample; S2: preparing an air-dried substrate; S3: preparing a dried biomass sample substrate; S4: treating the biomass sample by volatilization; S5: calculating moisture, volatile matter, fixed carbon, and ash content. This invention uses thermogravimetric analysis for industrial analysis of biomass materials. Based on the combustion characteristics of biomass materials, different temperature rise curves and gas atmospheres are set for different stages. The proportions of volatile matter, fixed carbon, ash, and moisture in the biomass material can be determined through mass changes in the thermogravimetric analysis.
[0008] The above method has the following limitations, as detailed below: First, all of these require actual samples and thermogravimetric instruments to conduct thermogravimetric experiments and obtain thermogravimetric curves. This process is time-consuming and material-intensive, and is sensitive to experimental conditions, which is not conducive to high-throughput material screening and early concept design. Second, existing virtual laboratories are mostly based on existing thermogravimetric data, focusing on simulating the effects of heating rate and sample amount on curve shape, but still require existing curves or preset parameters. Existing curves also need to be obtained through thermogravimetric experiments, and preset parameters are mostly based on human experience. Summary of the Invention
[0009] This invention provides a method, apparatus, electronic device, and storage medium for simulating the pyrolysis behavior of materials based on chemical and elemental composition. It overcomes the shortcomings of the prior art and effectively solves the problems that existing methods for obtaining thermogravimetric curves and micro-thermogravimetric curves require thermogravimetric instruments for thermogravimetric experiments, which is time-consuming, material-intensive, and sensitive to experimental conditions.
[0010] One of the technical solutions of this invention is achieved through the following measures: a method for simulating the pyrolysis behavior of materials based on chemical and elemental composition, comprising: Input the basic data of the target sample into the composition-parameter mapping model to obtain the predicted parameters of each thermal decomposition stage. The composition-parameter mapping model is obtained by deep learning using multiple samples. Each sample includes the basic data of the representative sample and the measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The basic data includes chemical composition, elemental composition and thermogravimetric experimental conditions. The measured parameters and predicted parameters of each thermal decomposition stage include peak parameters and apparent kinetic parameters. Peak parameters include peak area, peak temperature and peak width. Apparent kinetic parameters include apparent activation energy and pre-exponential factor. Simulated thermogravimetric curves and simulated differential thermogravimetric curves are generated by using the predicted parameters of each thermal decomposition stage to show the change of the target sample mass with temperature or time.
[0011] The following are further optimizations and / or improvements to the above-mentioned technical solution: The construction process of the above composition-parameter mapping model includes: Multiple samples were obtained and divided into training and testing sets according to the proportion. Each sample included basic data of representative samples and measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The pre-defined network model is trained using K-fold cross-validation based on the training set. Training ends when the training stopping condition is met, resulting in a composition-parameter mapping model. The trained composition-parameter mapping model is tested using a test set, the model parameters of the composition-parameter mapping model are optimized, and a composition-parameter mapping model that meets the test evaluation requirements is output.
[0012] The aforementioned preset network model is a multilayer feedforward artificial neural network.
[0013] The above-mentioned generation of simulated thermogravimetric curves and simulated differential thermogravimetric curves of target sample mass change with temperature or time using predicted parameters of each thermal decomposition stage includes: Construct Gaussian peaks or modified Gaussian peaks for each thermal decomposition stage based on the predicted parameters for each thermal decomposition stage. Normalize the peak area according to the total mass of the target sample to obtain the mass loss contribution at each stage. By superimposing and integrating each peak on the temperature axis or time axis, a thermogravimetric curve of mass change with temperature or time is obtained. Numerical differentiation of the thermogravimetric curve yields the differential thermogravimetric curve.
[0014] The above-mentioned generation of simulated thermogravimetric curves and simulated differential thermogravimetric curves of target sample mass change with temperature or time using predicted parameters of each thermal decomposition stage includes: The following formula is used to calculate the predicted parameters for each thermal decomposition stage to generate a simulated differential thermogravimetric curve of the target sample mass as a function of temperature or time. in, y To simulate the differential thermogravimetric value; exp[] is an exponential function with the natural constant e as the base; π Pi is a constant, i.e., the ratio of π to π. w Peak width; T p Peak temperature; T Temperature is the variable; A is the peak intensity coefficient; D is the peak shape correction factor. Numerical integration is performed on the simulated differential thermogravimetric curve to obtain the corresponding simulated thermogravimetric curve.
[0015] The above also includes calculating derived indices after generating simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass as a function of temperature or time using predicted parameters of each thermal decomposition stage. These derived indices include combustion characteristic index, pyrolysis stage temperature range, and maximum weight loss rate.
[0016] The second technical solution of the present invention is achieved through the following measures: a material pyrolysis behavior simulation device based on chemical and elemental composition, comprising: The parameter prediction unit inputs the basic data of the target sample into the composition-parameter mapping model to obtain the predicted parameters for each thermal decomposition stage. The composition-parameter mapping model is obtained by deep learning using multiple samples. Each sample includes the basic data of a representative sample and the measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The basic data includes chemical composition, elemental composition and thermogravimetric experimental conditions. The measured parameters and predicted parameters for each thermal decomposition stage include peak parameters and apparent kinetic parameters. Peak parameters include peak area, peak temperature and peak width. Apparent kinetic parameters include apparent activation energy and pre-exponential factor. The curve simulation unit generates simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass as a function of temperature or time using predicted parameters of each thermal decomposition stage.
[0017] The following are further optimizations and / or improvements to the above-mentioned technical solution: The above-mentioned curve simulation unit includes: The first simulation module includes: Construct Gaussian peaks or modified Gaussian peaks for each thermal decomposition stage based on the predicted parameters for each thermal decomposition stage. Normalize the peak area according to the total mass of the target sample to obtain the mass loss contribution at each stage. By superimposing and integrating each peak on the temperature axis or time axis, a thermogravimetric curve of mass change with temperature or time is obtained. Numerical differentiation of the thermogravimetric curve yields the differential thermogravimetric curve; The second simulation module includes: The following formula is used to calculate the predicted parameters for each thermal decomposition stage to generate a simulated differential thermogravimetric curve of the target sample mass as a function of temperature or time. in, y To simulate the differential thermogravimetric value; exp[] is an exponential function with the natural constant e as the base; π Pi is a constant, i.e., the ratio of π to π. w Peak width; T p Peak temperature; T Temperature is the variable; A is the peak intensity coefficient; D is the peak shape correction factor. Numerical integration is performed on the simulated differential thermogravimetric curve to obtain the corresponding simulated thermogravimetric curve.
[0018] The third technical solution of the present invention is achieved through the following measures: an electronic device, including a processor and a memory, wherein the memory stores a computer program, which is loaded and executed by the processor to implement the steps in the material pyrolysis behavior simulation method based on chemical and elemental composition.
[0019] The fourth technical solution of the present invention is achieved by the following measures: a storage medium storing a computer program that can be read by a computer, the computer program being configured to execute steps such as those in a material pyrolysis behavior simulation method based on chemical and elemental composition when running.
[0020] This invention introduces deep learning to establish a predictive link between chemical composition and elemental composition and thermogravimetric curves (TGA). It establishes a reliable mapping from "chemical composition and elemental composition + thermogravimetric conditions" to "thermogravimetric curves." Without requiring actual thermogravimetric experiments, TGA and differential TGA curves can be simulated solely based on chemical composition, elemental composition, and experimental conditions. This breaks the limitation of traditional TGA curve acquisition relying on thermogravimetric experiments. Furthermore, in systems such as coal, wood, polymers, and lubricating oils, industrial and elemental analysis are already very mature, making it relatively easy and cost-effective to obtain chemical and elemental compositions. Eliminating the need for actual thermogravimetric experiments eliminates the cost and time required for such experiments, reducing the cost of obtaining TGA curves and improving efficiency. This provides stable support for rapid thermogravimetric data in materials design, performance evaluation, and safety analysis. Attached Figure Description
[0021] Appendix Figure 1 This is a schematic diagram of an implementation environment provided for an embodiment of the present invention.
[0022] Appendix Figure 2 A schematic diagram of the process for simulating the pyrolysis behavior of materials based on chemical and elemental composition, provided in an embodiment of the present invention.
[0023] Appendix Figure 3 This is a schematic diagram of the composition-parameter mapping model construction method provided in an embodiment of the present invention.
[0024] Appendix Figure 4 This is a schematic diagram illustrating the acquisition of peak shape correction factors according to an embodiment of the present invention.
[0025] Figure 5 is a schematic diagram showing the relationship between the heating rate and the peak temperature provided in the embodiment of the present invention, wherein (a) is a schematic diagram showing the relationship of the first peak, (b) is a schematic diagram showing the relationship of the second peak, and (c) is a schematic diagram showing the relationship of the third peak.
[0026] Appendix Figure 6 This is a schematic diagram illustrating the comparison results provided in an embodiment of the present invention.
[0027] Appendix Figure 7 A schematic diagram of the structure of a material pyrolysis behavior simulation device based on chemical and elemental composition provided in an embodiment of the present invention. Detailed Implementation
[0028] The present invention is not limited to the following embodiments, and the specific implementation can be determined according to the technical solution of the present invention and the actual situation.
[0029] Those skilled in the art will understand that, unless specifically stated otherwise, in the embodiments of the present invention, a "module" or "unit" refers to a computer program or part of a computer program that has a predetermined function and works together with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.
[0030] In addition, in the embodiments of the present invention, "multiple" refers to two or more, and "first" and "second" are used to distinguish descriptions and should not be construed as implying relative importance.
[0031] This invention provides a method for simulating the pyrolysis behavior of materials based on chemical and elemental composition, comprising: inputting basic data of a target sample into a composition-parameter mapping model to obtain predicted parameters for each thermal decomposition stage. The composition-parameter mapping model is obtained through deep learning using multiple samples. Each sample includes basic data of a representative sample and the identification information of measured parameters for each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The basic data includes chemical composition, elemental composition, and thermogravimetric experimental conditions. The measured and predicted parameters for each thermal decomposition stage include peak parameters and apparent kinetic parameters. Peak parameters include peak area, peak temperature, and peak width. Apparent kinetic parameters include apparent activation energy and pre-exponential factor. The predicted parameters for each thermal decomposition stage are used to generate simulated thermogravimetric curves and simulated differential thermogravimetric curves showing the change of the target sample mass with temperature or time.
[0032] The method provided in this embodiment of the invention may involve artificial intelligence (AI) technology and may be implemented based on artificial intelligence technology, such as using deep learning to train a corresponding model using samples.
[0033] Machine Learning (ML) is a multidisciplinary field involving probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory, among others. It specifically studies how computers can simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to endow computers with intelligence; its applications span all areas of artificial intelligence.
[0034] Deep learning (DL) specifically refers to machine learning based on deep neural network models and methods. It has developed from statistical machine learning, artificial neural network algorithms, and other algorithms, combined with the advancements in big data and computing power. The most important technical feature of deep learning is its ability to automatically extract features.
[0035] The aforementioned machine learning and deep learning typically include techniques such as neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and instructional learning.
[0036] In deep learning, the loss function is used to predict the target value by comparing the predicted value with the target value. This is done by updating the weight vector of each layer of the neural network based on the difference between the two values (usually with an initialization process before the first update, where parameters are pre-configured for each layer) until the network can predict the target value or a value very close to it. Therefore, deep learning requires pre-defining "how to compare the difference between the predicted value and the target value," which is the loss function.
[0037] As attached Figure 1 The diagram illustrates an implementation environment provided by an embodiment of the present invention. This implementation environment may include: training equipment and usage equipment.
[0038] Both the training equipment and the equipment used are computer devices; optionally, the computer device is a terminal device, such as a mobile phone, tablet computer, PC (Personal Computer) or other electronic devices; or, the computer device is a server, which can be a single server, a server cluster composed of multiple servers, or a cloud computing service center. This embodiment of the invention does not limit this.
[0039] Training equipment refers to computer equipment capable of training and learning neural networks. Optionally, the training equipment has the ability to acquire neural networks and train and learn them according to application requirements. For example, the training equipment acquires neural networks from other devices through a network and then trains them with training samples according to application requirements, so that the neural network has the ability to obtain predictive parameters for each thermal decomposition stage. Optionally, the training equipment has the ability to build neural networks. It can build neural networks itself according to application requirements and then train and learn them. For example, in order to obtain predictive parameters for each thermal decomposition stage based on the chemical composition, elemental composition, and thermogravimetric experimental conditions of the target sample, the training equipment builds a neural network itself and then trains and learns it with samples according to application requirements.
[0040] The device used refers to a computer device that has the requirement to use a neural network. Optionally, the device uses a neural network from other devices through a network according to the application requirements. For example, if the device has the requirement to obtain the prediction parameters of each thermal decomposition stage, it can obtain the neural network that has completed training and learning the prediction parameters of each thermal decomposition stage from other devices through a network, and use the neural network to predict the prediction parameters of each thermal decomposition stage.
[0041] Based on this, the technical solution of the present invention will be described and explained below with reference to several examples.
[0042] Example 1: As shown in the attached document Figure 2 As shown, this invention discloses a method for simulating the pyrolysis behavior of materials based on chemical and elemental composition, including: Step S110: Input the basic data of the target sample into the composition-parameter mapping model to obtain the predicted parameters of each thermal decomposition stage. The composition-parameter mapping model is obtained by deep learning using multiple samples. Each sample includes the basic data of the representative sample and the measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The basic data includes chemical composition, elemental composition and thermogravimetric experimental conditions. The measured parameters and predicted parameters of each thermal decomposition stage include peak parameters and apparent kinetic parameters. The peak parameters include peak expansion area, peak temperature and peak width. The apparent kinetic parameters include apparent activation energy and pre-exponential factor. Step S120: Using the predicted parameters of each thermal decomposition stage, generate simulated thermogravimetric curves and simulated differential thermogravimetric curves showing the change of the target sample mass with temperature or time.
[0043] In this embodiment, the basic data includes chemical composition, elemental composition, and thermogravimetric experimental conditions; the chemical composition includes volatile matter, fixed carbon, ash content, and moisture content, or the proportions of saturated matter, aromatic matter, resin matter, asphaltenes, or other indicators that can represent the distribution of components; the elemental composition includes the mass fractions of C, H, O, N, and S, and optionally the content of Cl or other metal elements; the thermogravimetric experimental conditions include the reference heating rate, starting temperature, ending temperature, atmosphere type (such as air, nitrogen, carbon dioxide, or water vapor), and reference sample amount.
[0044] It should also be noted that the basic data of the target sample is the input feature vector formed by encoding the basic data in numerical form.
[0045] This invention discloses a method for simulating the pyrolysis behavior of materials based on chemical and elemental composition. It introduces deep learning to establish a predictive link between chemical and elemental composition and thermogravimetric curves (TGA). A reliable mapping from "chemical and elemental composition + thermogravimetric conditions" to "thermogravimetric curve" is established. Without requiring actual thermogravimetric experiments, TGA and differential TGA curves can be simulated solely based on chemical composition, elemental composition, and experimental conditions. This overcomes the limitation of traditional TGA curve acquisition relying on thermogravimetric experiments. Furthermore, in systems such as coal, wood, polymers, and lubricating oils, industrial and elemental analysis are already very mature, making it relatively easy and cost-effective to obtain chemical and elemental compositions. Eliminating the need for actual thermogravimetric experiments eliminates the cost and time required for such experiments, reducing the cost of obtaining TGA curves and improving efficiency. This provides stable support for rapid thermogravimetric data in material design, performance evaluation, and safety analysis.
[0046] Example 2: As shown in the attached document Figure 3 As shown, the embodiments of the present invention are further optimizations of the above embodiments, wherein the construction process of the composition-parameter mapping model includes: Step S210: Obtain multiple samples and divide them into training set and test set according to the proportion. Each sample includes basic data of representative samples and measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. In this embodiment, each sample includes basic data of a representative sample and measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The basic data includes chemical composition, elemental composition, and thermogravimetric experimental conditions. After obtaining the basic data, it is encoded in numerical form to form an input feature vector. It should also be noted that the thermogravimetric curves or differential thermogravimetric curves of the representative samples are all obtained through actual thermogravimetric experiments.
[0047] Understandably, since differential thermogravimetric curves (TGAs) directly reflect the peak characteristics (peak temperature, peak area, peak width, etc.) of each thermal decomposition stage, the physical meaning of the parameters is clearer, and the separation of each thermal decomposition stage is higher, which is beneficial for parameter identification and model training, the measured parameters of each thermal decomposition stage in the differential TGAs are preferred. However, to improve the applicability of the method, this embodiment also allows the extraction of measured parameters of each thermal decomposition stage based on the TGAs. Specifically, the differential TGAs are obtained by numerically differentiating the TGAs, and then the measured parameters of each thermal decomposition stage are extracted, or the measured parameters of each thermal decomposition stage are directly derived by piecewise fitting of the TGAs.
[0048] Since the type of measured parameters input into the model for all samples is uniform, the mixed use of measured parameters from each thermal decomposition stage in the thermogravimetric curve and the measured parameters from each thermal decomposition stage in the differential thermogravimetric curve will not affect the model construction and prediction accuracy.
[0049] In this embodiment, the measured parameters of each thermal decomposition stage in the differential thermogravimetric curve of a representative sample are obtained, specifically including: Under a standardized experimental protocol, thermogravimetric experiments were conducted on representative samples, and data on the change of mass with temperature or time were recorded. The thermogravimetric curves were then processed, and numerical differentiation was performed to obtain differential thermogravimetric curves. The differential thermogravimetric curve is fitted with a modified multi-peak Gaussian fit to break down the overall curve into several peaks, each corresponding to a relatively independent thermal decomposition stage, and the peak parameters of each peak are obtained, including peak area, peak temperature, and peak width. For the temperature range corresponding to each peak, a fully automatic kinetic parameter inversion algorithm is used to obtain the apparent kinetic parameters based on the mass loss rate data. The apparent kinetic parameters include apparent activation energy and pre-exponential factor.
[0050] It should also be noted that the differential thermogravimetric curve is fitted with a modified multi-peak Gaussian fit, that is, the non-normal peaks are corrected by introducing a correction factor to improve the consistency between the simulated curve and the measured curve in terms of the initial decomposition temperature, peak temperature and termination decomposition temperature.
[0051] Step S220: Train the preset network model using K-fold cross-validation based on the training set. End the training when the training stopping condition is met to obtain the composition-parameter mapping model. The preset network model in this embodiment can be set as needed. For example, a multi-layer feedforward artificial neural network or a machine learning network model such as ensemble regression can be used.
[0052] During training, the measured parameters of each thermal decomposition stage in the samples serve as supervision signals. The model weights are updated by minimizing the error between the predicted and measured parameters. Furthermore, K-fold cross-validation is employed to evaluate the model's generalization ability across different material systems. Additionally, the number of samples can be increased or the model structure adjusted as needed.
[0053] Step S230: Test the trained composition-parameter mapping model using the test set, optimize the model parameters of the composition-parameter mapping model, and output the composition-parameter mapping model that meets the test evaluation requirements.
[0054] Furthermore, in this embodiment, the composition-parameter mapping model can also be self-calibrated and extended. That is, when new real thermogravimetric experimental data are accumulated, samples with the same or similar compositions can be compared, and the difference between the simulated curve and the real curve can be used as feedback to incrementally train the composition-parameter mapping model, thereby gradually improving the prediction accuracy of different material sample types and different thermogravimetric conditions.
[0055] Furthermore, a database can be established for all representative samples. As the types of materials in the database expand, the method in this embodiment can cover a variety of systems such as coal, wood, lubricating oil, plastics, and mixed solid waste, and build models applicable to various systems to form a universal thermogravimetric curve simulation platform.
[0056] Example 3: This embodiment of the invention is a further optimization of the above embodiments, wherein generating simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass changing with temperature or time using predicted parameters of each thermal decomposition stage may include: Step S311: Construct Gaussian peaks or modified Gaussian peaks for each thermal decomposition stage based on the predicted parameters for each thermal decomposition stage. The above-mentioned construction of Gaussian peaks or modified Gaussian peaks for each thermal decomposition stage based on the predicted parameters of each thermal decomposition stage can be achieved using existing software such as Python, MATLAB, and Origin. Step S312: Normalize the peak area according to the total mass of the target sample to obtain the mass loss contribution at each stage. Step S313: Superimpose and integrate each peak on the temperature axis or time axis to obtain a simulated thermogravimetric curve showing the change of mass with temperature or time. Step S314: Perform numerical differentiation on the thermogravimetric curve to obtain the simulated differential thermogravimetric curve.
[0057] Example 4: This embodiment of the invention is a further optimization of the above embodiments, wherein generating simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass change with temperature or time using predicted parameters of each thermal decomposition stage may further include: Step S321: Calculate the predicted parameters for each thermal decomposition stage using the following formula to generate a simulated differential thermogravimetric curve showing the change of the target sample mass with temperature or time. in, y To simulate the differential thermogravimetric value; exp[] is an exponential function with the natural constant e as the base; π Pi is a constant, i.e., the ratio of π to π. w The peak width is the temperature width corresponding to 0.607 times the peak height, representing the thermogravimetric reaction range of this component. T p The peak temperature is the temperature corresponding to the maximum weight loss rate, which indicates the position of the lost component peak in the simulated thermogravimetric curve. T A is the temperature variable; A is the peak intensity coefficient, which characterizes the peak intensity or maximum weight loss rate level of this thermal decomposition stage. Its value varies with the sample composition and pyrolysis behavior and is not a fixed constant; D is the peak shape correction factor, used to describe asymmetric peak shapes (such as tailing effects).
[0058] It should be noted that a unique A and D are assigned to each type of sample, which are obtained by fitting the differential thermogravimetric curves obtained from thermogravimetric experiments using one or more representative samples of the same type.
[0059] Step S322: Perform numerical integration on the simulated differential thermogravimetric curve to obtain the corresponding simulated thermogravimetric curve.
[0060] In the above embodiments, a peak temperature correction model can be used to correct the peak temperature in the peak parameters, and the predicted parameters for each thermal decomposition stage of the target sample can be obtained after correction. The peak temperature correction model is shown below: ; in, T p V is the peak temperature; V is the heating rate; m is the target sample amount. The baseline heating rate is (e.g., 10 °C / min). The reference sample amount (e.g., 10.0 mg); denoted as , where is the peak temperature of the target sample under reference conditions; M is the influence coefficient of the heating rate on the peak temperature; and N is the influence coefficient of the amount of target sample on the peak temperature.
[0061] Example 5: This embodiment of the invention is a further optimization of the above embodiment. It also includes generating simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass change with temperature or time using the predicted parameters of each thermal decomposition stage, and then calculating derived indices, including combustion characteristic index, pyrolysis stage temperature range and maximum weight loss rate.
[0062] The derived indices can be calculated based on simulated differential thermogravimetric curves or simulated thermogravimetric curves, with simulated thermogravimetric curves being the preferred method for extracting derived indices. Specifically: Maximum rate of weightlessness Determined directly from simulated thermogravimetric curves: , where W is the remaining mass or normalized mass (mass fraction) of the target sample at temperature T, and T is the temperature variable; peak temperature : Corresponding peak position of the simulated differential thermogravimetric curve: ; pyrolysis stage temperature range It can be determined by thresholding or peak separation, for example: The corresponding temperature range is the reaction range for that stage (wherein) To set a threshold, such as 1% to 5% of the maximum peak value). weightlessness The result is obtained by integrating the simulated differential thermogravimetric curve: ,in The starting temperature of this thermal decomposition stage Sample quality at that time The termination temperature of this thermal decomposition stage Sample quality at that time; Combustion characteristic index: Alternatively, a weighted approach may be used depending on the criteria used (this method does not limit the specific form).
[0063] The aforementioned derived indices can be obtained through the following methods: software methods: Origin (peak identification + integration), MATLAB, Python (NumPy / SciPy); algorithmic methods: numerical differentiation (central difference); peak search (extreme value judgment); numerical integration (trapezoidal method / Simpson method).
[0064] In this embodiment, derived indices are calculated to provide data support for subsequent combustion performance evaluation and material selection.
[0065] Example 6: This example uses copper sulfate pentahydrate as the target sample to verify the effectiveness of the present invention, specifically including: (1) Copper sulfate pentahydrate was selected as the target sample. A commercial thermogravimetric analyzer was used to perform a programmed temperature rise test under the set thermogravimetric experimental conditions in a nitrogen atmosphere to obtain the corresponding true thermogravimetric curve and true differential thermogravimetric curve. Specifically, the target sample was placed in a platinum sample pan with a sample amount of about 10 mg. The temperature range was 30–820 °C, and the heating rate was 5, 10, 15, and 20 °C / min. Then, the thermogravimetric experiment was performed.
[0066] (2) The differential thermogravimetric curve is fitted with a modified multi-peak Gaussian curve to split the overall curve into several peaks, each corresponding to a relatively independent thermal decomposition stage, and the peak parameters of each peak are obtained, including peak area, peak temperature and peak width; for the temperature range corresponding to each peak, the apparent kinetic parameters are obtained from the mass loss rate data using a fully automatic kinetic parameter inversion algorithm, including apparent activation energy and pre-exponential factor; in this step, the differential thermogravimetric curve obtained in the experiment is compared with the differential thermogravimetric curve after the modified multi-peak Gaussian fitting, as shown in the appendix. Figure 4 As shown, determine the peak shape correction factor D; (3) Determine the basic data of the target sample, including chemical composition, elemental composition and thermogravimetric experimental conditions, and input them into the composition-parameter mapping model to obtain the corresponding predicted parameters for each thermal decomposition stage; (4) Use the peak temperature correction model to correct the peak temperature in the peak parameters, and obtain the predicted parameters of each thermal decomposition stage of the target sample after correction. The peak temperature correction model is shown below: ; in, T p V is the peak temperature; V is the heating rate; m is the target sample amount. The baseline heating rate is (e.g., 10 °C / min). The reference sample amount (e.g., 10.0 mg); M represents the peak temperature of the target sample under reference conditions; M is the influence coefficient of the heating rate on the peak temperature; N is the influence coefficient of the amount of target sample on the peak temperature. The above-mentioned process of obtaining M and N is based on thermogravimetric experiments on a large number of samples, analyzing the linear relationship between the peak temperature of the differential thermogravimetric curve and the heating rate and sample amount. That is, by regressing the experimental data under different heating rates and sample amounts, the linear relationship between the peak temperature and the changes in heating rate and sample amount is established, and the corresponding coefficients are given, as shown in Figure 5.
[0067] (5) Determine the peak width based on the predicted parameters of each thermal decomposition stage corresponding to the target sample. w Peak temperature T p The peak intensity coefficient A and peak shape correction factor D (as shown in Table 1) will be substituted into the following formula to construct the peak shape of a single thermal decomposition stage and form a differential thermogravimetric curve.
[0068] .
[0069] Table 1. Parameter values of copper sulfate pentahydrate .
[0070] (6) Compare the simulated differential thermogravimetric curve with the actual differential thermogravimetric curve. The comparison results are shown in the appendix. Figure 6 As shown, the deviation between the two is within a reasonable range, therefore the method of the present invention is effective, and the simulated differential thermogravimetric curve is accurate.
[0071] Example 7: As attached Figure 7 As shown, this embodiment of the invention discloses a material pyrolysis behavior simulation device based on chemical and elemental composition, comprising: The parameter prediction unit inputs the basic data of the target sample into the composition-parameter mapping model to obtain the predicted parameters for each thermal decomposition stage. The composition-parameter mapping model is obtained by deep learning using multiple samples. Each sample includes the basic data of a representative sample and the measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The basic data includes chemical composition, elemental composition and thermogravimetric experimental conditions. The measured parameters and predicted parameters for each thermal decomposition stage include peak parameters and apparent kinetic parameters. Peak parameters include peak area, peak temperature and peak width. Apparent kinetic parameters include apparent activation energy and pre-exponential factor. The curve simulation unit generates simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass as a function of temperature or time using predicted parameters of each thermal decomposition stage.
[0072] The curve simulation unit includes: The first simulation module includes: Construct Gaussian peaks or modified Gaussian peaks for each thermal decomposition stage based on the predicted parameters for each thermal decomposition stage. Normalize the peak area according to the total mass of the target sample to obtain the mass loss contribution at each stage. By superimposing and integrating each peak on the temperature axis or time axis, a thermogravimetric curve of mass change with temperature or time is obtained. Numerical differentiation of the thermogravimetric curve yields the differential thermogravimetric curve.
[0073] The second simulation module includes: The following formula is used to calculate the predicted parameters for each thermal decomposition stage to generate a simulated differential thermogravimetric curve of the target sample mass as a function of temperature or time. in, y To simulate the differential thermogravimetric value; exp[] is an exponential function with the natural constant e as the base; π Pi is a constant, i.e., the ratio of π to π. w Peak width; T p Peak temperature; T Temperature is the variable; A is the peak intensity coefficient; D is the peak shape correction factor. Numerical integration is performed on the simulated differential thermogravimetric curve to obtain the corresponding simulated thermogravimetric curve.
[0074] Example 8: This embodiment of the invention discloses a storage medium storing a computer program that can be read by a computer. The computer program is configured to execute steps in a material pyrolysis behavior simulation method based on chemical and elemental composition when it runs.
[0075] The aforementioned storage media may include, but are not limited to, USB flash drives, read-only memory, portable hard drives, magnetic disks, optical disks, and other media capable of storing computer programs.
[0076] Example 9: This embodiment of the invention discloses an electronic device, including a processor and a memory, wherein the memory stores a computer program, which is loaded and executed by the processor to implement the steps in a method for simulating the pyrolysis behavior of materials based on chemical and elemental composition.
[0077] The processor described above can be a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an ASIC, an FPGA, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this invention. It can also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc. The memory can include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, portable hard drives, magnetic disks, or optical disks.
[0078] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.
[0079] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0080] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0081] The above content is only a specific embodiment of the present invention, which has strong adaptability and implementation effect. However, the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be covered within the protection scope of the present invention. Therefore, equivalent changes made in accordance with the claims of the present invention are still within the scope of the present invention.
Claims
1. A method for simulating the pyrolysis behavior of materials based on chemical and elemental composition, characterized in that, include: Input the basic data of the target sample into the composition-parameter mapping model to obtain the predicted parameters for each thermal decomposition stage. The composition-parameter mapping model is obtained by deep learning using multiple samples. Each sample includes the basic data of a representative sample and the measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The basic data includes chemical composition, elemental composition and thermogravimetric experimental conditions. The measured and predicted parameters for each thermal decomposition stage include peak parameters and apparent kinetic parameters. Peak parameters include peak area, peak temperature and peak width. Apparent kinetic parameters include apparent activation energy and pre-exponential factor. Simulated thermogravimetric curves and simulated differential thermogravimetric curves are generated by using the predicted parameters of each thermal decomposition stage to show the change of the target sample mass with temperature or time.
2. The method for simulating the pyrolysis behavior of materials based on chemical and elemental composition according to claim 1, characterized in that, The process of constructing the composition-parameter mapping model includes: Multiple samples were obtained and divided into training and testing sets according to the proportion. Each sample included basic data of representative samples and measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The pre-defined network model is trained using K-fold cross-validation based on the training set. Training ends when the training stopping condition is met, resulting in a composition-parameter mapping model. The trained composition-parameter mapping model is tested using a test set, the model parameters of the composition-parameter mapping model are optimized, and a composition-parameter mapping model that meets the test evaluation requirements is output.
3. The method for simulating the pyrolysis behavior of materials based on chemical and elemental composition according to claim 2, characterized in that, The preset network model is a multi-layer feedforward artificial neural network.
4. The method for simulating the pyrolysis behavior of materials based on chemical and elemental composition according to claim 1, 2, or 3, characterized in that, Simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass as a function of temperature or time are generated using predicted parameters for each thermal decomposition stage, including: Construct Gaussian peaks or modified Gaussian peaks for each thermal decomposition stage based on the predicted parameters for each thermal decomposition stage. The peak areas are normalized according to the total mass of the target sample to obtain the mass loss contribution at each stage. By superimposing and integrating each peak on the temperature axis or time axis, a thermogravimetric curve of mass change with temperature or time is obtained. Numerical differentiation of the thermogravimetric curve yields the differential thermogravimetric curve.
5. The method for simulating the pyrolysis behavior of materials based on chemical and elemental composition according to claim 1, 2, or 3, characterized in that, Simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass as a function of temperature or time are generated using predicted parameters for each thermal decomposition stage, including: The following formula is used to calculate the predicted parameters for each thermal decomposition stage to generate a simulated differential thermogravimetric curve of the target sample mass as a function of temperature or time. in, y To simulate the differential thermogravimetric value; exp[] is an exponential function with the natural constant e as the base; π Pi is a constant, i.e., the ratio of π to π. w Peak width; T p Peak temperature; T Temperature is the variable; A is the peak intensity coefficient; D is the peak shape correction factor. Numerical integration is performed on the simulated differential thermogravimetric curve to obtain the corresponding simulated thermogravimetric curve.
6. The method for simulating the pyrolysis behavior of materials based on chemical and elemental composition according to claim 1, 2, or 3, characterized in that, It also includes calculating derived indices after generating simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass as a function of temperature or time using predicted parameters of each thermal decomposition stage. These derived indices include combustion characteristic index, pyrolysis stage temperature range, and maximum weight loss rate.
7. A device for simulating the pyrolysis behavior of materials based on chemical and elemental composition, using the method described in any one of claims 1 to 6, characterized in that, include: The parameter prediction unit inputs the basic data of the target sample into the composition-parameter mapping model to obtain the predicted parameters for each thermal decomposition stage. The composition-parameter mapping model is obtained by deep learning using multiple samples. Each sample includes the basic data of a representative sample and the measured parameters of each thermal decomposition stage in the corresponding thermogravimetric curve or differential thermogravimetric curve. The basic data includes chemical composition, elemental composition and thermogravimetric experimental conditions. The measured parameters and predicted parameters for each thermal decomposition stage include peak parameters and apparent kinetic parameters. Peak parameters include peak area, peak temperature and peak width. Apparent kinetic parameters include apparent activation energy and pre-exponential factor. The curve simulation unit generates simulated thermogravimetric curves and simulated differential thermogravimetric curves of the target sample mass as a function of temperature or time using predicted parameters of each thermal decomposition stage.
8. The material pyrolysis behavior simulation device based on chemical and elemental composition according to claim 7, characterized in that, Curve simulation unit, including: The first simulation module includes: Construct Gaussian peaks or modified Gaussian peaks for each thermal decomposition stage based on the predicted parameters for each thermal decomposition stage. The peak areas are normalized according to the total mass of the target sample to obtain the mass loss contribution at each stage. By superimposing and integrating each peak on the temperature axis or time axis, a thermogravimetric curve of mass change with temperature or time is obtained. Numerical differentiation of the thermogravimetric curve yields the differential thermogravimetric curve; The second simulation module includes: The following formula is used to calculate the predicted parameters for each thermal decomposition stage to generate a simulated differential thermogravimetric curve of the target sample mass as a function of temperature or time. in, y To simulate the differential thermogravimetric value; exp[] is an exponential function with the natural constant e as the base; π Pi is a constant, i.e., the ratio of π to π. w Peak width; T p Peak temperature; T Temperature is the variable; A is the peak intensity coefficient; D is the peak shape correction factor. Numerical integration is performed on the simulated differential thermogravimetric curve to obtain the corresponding simulated thermogravimetric curve.
9. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, which is loaded and executed by the processor to implement the steps of the method as claimed in any one of claims 1 to 6.
10. A storage medium, characterized in that, The storage medium stores a computer program that can be read by a computer, the computer program being configured to execute the steps of the method as described in any one of claims 1 to 6 when it is run.