A method and system for sustainable aviation fuel kitchen waste oil traceability management

Abstract

The application discloses a kind of kitchen waste oil traceability management method and system for sustainable aviation fuel, it is related to the technical field of efficient energy-saving engineering management service.The method comprises the following steps: a collaborative network of waste oil generating mechanism, collection and transportation company, processing company and oil refinery is constructed; waste oil is regenerated into sustainable aviation fuel after being treated by each collaborative subject; when the waste oil generating mechanism generates the fourth delivery behavior of waste oil, a delivery account is generated; when the collection and transportation company receives the third entry behavior, the third entry information is audited; when the oil refinery receives the second entry behavior, the second entry information is audited; after the audit is passed, the second sampling report is uploaded, and the record of second production treatment is increased.The application realizes the whole-chain data traceability management of the regeneration process of kitchen waste oil from food waste to sustainable aviation fuel.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of high-efficiency energy-saving engineering management services, in particular to a kitchen waste oil traceability management method and system for sustainable aviation fuel. BACKGROUND

[0002] Sustainable aviation fuel (SAF) refers to aviation alternative fuel (non-petroleum-based) that meets both aviation safety and suitability standards and sustainability evaluation standards. SAF is compatible with existing aircraft and civil aviation infrastructure, and the carbon emissions in the whole life cycle are reduced by more than 10% compared with fossil-based aviation fuel. Only aviation alternative fuel that has passed sustainable certification can be regarded as SAF and can be used for compliance. According to the definition of the International Air Transport Association (IATA) and the International Renewable Energy Agency (IRENA), the core raw material of SAF must meet the following conditions: non-food biomass, waste property and low carbon emission. Kitchen waste oil (hereinafter referred to as kitchen waste oil) meets the above conditions and is the core raw material of SAF.

[0003] Based on the above, the traceability management of kitchen waste oil for sustainable aviation fuel is not simply process control, but a key hub connecting raw material compliance, process stability, policy compliance and market trust.

[0004] Therefore, there is an urgent need for a kitchen waste oil traceability management method and system for the production of sustainable aviation fuel. SUMMARY

[0005] The embodiments of the application provide a kitchen waste oil traceability management method and system for sustainable aviation fuel to solve the above technical problems.

[0006] According to one aspect of the embodiments of this application, a method for traceability management of kitchen waste oil for sustainable aviation fuel is provided, comprising: constructing a collaborative network of waste oil generating institutions, collection and transportation companies, processing companies, and refineries, wherein the control platform of the collaborative network synchronizes traceability data with each collaborative entity in real time; wherein the traceability data is data generated in each stage of waste oil collection and processing; after being processed by each collaborative entity, the waste oil is recycled from waste into sustainable aviation fuel; when the waste oil generating institution generates a fourth outbound action for waste oil, an outbound ledger is generated; when the collection and transportation company receives a third inbound action, the third inbound information is reviewed; after the review is passed, a record of the third production and processing and the third outbound information are added; when the processing company receives a first inbound action, the first inbound information is reviewed; after the review is passed, a first sampling inspection report is uploaded, and a record of the first production and processing and the first outbound information are added; when the refinery receives a second inbound action, the second inbound information is reviewed; after the review is passed, a second sampling inspection report is uploaded, and a record of the second production and processing is added.

[0007] Furthermore, the deployment of the graph neural network includes:

[0008] In the first processing stage, a lightweight graph neural network is embedded in a portable spectrometer to generate the substructure fingerprint in real time through a local feature extraction algorithm, and transmits it to the cloud database based on a lightweight compression protocol; in the second processing stage, a high-performance graph neural network is deployed in a chromatography-mass spectrometry device to analyze the graph structure fingerprint and predict the freezing point and density of sustainable aviation fuel products.

[0009] The graph neural network shares the node feature weights of the first processing stage and the second processing stage to establish a correlation model for cross-stage molecular structure evolution.

[0010] Furthermore, the generation of the graph structure fingerprint further includes:

[0011] Multimodal detection data integrating infrared spectroscopy, Raman spectroscopy, and gas chromatography-mass spectrometry are used to model the dynamic evolution of molecular structure in the first and second processing stages through a variational autoencoder combined with a graph convolutional network. Specifically, this includes:

[0012] The molecular bond breaking events such as ester bond cleavage and triglyceride hydrolysis in the first treatment stage;

[0013] The second processing stage includes double bond hydrogenation saturation, oxygen-containing compound deoxygenation, and functional group elimination events.

[0014] The dynamic evolution process is bound with the space-time dimension of the oil refining process parameters, and the hydrogen consumption control strategy of the hydrogenation reactor, the catalyst regeneration cycle and the isomerization reaction temperature interval are adjusted in real time through a model optimization algorithm, and the process adjustment instructions are encrypted and transmitted to the control center for processing.

[0015] According to another aspect of the embodiments of the present application, a kitchen waste oil traceability management system for sustainable aviation fuel is provided, comprising: a collaborative network construction unit for constructing a collaborative network of waste oil generating institutions, collection and transportation companies, processing companies and oil refineries, and a control center of the collaborative network synchronously transmitting traceability data to each collaborative subject in real time; wherein the traceability data is data generated in each link of waste oil collection and processing; the waste oil is regenerated into sustainable aviation fuel after being processed by each collaborative subject; a waste oil out-of-stock and transfer unit for generating an out-of-stock account when the waste oil generating institution generates a fourth out-of-stock behavior of the waste oil; auditing the third entry information when the collection and transportation company receives the third entry behavior; after the audit is passed, adding the third production processing record and the third entry information; a preprocessing unit for auditing the first entry information when the processing company receives the first entry behavior; after the audit is passed, uploading the first sampling report and adding the first production processing record and the first entry information; a conversion unit for auditing the second entry information when the oil refinery receives the second entry behavior; after the audit is passed, uploading the second sampling report and adding the second production processing record.

[0016] Based on the method and system provided by the present application, the following beneficial effects are achieved:

[0017] Each collaborative subject (including waste oil generating institutions, collection and transportation companies, processing companies and oil refineries) needs to generate an out-of-stock account, an entry audit record, a sampling report and a production processing record in the waste oil circulation process, forming a complete data chain from "waste oil generating institution waste oil out-of-stock" to "oil refinery entry conversion". For example, the out-of-stock behavior of the waste oil generating institution triggers the generation of the account, the collection and transportation company records the third production processing information after the audit is passed, the processing company uploads the sampling report and records the first production processing information after the audit is passed, and the oil refinery receives and audits again and supplements the second production processing record, ensuring that each link operation leaves a trace and avoiding data loss or discontinuity in the traditional process, providing traceable original data support for SAF raw material compliance verification. Through the real-time synchronization of the control center, each party of the waste oil generating institution, the collection and transportation company, the processing company and the oil refinery can obtain the key information (such as the out-of-stock quantity, the audit result and the processing progress) of the upstream and downstream links in real time, replacing the traditional manual transmission or isolated system storage mode, and ensuring the consistency and real-time checkability of the whole chain data.

[0018] The processing company and the refinery perform the first and second warehouse audits respectively, forming double quality control of upstream and downstream. The processing company audits the first warehouse information, and the refinery reviews the second warehouse information. The audit results are archived through the control platform. This ensures that the waste oil entering the SAF conversion process meets the phased processing standards, reducing the risk of unqualified raw materials flowing into the subsequent process. After each audit, the sampling report must be uploaded and the production processing information must be recorded, binding the quality responsibility to the specific subject. When there is a problem with the raw material quality, the problem can be quickly located through the record timestamp and subject association relationship of the control platform (for example, if the refinery audit fails, the first sampling report and production record of the processing company can be directly traced). This achieves technicalization of responsibility definition and traceability, avoiding the slow responsibility tracing problem in the traditional mode.

[0019] In summary, the application provides bottom support for the large-scale application of SAF and the carbon neutralization target of the aviation industry through the whole-chain data traceability management of the regeneration process of kitchen waste oil from food waste to sustainable aviation fuel. BRIEF DESCRIPTION OF DRAWINGS

[0020] The accompanying drawings, which are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and serve to explain the principles of the application, and do not limit the application in any way. In the drawings:

[0021] Figure 1 A flowchart of an optional kitchen waste oil traceability management method for sustainable aviation fuel according to an embodiment of the application;

[0022] Figure 2 An interaction flowchart between the waste oil generating mechanism, the collection and transportation company, the processing company and the refinery according to an optional embodiment of the application;

[0023] Figure 3 An interaction flowchart between the waste oil generating mechanism, the collection and transportation company, the processing company and the refinery according to another optional embodiment of the application;

[0024] Figure 4 A schematic diagram of an optional first production processing sub-chain and second production processing sub-chain implementing a responsibility traceback and compensation mechanism according to an embodiment of the application;

[0025] Figure 5 A structure diagram of an optional kitchen waste oil traceability management system for sustainable aviation fuel according to an embodiment of the application;

[0026] Figure 6 An interface diagram of an optional supplier management module of a kitchen waste oil traceability management system for sustainable aviation fuel according to an embodiment of the application;

[0027] Figure 7 An interface diagram of a warehouse management module of a kitchen waste oil traceability management system for sustainable aviation fuel according to an embodiment of the present application.

[0028] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0029] In order for those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings of the embodiments of the present application.

[0030] It should be noted that the terms "first", "second", and the like in the specification of the present application, the claims and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0031] At present, aviation fuel accounts for more than 97% of the total emissions of China's civil aviation industry, and electric and hydrogen energy aircraft cannot meet the emission reduction needs of the aviation industry in the short term. SAF is the better efficient energy-saving solution at present. SAF is a new direction and solution for the comprehensive utilization of existing resources and technologies in China. The development of SAF helps to promote the upgrading and innovation of chemical and energy technologies, and solves the problems of energy shortage and energy security.

[0032] Kitchen waste oil refers to waste edible animal and plant oil and wastewater containing edible animal and plant oil generated in activities other than daily life, such as catering services (including unit catering, hereinafter referred to as "catering services"), food production and processing, and food production and processing. Kitchen waste oil is the core raw material of SAF, which is a non-grain biomass waste, and meets the three requirements of the International Air Transport Association and the International Renewable Energy Agency for SAF raw materials: non-competitive: not competing for resources with grain production; waste property: from catering industry byproducts, with an annual global output of about 15 million tons; low carbon emissions: full life cycle carbon emissions ≥ 50% (actual 80%-85%). The chemical structure of kitchen waste oil is mainly composed of triglycerides, which has similar carbon chain length (C15-C18) to traditional aviation coal (petroleum-based hydrocarbons), providing a basis for subsequent conversion.

[0033] Currently, there are some technical solutions for the management of kitchen waste oil. For example, the patent for invention with publication number CN119168632A discloses a kitchen waste oil management system and method based on AI technology, wherein the system includes an oil-water separator, an intelligent pressure sensing base, a wireless communication module, an AI processing module, a terminal display APP and a cloud database. Through these components, the system can monitor the waste oil inventory in real time, automatically track the whereabouts of waste oil, and record equipment maintenance. The AI processing module dynamically adjusts the maintenance cycle by analyzing the sensor data and generates maintenance notifications.

[0034] However, the technical solution disclosed in the patent with publication number CN119168632A focuses on the detection of kitchen waste oil inventory and equipment maintenance management, and involves less in the monitoring and tracing of kitchen waste oil quality. The quality of kitchen waste oil from different sources often differs, and if not traced, some non-standard kitchen waste oil raw materials may enter the conversion link, thereby affecting the production safety and production effect of SAF.

[0035] Based on this, the applicant proposes a kitchen waste oil traceability management method and system for sustainable aviation fuel, aiming to manage the full-chain data traceability of kitchen waste oil from "food waste" to the regeneration process of "sustainable aviation fuel", and to provide bottom support for the large-scale application of SAF and the carbon neutralization target of the aviation industry.

[0036] According to an aspect of an embodiment of the present application, a kitchen waste oil traceability management method for sustainable aviation fuel is provided, as shown in Figure 1 The method comprises:

[0037] S101, a collaborative network of waste oil generating institutions, processing companies and oil refineries is constructed, and a control center of the collaborative network synchronizes traceability data to each collaborative subject in real time; wherein the traceability data is data generated in each link of waste oil collection and processing; the waste oil is regenerated from waste to sustainable aviation fuel after being processed by each collaborative subject;

[0038] Wherein, the waste oil generating institutions include but are not limited to restaurants, slaughterhouses and food factories, etc. which can generate kitchen waste oil, and the waste oil generating institutions can also be called front-end enterprises; each link includes but is not limited to contract signing, invoicing, reverse invoicing, warehouse-out, warehouse-in, production processing, weighing, transportation and sampling inspection;

[0039] S102, when the waste oil generating institution generates a fourth warehouse-out behavior on the waste oil, a warehouse-out account is generated; S103, when the processing company receives a first warehouse-in behavior, the first warehouse-in information is audited; after the audit is passed, the first sampling inspection report is uploaded, and the records of the first production processing and the first warehouse-out information are added;

[0040] Wherein, after uploading the first inspection report, the processing company side determines whether the supplier information (i.e. the information of the waste oil producing mechanism) exists; if it exists, the state of the first warehouse information is modified to completed; if it does not exist, the supplier information is automatically added, and the state of the first warehouse information is modified to completed.

[0041] S104, when the refinery receives the second warehouse behavior, the second warehouse information is audited; after the audit is passed, the second inspection report is uploaded, and the record of the second production processing is added.

[0042] Wherein, after uploading the second inspection report, the refinery side determines whether the supplier information (i.e. the information of the waste oil producing mechanism and the information of the processing company) exists; if it exists, the state of the second warehouse information is modified to completed; if it does not exist, the supplier information is automatically added, and the state of the second warehouse information is modified to completed.

[0043] Wherein, the supplier information can include but is not limited to basic information (including enterprise name, registered address and production address), contact information and contract information, etc.

[0044] Reference Figure 1 The kitchen waste oil traceability management method for sustainable aviation fuel provided by the application belongs to the technical field of high-efficiency energy-saving engineering management services.

[0045] It should be noted that the first production processing is also called pretreatment, and the first production processing includes but is not limited to heating, sedimentation, filtration, adsorption decolorization and deacidification and deodorization, etc. The second production processing is also called conversion processing, and the second production processing includes but is not limited to ester exchange reaction, hydrogenation treatment (including hydrogenation deoxidation, hydrogenation cracking and isomerization) and refining and separation, etc.

[0046] Based on the method, the method and the system provided by the application, the following beneficial effects are obtained:

[0047] Each collaborative subject (including waste oil generating mechanism, processing company and refinery) needs to generate out-of-stock account, in-stock audit record, sampling report and production processing record in the process of waste oil circulation, forming a complete data chain from "waste oil generating mechanism waste oil out-of-stock" to "refinery in-stock conversion". For example, the out-of-stock behavior of the waste oil generating mechanism triggers the generation of the account, and after the processing company passes the audit, it must upload the sampling report and record the first production processing information. When the refinery receives it, it audits again and supplements the record of the second production processing, ensuring that each link operation leaves a trace and avoiding data loss or discontinuity in the traditional process, providing traceable original data support for SAF raw material compliance verification. Through real-time synchronization of the above data by the control platform, the waste oil generating mechanism, processing company and refinery can obtain the key information (such as out-of-stock quantity, audit result and processing progress) of the upstream and downstream links in real time, replacing the traditional manual transmission or isolated system storage mode, and ensuring the consistency and real-time checkability of the whole chain data.

[0048] The processing company and the refinery perform first and second in-stock audits respectively, forming double quality control of upstream and downstream. The processing company audits the first in-stock information (such as waste oil appearance, preliminary impurities and basic physical properties), and the refinery reviews the second in-stock information (such as the state of the pretreated oil and fat and its adaptability to the conversion process), and the audit results are archived through the control platform. Thus, it is ensured that the waste oil entering the SAF conversion link meets the phased processing standards, reducing the risk of unqualified raw materials flowing into the subsequent process. After each audit, the sampling report and production processing information need to be uploaded, binding the quality responsibility to the specific subject. When there is a problem with the raw material quality, the problem link can be quickly located through the record timestamp and subject association relationship of the control platform (such as when the refinery audit fails, the first sampling report and production record of the processing company can be directly traced), realizing the technicalization of responsibility definition and traceability, and avoiding the slow problem of responsibility tracing in the traditional mode.

[0049] In summary, the present application provides bottom support for the large-scale application of SAF and the carbon neutralization goal of the aviation industry through the whole-chain data traceability management of the regeneration process of kitchen waste oil from "food waste" to "sustainable aviation fuel".

[0050] The "out-of-stock, in-stock audit and record addition" operation process of the present application establishes standardized data collection nodes for the whole process of converting kitchen waste oil into SAF. For example, the out-of-stock account of the waste oil generating mechanism needs to include core information such as waste oil source and quantity (to help confirm whether it meets the "waste attribute" certification requirements of SAF raw materials), and the records of the processing company and the refinery need to specify the processing technology and operation time. These data directly constitute the original evidence chain required for sustainable certification (corresponding to the core requirement of "SAF needs to pass sustainable certification").

[0051] As an optional solution, the collaborative network further includes a collection and transportation company;

[0052] When the third warehousing behavior is received by the collection and transportation company, the third warehousing information is audited; after the audit is passed, the record of the third production treatment and the third delivery information are added; if the audit is not passed, the account status of the waste oil generating mechanism is changed to rejected; wherein, the record of the third production treatment is also called transfer treatment.

[0053] The third production treatment includes but is not limited to removing large impurities in kitchen waste oil, performing preliminary settlement or centrifugation, and preliminarily separating oil and water.

[0054] Wherein, after the audit is passed, the collection and transportation company side judges whether the supplier information (i.e. the information of the waste oil generating mechanism) exists; if it exists, the state of the third warehousing information is modified to completed; if it does not exist, the supplier information is automatically added, and the state of the third warehousing information is modified to completed;

[0055] The first delivery information corresponds to the first delivery behavior; the third delivery information corresponds to the third delivery behavior.

[0056] The first warehousing behavior occurs based on the third delivery behavior or the delivery behavior; the first warehousing information is associated with the record of the first production treatment; the first delivery information is associated with the record of the first production treatment and the first sampling report; the first production treatment includes converting waste oil into intermediate products;

[0057] The second warehousing behavior occurs based on the first delivery behavior; the second warehousing information is associated with the record of the second production treatment and the second sampling report; the second production treatment includes converting intermediate products into sustainable aviation fuel.

[0058] Wherein, the intermediate product can include industrial grade mixed oil (Used Cooking Oil, UCO).

[0059] In some embodiments of the present application, in S103, if the first warehousing information is not passed, the account status of the waste oil generating mechanism is modified to rejected; or, the state of the third delivery information is modified to rejected; in S104, if the second warehousing information is not passed, the state of the first delivery information is modified to rejected.

[0060] The outbound account is divided into an outbound account for the collection and transportation company or an outbound account for the processing company according to the receiving object. Specifically, the outbound account for the collection and transportation company includes but is not limited to the serial number, date, information of the generating unit (including the restaurant name, address, unit price, weight, and oil type), and information of the receiving unit (including the license plate number, address, collection and transportation personnel, and enterprise name). The outbound account for the processing company includes but is not limited to the order number, order date, finished product information (including the production batch number, weight, and related certificate), buyer information (including the enterprise name and address), and seller information (enterprise name, address, driver, license plate number, and contact person).

[0061] The first production processing record and the second production processing record can include but are not limited to the raw material type, raw material weight, production batch number, production date, yield, detection standard, operator, quality inspector, supervisor, and production certificate.

[0062] The first inbound information, the second inbound information, and the third inbound information can include but are not limited to the serial number, supplier name, raw material type, raw material name, net weight, used weight, remaining weight, and inbound date.

[0063] The first outbound information and the third outbound information can include but are not limited to the order number, order date, finished product information, buyer information, and seller information.

[0064] As a specific embodiment, as shown in FIG. 1, it is an interaction flow chart among the waste oil generating mechanism, the collection and transportation company, the processing company, and the refinery. Figure 2

[0065] As a specific embodiment, as shown in FIG. 2, it is another interaction flow chart among the waste oil generating mechanism, the collection and transportation company, the processing company, and the refinery. Figure 3

[0066] As an optional solution, the traceability data is stored in a cloud database.

[0067] The traceability data is encrypted by a hash algorithm and connected in a private chain by a blockchain traceability strategy.

[0068] Each link is a blockchain node, supporting data verification and content block verification.

[0069] If the content of a node is modified, it cannot pass the blockchain verification method, unless the blockchain data of the entire database is modified.

[0070] The private chain connection refers to a private chain or alliance chain architecture (rather than a public chain), which has the following characteristics:

[0071] ​​Permission control: on-chain nodes (such as internal systems of enterprises, partners) need to be authorized to access, ensuring data privacy;

[0072] Data correlation: through the transaction hash, block height and other fields of the blockchain, the hash values of data at different stages are linked in time sequence to form a complete traceability chain.

[0073] Mapping with cloud database: each on-chain record corresponds to one or a batch of data in the cloud database, and is indexed and associated through a unique identifier (such as data ID, transaction ID) to realize the cooperation of on-chain verification and off-chain query. If the data in the cloud database is tampered with, the hash value on the chain will not change synchronously, and through regular or real-time verification (such as automatic comparison of hash by smart contract), data tampering can be quickly discovered and an alarm can be given.

[0074] In some embodiments of the present application, in the blockchain, a carbon reduction certificate (such as life cycle carbon footprint data) is bound to each node. During the storage process, if additional carbon reduction is generated due to environmental management, a tradable carbon credit is automatically generated. These credits can be used to offset storage costs or realized in the carbon market, forming a "environmental protection-cost" positive cycle. The traceability data is connected with the energy trading platform to realize real-time accounting of storage energy consumption costs. For example, SAF batches using the energy trading platform can obtain a lower rate, and the difference in fees is automatically settled by the smart contract, promoting the low-carbon transformation of the storage link.

[0075] As an optional solution, in each link, the deployed Internet of Things detection equipment is used to detect the quality parameters of waste oil in real time, and the detection results are synchronized to each collaborative subject through the control platform; wherein the quality parameters include indicators that directly affect the conversion process; the conversion process is the process of regenerating waste oil from waste into sustainable aviation fuel;

[0076] In other words, "conversion" refers to the whole process of chemical and physical modification of waste oil from waste to qualified aviation fuel, and the core is to convert the molecular structure of waste oil into hydrocarbon compounds that meet the performance requirements of aviation fuel through industrial processing.

[0077] Currently, there is a lack of real-time and accurate monitoring and recording of key quality indicators that affect the quality of SAF production, such as fatty acid composition, oxidation stability, heavy metal content, etc. The fatty acid composition of kitchen waste oil from different sources varies greatly, which will directly affect the reaction conditions and product quality in the subsequent SAF production process. However, existing technologies cannot effectively trace these subtle quality differences, making it difficult to make precise process adjustments according to the actual situation of the raw materials in the SAF production process.

[0078] Among them, the Internet of Things detection equipment includes a spectrum analyzer and a weighing device for real-time acquisition of weight data and molecular fingerprint data of waste oil.

[0079] Based on the embodiments provided in the present application, molecular level quality detection is realized through spectral analysis, and real-time calibration of raw material loss is realized by combining with weighing data, solving the problems of traditional traceability management data lag and rough detection, and improving the transparency and synergy efficiency of waste oil conversion process. The control center synchronizes data in real time, so that oil refineries, recycling companies, airlines and other main bodies can real-time master the quality of raw materials, avoid information silos, and provide real-time data support for subsequent value assessment and responsibility tracing.

[0080] In the embodiments of the present application, for the SAF raw material of kitchen waste oil, the molecular fingerprint data is associated with the molecular stability data. The molecular fingerprint data reflects the molecular structure and component information of the raw material, and the structure and component of the molecule determine its stability. For example, the molecular structure containing more unsaturated bonds has specific characteristic peaks in the molecular fingerprint, and such structure is relatively unstable and prone to oxidation and other reactions. By analyzing the molecular fingerprint data, the molecular stability characteristics of the raw material can be preliminarily judged; and the change of the molecular stability data will also be reflected in the molecular fingerprint data to a certain extent, such as the appearance of new characteristic peaks or the change of the intensity of the original peaks in the molecular fingerprint spectrum of the raw material after oxidation. The molecular fingerprint image is a direct and characteristic representation of the molecular structure. Different molecules have unique fingerprint images, just like human fingerprints, which have uniqueness and specificity. Through the analysis of the fingerprint image, different molecules can be accurately identified and distinguished, and the structure of the molecule can be judged, and whether there are impurities or isomers can be judged, which provides a reliable basis for quality control and analysis. In practical application, fingerprint image analysis can realize automation and rapid detection. Using computer vision and machine learning technology, a large number of molecular fingerprint images can be quickly processed, improving the detection efficiency and accuracy. At the same time, the fingerprint image can also be combined with other analysis techniques such as spectral analysis, chromatographic analysis, etc., to form a more comprehensive analysis method, providing strong technical support for chemical production.

[0081] The weight data and the molecular fingerprint data are stored in a cloud database and a blockchain for phase-by-phase tracing;

[0082] As an optional solution, the molecular fingerprint data is generated through phase-by-phase detection, including:

[0083] In the first processing stage, the key physicochemical parameters and functional group characteristics of waste oil are identified based on a spectral detection strategy, specifically including acid value, moisture content, metal ion content, triglyceride concentration and functional group characteristics of impurity molecules, to generate a substructure fingerprint containing functional group existence, oil component proportion and impurity molecular structure identification; wherein the spectral detection strategy includes identification by an infrared spectrometer and / or identification by a Raman spectrometer;

[0084] In the second processing stage, the molecular structure of the waste oil is analyzed based on a mass spectrometry strategy, the fatty acid composition, carbon chain length distribution, double bond position and branched structure are mapped into a graph structure with atoms as nodes and chemical bonds as edges, graph neural network is used to extract molecular skeleton connection features and functional group spatial distribution features, and a graph structure fingerprint is generated. The mass spectrometry strategy includes analysis by gas chromatography-mass spectrometry and / or analysis by liquid chromatography-mass spectrometry.

[0085] It should be noted that the first processing stage is also referred to as a pretreatment stage, and the second processing stage is also referred to as a conversion stage.

[0086] In the embodiments of the present application, taking Raman spectrum imaging as an example, it is to irradiate the kitchen waste oil sample with laser, when the laser interacts with the molecules, Raman scattering light will be generated, different molecular structures will produce different characteristic Raman scattering spectra, by collecting and analyzing these spectra through the Raman spectrum detector, the spectrum reflecting the molecular structure information can be obtained. These spectra are unique like human fingerprints, which can be used to identify different substance components and their distribution.

[0087] Mass spectrometry imaging technology is to ionize the sample, separate and detect according to the mass-to-charge ratio of the ions, and also can obtain the characteristic information of the molecules and the spatial distribution image of the molecules in the sample.

[0088] During the pretreatment and conversion of the kitchen waste oil, the molecular fingerprint usually changes, and this change also has important monitoring and tracing significance, because:

[0089] Reflecting the processing and quality: The changes in molecular fingerprints can reflect the chemical reactions and quality changes of kitchen waste oil during the pretreatment and conversion process. For example, during the pretreatment process, by monitoring the molecular fingerprint, the effects of removing impurities, water, and adjusting the molecular structure of fatty acids can be understood. During the conversion to SAF, changes in the molecular fingerprint can indicate the degree of reaction and the quality of the product, helping operators to adjust process parameters in a timely manner to ensure that the product meets quality standards. Achieve accurate quality control: Using molecular fingerprint technology can accurately monitor each processing link. Once the molecular fingerprint changes abnormally, the problem can be quickly located, and appropriate measures can be taken to optimize and improve. This helps to improve the controllability of the entire production process and the stability of product quality, reducing the increase in costs and waste of resources caused by quality fluctuations. Establish a quality traceability system: Molecular fingerprint technology can establish a complete quality traceability system for the production of SAF. From the procurement of raw materials to the delivery of the final product, the molecular fingerprint data of each link can be recorded and saved. Through the analysis of these data, the root cause of product quality problems can be traced, and strong technical support is provided for product quality certification and market supervision. That is, the changes in the molecular fingerprint of kitchen waste oil during pretreatment and conversion provide key information for process monitoring, quality control, and traceability, making molecular fingerprint technology an important application in SAF production and a reliable and effective technical means.

[0090] Based on the embodiments provided in this application, the first stage identifies key parameters such as acid value and metal ions through spectroscopy, and the second stage analyzes molecular structure details (such as carbon chain length and double bond position) through mass spectrometry and graph neural networks, forming a hierarchical detection system from "functional group existence" to "molecular spatial structure". Compared with a single detection method, it can more accurately locate molecular-level defects (such as ester bond rupture and abnormal carbon chain cleavage) during raw material conversion. Substructure fingerprints serve quality control in the pretreatment stage, and graph structure fingerprints serve deep analysis in the hydrogenation conversion stage. The phased detection strategy matches the "pretreatment-conversion" process logic of waste oil to SAF, improving detection efficiency and relevance.

[0091] Specifically, substructure fingerprints are used for quality control in the preprocessing stage, which can quickly identify and characterize specific substructures in molecules. In the preprocessing stage, the main purpose is to preliminarily screen and quality control a large number of compounds, and substructure fingerprints can efficiently find compounds that do not meet the requirements, such as containing specific impurity substructures or not meeting the basic structure requirements, thereby quickly excluding these unqualified samples and improving the efficiency of subsequent processing. The advantages of substructure fingerprints include: substructure fingerprints (such as MACCS keys, Atom-Pair fingerprints) extract specific substructures (such as double bonds, hydroxyl groups, aromatic rings, etc.) through pre-set rules, which can directly match "quality control indicators" (for example: detect whether it contains sulfur / nitrogen heterocyclic ring - such structures are easy to cause catalyst poisoning). Compared with graph structure fingerprints, substructure fingerprints are faster in generation and comparison (low time complexity), suitable for high-throughput sample preliminary screening (such as tens of thousands of raw material samples per day, requiring millisecond-level response). In industrial scenarios, a "prohibited substructure list" is often pre-set (such as the prohibition of polycyclic aromatic hydrocarbons in SAF production), and substructure fingerprints can quickly trigger quality control alarms through Boolean logic (presence / absence).

[0092] Graph structure fingerprints are used for deep analysis in the hydrogenation stage, and graph structure fingerprints can more comprehensively and meticulously describe the overall structure and chemical bond information of molecules. In the hydrogenation stage, it is necessary to deeply understand the structural changes and interactions of molecules in the reaction, and graph structure fingerprints can accurately reflect the breaking and formation of chemical bonds, the rearrangement of atoms, etc. during the hydrogenation process, which helps to accurately analyze the reaction mechanism, predict the reaction products, and optimize the reaction conditions. The advantages of graph structure fingerprints include: graph structure fingerprints (such as ECFP, Weisfeiler-Lehman fingerprints) consider molecules as "graph models" (atoms as nodes, chemical bonds as edges), record atom types, connection modes, ring structures, etc. Information can distinguish between isomers (such as cis / trans double bonds have significant differences in hydrogenation activity). In the construction of reaction kinetics model and catalyst structure-activity relationship (SAR), graph structure fingerprints can be used as input features to capture "structure-activity" correlation (for example: predict the influence of specific branched structures on hydrogenation rate) through machine learning (such as GNN graph neural network). In the hydrogenation process, the change of molecular skeleton (such as ring opening of naphthenes, migration of double bonds) needs to rely on dynamic analysis of graph structure, and substructure fingerprints cannot provide such global information.

[0093] As an optional solution, molecular fingerprint data is further used to evaluate the reduction of raw material value from the molecular level, which specifically includes:

[0094] In the first processing stage, the loss of effective molecules in the processing process (including physical processing and chemical processing) is evaluated by the attenuation rate of triglyceride peaks, the retention rate of ester bond functional groups, and the increase of impurity characteristic peaks in the substructure fingerprint; the loss of effective molecules includes but is not limited to the volatilization of target components caused by distillation, and the loss of ester bond rupture caused by acid-base treatment;

[0095] In the second processing stage, the molecular level loss in the hydrodeoxygenation reaction is evaluated by the degree of carbon chain rupture, the residual proportion of double bonds, and the node abundance of oxygen-containing compounds in the graph structure fingerprint; wherein the molecular level loss in the hydrodeoxygenation reaction includes but is not limited to the residual of oxygen-containing compounds that have not been deoxygenated, and the carbon chain cleavage loss caused by side reactions;

[0096] The loss data obtained by evaluation is generated by a blockchain smart contract to generate a raw material value dynamic evaluation mechanism, and the raw material is priced at a premium or discount based on the molecular level loss quantification results.

[0097] Molecular level loss quantification refers to the change of the loss of SAF raw materials (kitchen waste oil in the embodiments of the present application) from the traditional statistics based on weight or volume to the evaluation of the reduction of raw material value from the molecular level. Since the quality and value of SAF raw materials are closely related to their molecular structure and composition, when the molecular structure of the raw material changes (such as oxidation, degradation, etc.) during storage, even if the weight or volume does not change significantly, the value of the raw material as SAF may have decreased. For example, after some fatty acid molecules in kitchen waste oil are oxidized, they may affect the efficiency and quality of their subsequent conversion into SAF, thereby causing the value of this part of the raw material to decrease. Molecular level value loss is to more accurately calculate the actual value loss of the raw material during storage by monitoring and evaluating the changes at the molecular level of the raw material. As an optional solution, the method further comprises:

[0098] The control center dynamically adjusts the sampling strategy of each collaborative subject according to the real-time monitored quality parameters, and triggers a data return instruction when the sampling is unqualified to feedback the waste oil quality defects;

[0099] Based on the embodiments provided in the present application, the molecular level loss parameters such as triglyceride attenuation rate and carbon chain rupture degree are converted into economic indicators, solving the problem of ignoring the quality difference of raw materials in the traditional "pricing by weight" and realizing a scientific pricing mechanism of "high-quality high price, low-quality discount" (for example: the ester bond retention rate decreases by 1%, and the unit price of the raw material decreases by 0.5%). The loss data is stored on the chain for evidence, and the smart contract automatically executes the pricing rules, avoiding human intervention, enhancing the trust of each subject in value evaluation, and promoting the standardization of the waste oil recycling industry chain.

[0100] The dynamic adjustment of the sampling strategy includes:

[0101] In the first processing stage, if the substructure fingerprint shows that the attenuation rate of triglyceride peak intensity exceeds the threshold value, the ester bond characteristic peak abnormally weakens, or the metal ion residue exceeds the standard, the ion exchange resin treatment effect review, vacuum distillation temperature parameter verification, or impurity removal process special sampling inspection is triggered;

[0102] In the second processing stage, if the graph structure fingerprint shows that the carbon chain distribution deviates from the target alkane chain length range, the double bond residual amount exceeds the standard, or the oxygen-containing compound node proportion is abnormal, the hydrogenation reaction pressure or temperature parameter review, catalyst activity detection, or regeneration process instruction is triggered;

[0103] Wherein, the sampling strategy needs to be cross-verified with the pretreatment process parameter records (such as distillation temperature, acid and alkali addition amount) and conversion process parameter records (such as catalyst use time, hydrogen consumption) stored in the blockchain to ensure the relevance of loss assessment and process operation.

[0104] In some embodiments of the present application, in the first processing stage (pretreatment stage):

[0105] Triglyceride peak intensity attenuation rate threshold: threshold setting: taking the triglyceride peak intensity before ester exchange reaction as the benchmark, the attenuation rate threshold is set to be ≥90% (i.e. the residual amount of triglyceride after reaction ≤10%). The triggering condition is that if the detected attenuation rate is <90% (such as 85%), it indicates that the ester exchange reaction is not complete, and the residual glycerol may block the hydrogenation catalyst pores, triggering the ion exchange resin treatment effect review (detecting whether the resin exchange capacity has decreased by more than 20%).

[0106] Abnormal weakening of ester bond characteristic peak: detection index: through infrared spectrum detection of ester carbonyl (1740 cm⁻¹) peak intensity, compared with the standard spectrum, the abnormal weakening amplitude is set to be >15%. The triggering condition is that if the ester bond peak intensity is weakened by more than 15%, it may be due to the decomposition of esters caused by too high vacuum distillation temperature (such as >190℃), triggering the distillation temperature parameter verification (checking whether the temperature control system deviates from the process window of 180±5℃).

[0107] Metal ion residue exceeds the standard: threshold setting: single ion (Na⁺ / K⁺) residue >0.8ppm or total metal ion >1.5ppm (reference the poisoning threshold of hydrogenation catalyst Ni-Mo / Al2O3). The triggering condition is that if it exceeds the standard, it indicates that the impurity removal in pretreatment is not complete, triggering the impurity removal process special sampling inspection (such as checking whether the amount of white clay adsorbent is less than 10% of the design value).

[0108] In the second processing stage (conversion stage):

[0109] Carbon chain distribution deviates from the target alkane chain length range: target range: the main components of SAF are C10-C18 alkanes, and the proportion of carbon chain length <10 or >18 set to >5% is deviated. Trigger condition: if the deviation exceeds the standard, it may be caused by insufficient hydrogenation reaction pressure (such as <5 MPa) leading to incomplete cracking, triggering reaction pressure parameter review (comparing with standard pressure 5.5-6.5 MPa).

[0110] Double bond residual exceeds the standard: detection index: the content of unsaturated bond (C=C) is detected by mass spectrometry, and the residual amount >3% is set as abnormal (SAF requires that the proportion of saturated alkanes is ≥97%). Trigger condition: if it exceeds the standard, it indicates that the catalyst activity decreases (such as Pt / Pd load decreases by 15%), triggering catalyst activity detection (measuring whether the hydrogen adsorption amount is lower than 20% of the initial value). Wherein, "Pt / Pd load decrease" refers to the actual content of active metals platinum (Pt) and palladium (Pd) in the catalyst decreases during use, leading to the phenomenon of decreased catalytic performance.

[0111] Abnormal proportion of oxygen-containing compound nodes: threshold setting: the proportion of oxygen-containing compound (such as carboxylic acid, aldehyde) molecular nodes >2% (target is ≤1%, to avoid affecting the fuel heat value). Trigger condition: if it is abnormal, it indicates that the hydrogenation deoxidization is not complete, triggering the regeneration process instruction (automatically starting the catalyst hydrogen reduction process, the temperature is raised to 300°C for 2 hours).

[0112] Based on the embodiments provided in the present application, through real-time triggering of targeted sampling inspection (such as checking the distillation temperature when the ester bond is abnormal) based on molecular fingerprints, compared with the traditional "fixed frequency sampling inspection", the "problem-oriented" quality control is realized, the invalid sampling inspection is reduced, and the production efficiency is improved. Cross verification of the sampling inspection results and the process parameters (such as ion exchange time, hydrogenation pressure) stored in the block chain can accurately locate the problem link (for example: when the metal ion exceeds the standard, the insufficient white clay dosage in the pretreatment stage is traced back), and the troubleshooting time of quality abnormality is shortened.

[0113] As an optional solution, the block chain nodes are divided into independent sub-chains according to the processing stages, and wherein:

[0114] The first production processing sub-chain stores the sub-structure fingerprint, the pretreatment process parameter record and the first sampling inspection report;

[0115] The second production processing sub-chain stores the graph structure fingerprint, the conversion process parameter record, the second sampling inspection report and the detection results of the sustainable aviation fuel product; the detection results include the distillation range, the freezing point and the density parameters;

[0116] When the detection results of the sustainable aviation fuel product do not meet the standard, the responsibility nodes of the first production processing sub-chain or the second production processing sub-chain are traced back through the cross-chain protocol, and the compensation mechanism is triggered based on the molecular fingerprint data.

[0117] As Figure 4 shown, it is a schematic diagram of a first production processing subchain and a second production processing subchain to implement a responsibility traceability and compensation mechanism;

[0118] In some embodiments of the present application, the cross-chain protocol can be implemented based on a notary mechanism, a cross-chain smart contract, or an Inter-Blockchain Communication Protocol (IBC protocol).

[0119] Specifically, the notary mechanism (Notary Schemes): a multi-signature notary node (such as an alliance composed of 3 third-party testing institutions) is deployed between the first production processing subchain and the second production processing subchain, which is responsible for listening to the key data (such as preprocessing parameters, conversion stage product detection results) of the two chains. When the SAF product detection result is unqualified, the notary node generates a cross-chain verification report through a threshold signature algorithm (Threshold ECDSA) to trigger the responsibility traceability process. When it is necessary to quickly locate the metal ion residue exceeding the standard in the preprocessing stage (such as abnormal ion exchange resin running time of the first production processing subchain), the notary node can directly call the preprocessing process parameter record for cross-validation.

[0120] Smart contract cross-chain calling (Hyperledger Cactus): cross-chain smart contracts are deployed on the first production processing subchain and the second production processing subchain respectively, and data interaction is realized through the cross-chain communication engine of Hyperledger Cactus. For example, when the second production processing subchain detects that the density of the SAF product is abnormal, the smart contract automatically calls the preprocessing process parameter record of the first production processing subchain to verify whether the distillation temperature exceeds the threshold value (such as > 250℃). When it is necessary to verify the correlation between the ester bond functional group retention rate in the preprocessing stage (first production processing subchain) and the carbon chain breaking degree in the conversion stage (second production processing subchain), the cross-chain smart contract can directly read the molecular fingerprint data of the two chains.

[0121] IBC protocol: a light client connection is established between the first production processing subchain and the second production processing subchain, and the verification request is transmitted through the data packet relay mechanism of the Cosmos IBC protocol. For example, the verification node of the second production processing subchain requests the ion exchange resin running time record of the preprocessing stage from the first production processing subchain through the IBC protocol, and completes the data verification based on the light client verification (only synchronizing block headers). When it is necessary to trace the correlation between the catalyst activity detection result in the conversion stage (second production processing subchain) and the metal ion residue in the preprocessing stage (first production processing subchain), the IBC protocol can efficiently transmit the verification request.

[0122] Based on the embodiments provided in the present application, the first sub-chain stores pretreatment data, and the second sub-chain stores conversion data. When the SAF product is unqualified (such as the alkane chain length not meeting the standard), the ester exchange reaction parameter abnormality in the pretreatment stage or the missing catalyst regeneration record in the conversion stage can be quickly located through the cross-chain protocol, solving the problem of “fuzzy multi-link responsibility” in traditional tracing. Based on the molecular fingerprint data (such as too high residual triglyceride), automatic compensation (such as requiring the recycling enterprise to bear the catalyst replacement cost) is triggered, reducing manual disputes and improving the collaborative efficiency of the industry chain.

[0123] Further, the method further comprises:

[0124] Based on the association data of the substructure fingerprint and the graph structure fingerprint, the waste oil batch is matched with the process demand of the refinery;

[0125] The high-acid-value waste oil is directed to the refinery with strong deacidification capability, the waste oil containing long carbon chains is directed to the refinery with high isomerization efficiency, and the minimum transportation cost is calculated in combination with real-time geographic location; wherein the waste oil with more than 15 carbon atoms in the carbon chain can be considered as waste oil containing long carbon chains;

[0126] The matching result and the transportation path data are written into the blockchain node.

[0127] Based on the embodiments provided in the present application, based on the association data of the substructure fingerprint (acid value, carbon chain length) and the graph structure fingerprint (carbon chain distribution, functional group characteristics), the waste oil is directed, the minimum transportation cost is calculated in combination with the geographic location, and the matching result is chained. Thus, the traditional “one-size-fits-all” waste oil distribution mode is changed, the waste oil is directed to the refinery most suitable for processing the type of raw material (for example: high-acid-value waste oil is preferentially distributed to the refinery equipped with strong acid adsorption resin) through the characteristics of acid value, carbon chain length and other characteristics at the molecular level, the acid value reduction efficiency in the pretreatment stage is improved, and the invalid process steps (such as repeated deacidification) are reduced. The real-time geographic location data (such as GPS positioning of the recycling point and the location of the refinery) are fused, the minimum transportation cost is calculated through the intelligent algorithm (such as Dijkstra shortest path), the transportation energy consumption is reduced compared with the traditional manual scheduling, and the low-carbon environmental protection requirement of SAF production is met. The matching result and the transportation path are chained, ensuring that the flow of waste oil is traceable (such as the distribution record, transportation mileage, and isomerization process parameters of the refinery of a batch of long carbon chain waste oil), providing full-link data support for subsequent value assessment and responsibility definition.

[0128] Further, the deployment of the graph neural network comprises:

[0129] In the first processing stage, a lightweight graph neural network is embedded in the portable spectrometer to generate substructure fingerprints in real time through a local feature extraction algorithm, and is transmitted to the cloud database based on a lightweight compression protocol; in the second processing stage, a high-performance graph neural network is deployed in the GC-MS device to analyze the graph structure fingerprint and predict the freezing point and density of the sustainable aviation fuel product; wherein the lightweight compression protocol is, for example, TensorFlowLite;

[0130] The graph neural network shares the node feature weight of the first processing stage and the second processing stage, and establishes a correlation model of the evolution of the molecular structure across the stages, such as the feature mapping of the triglyceride cracking path and the alkane generation path.

[0131] Based on the embodiments provided in the present application, the first processing stage embeds a lightweight graph neural network in the portable spectrometer to generate substructure fingerprints in real time and compresses the transmission; the second processing stage deploys a high-performance graph neural network in the mass spectrometer device, shares the node feature weight across the stages, and establishes a feature mapping model of triglyceride cracking and alkane generation. Thus, by using a lightweight model to process spectral data in real time on a portable device, on-site rapid detection of waste oil quality is achieved. Sharing the node feature weight of the two-stage model (such as correlating the ester bond strength detected in the first stage with the alkane freezing point predicted in the second stage), the cross-stage mapping of the evolution of the molecular structure is established, which improves the prediction accuracy of the product properties and provides early warning of process abnormalities (such as automatically adjusting the hydrogenation temperature when the freezing point is predicted to be out of specification). The lightweight model processes low-complexity substructure fingerprints (functional group existence), and the high-performance model analyzes high-complexity graph structure fingerprints (molecular spatial distribution), forming a hierarchical architecture of “edge device real-time preliminary screening + cloud device deep analysis”. Compared with a single high-performance model, the hierarchical architecture saves computing resources and adapts to the distributed device deployment requirements of the waste oil recycling scene.

[0132] Further, the generation of the graph structure fingerprint further comprises:

[0133] Fusing multi-modal detection data of infrared spectrum, Raman spectrum, and GC-MS data, modeling the dynamic evolution process of the molecular structure in the first processing stage and the second processing stage through a variational autoencoder combined with a graph convolution network, specifically including:

[0134] Ester bond cracking, triglyceride hydrolysis, and other molecular bond cracking events in the first processing stage;

[0135] Double bond hydrogenation saturation, deoxygenation of oxygen-containing compounds, and other bond saturation and functional group elimination events in the second processing stage;

[0136] The dynamic evolution process is bound with the space-time dimension of the oil refining process parameters, and the hydrogen consumption control strategy of the hydrogenation reactor, the catalyst regeneration cycle and the isomerization reaction temperature interval are adjusted in real time through a model optimization algorithm, and the process adjustment instructions are encrypted and transmitted to the control center. The oil refining process parameters can include but are not limited to distillation temperature, acid and base addition amount in the pretreatment stage, hydrogen pressure, catalyst type in the conversion stage.

[0137] Based on the embodiments provided in the present application, infrared, Raman, chromatography-mass spectrometry multi-modal data are fused, and through a variational autoencoder + graph convolutional network, molecular bond breaking (ester bond breaking, triglyceride hydrolysis) and bond saturation events (double bond hydrogenation, deoxidation) are modeled, and process parameters (distillation temperature, hydrogen pressure) are bound in space-time, and hydrogen consumption, catalyst regeneration cycle, etc. are adjusted in real time. Thus, the limitations of a single detection method (such as relying only on spectroscopy or mass spectrometry) are broken through, multi-modal data fusion (spectrum looks at functional groups, mass spectrometry looks at molecular structure, chromatography looks at component distribution) is realized, full-view modeling of molecular dynamic evolution (for example: capturing the degree of ester bond breaking and carbon chain isomerization path at the same time) is realized, and the accuracy of identifying molecular defects is significantly improved. The molecular bond breaking or saturation event is real-time bound with the process parameter (such as automatically increasing the distillation temperature by 5°C when the ester bond breaking rate is accelerated), and the hydrogen consumption (reducing invalid hydrogen consumption) and the catalyst regeneration cycle are dynamically adjusted through a model optimization algorithm (such as gradient descent), realizing the closed-loop control of "molecular-level change-process-level response", and significantly improving the SAF conversion rate. Through the graph convolutional network, the molecular bond evolution process can be visualized (such as using different colors to mark the ester bond breaking position and the double bond hydrogenation path), and the space-time correlation of the process parameters (such as incomplete deoxidation caused by hydrogen pressure fluctuation in a certain period) is combined, providing an interpretable quality anomaly analysis tool for engineers, and reducing the process troubleshooting time.

[0138] According to another aspect of the embodiments of the present application, a kitchen waste oil traceability management system for sustainable aviation fuel is provided, as shown in Figure 5 The system comprises:

[0139] A collaborative network construction unit 501 is configured to construct a collaborative network of waste oil generating institutions, collection and transportation companies, processing companies and oil refineries, and a control center of the collaborative network synchronizes traceability data to each collaborative subject in real time; wherein the traceability data is data generated in each link of waste oil collection and processing; the waste oil is regenerated into sustainable aviation fuel after being processed by each collaborative subject;

[0140] A waste oil delivery and transfer unit 502 is configured to generate a fourth delivery account when a waste oil generating institution generates a delivery behavior of waste oil; when a collection and transportation company receives a third entry behavior, the third entry information is audited; after the audit is passed, the third production processing record and the third delivery information are added;

[0141] The preprocessing unit 503 is configured to audit the first warehousing information when the processing company receives the first warehousing behavior; after the audit is passed, upload the first sampling report, and add the record of the first production processing and the first delivery information;

[0142] The conversion unit 504 is configured to audit the second warehousing information when the refinery receives the second warehousing behavior; after the audit is passed, upload the second sampling report, and add the record of the second production processing.

[0143] Reference Figure 5 The kitchen waste oil traceability management system for sustainable aviation fuel provided in the present application belongs to the technical field of efficient and energy-saving engineering management service.

[0144] As a specific embodiment, as shown in Figure 6 is an interface diagram of a supplier management module of a kitchen waste oil traceability management system for sustainable aviation fuel;

[0145] As a specific embodiment, as shown in Figure 7 is an interface diagram of a warehousing management module of another kitchen waste oil traceability management system for sustainable aviation fuel;

[0146] Optionally, in the present embodiment, the embodiments to be implemented by each unit module described above can refer to the above-mentioned method embodiments, which will not be described here.

[0147] In the above embodiments of the present application, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0148] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled in the art, without departing from the principle of the present application, a number of improvements and refinements can be made, which should be considered as the protection scope of the present application.

Claims

1. A method for the provenance management of kitchen waste oil for sustainable aviation fuel, characterized by, include: A collaborative network is constructed, comprising waste oil generating entities, collection and transportation companies, processing companies, and refineries. The control platform of this collaborative network synchronizes traceability data with each collaborative entity in real time. The traceability data consists of data generated at each stage of waste oil collection and processing. After processing by each collaborative entity, the waste oil is recycled from waste into sustainable aviation fuel. The traceability data is stored in a cloud database; the traceability data is encrypted using a hash algorithm and then privately chained using a blockchain traceability strategy; each link acts as a blockchain node, supporting data verification and content block verification, and each blockchain node is bound with a carbon emission reduction certificate; during the warehousing process, if additional carbon emission reductions are generated due to environmental management, carbon credits are automatically generated to offset warehousing costs or for trading in the carbon market; each link includes: contract signing, invoicing, reverse invoicing, outbound, inbound, production processing, weighing, transportation, and sampling inspection; In each stage, the deployed IoT detection equipment is used to detect the quality parameters of waste oil in real time, and the detection results are synchronized to each collaborating entity through the control platform; wherein, the IoT detection equipment includes a spectrometer and a weighing device, and the quality parameters include the weight data and molecular fingerprint data of waste oil collected by the spectrometer and the weighing device. The molecular fingerprint data is generated through staged detection, including: In the pretreatment stage, key physicochemical parameters and functional group characteristics of waste oil are identified based on a spectral detection strategy, specifically including acid value, moisture content, metal ion content, triglyceride concentration, and functional group characteristics of impurity molecules, generating a substructure fingerprint that includes the presence of functional groups, the proportion of oil components, and the structural identifier of impurity molecules; the spectral detection strategy includes identification using an infrared spectrometer and identification using a Raman spectrometer. During the conversion stage, the molecular structure of waste oil is analyzed using a mass spectrometry strategy. The fatty acid composition, carbon chain length distribution, double bond position, and branched structure are mapped into a graph structure with atoms as nodes and chemical bonds as edges. A graph neural network is used to extract the molecular skeleton connection features and functional group spatial distribution features to generate a graph structure fingerprint. The mass spectrometry analysis strategy includes analysis using gas chromatography-mass spectrometry and analysis using liquid chromatography-mass spectrometry. Based on the generated substructure fingerprint and graph structure fingerprint, the reduction in raw material value is assessed at the molecular level, including: In the pretreatment stage, the loss of effective molecules during the treatment process is evaluated by the attenuation rate of triglyceride peak intensity, the retention rate of ester bond functional groups, and the increase of impurity characteristic peaks in the substructure fingerprint. The loss of effective molecules includes the volatilization of target components due to distillation and the loss of ester bond breakage due to acid and alkali treatment. In the conversion stage, the molecular level loss of incomplete conversion in the hydrodeoxygenation reaction is evaluated by the degree of carbon chain breakage, the proportion of double bond residues and the abundance of oxygen-containing compound nodes in the graph structure fingerprint; wherein, the conversion is to convert the molecular structure of waste oil into hydrocarbon compounds meeting the performance requirements of aviation fuel; the molecular level loss of incomplete conversion in the hydrodeoxygenation reaction includes the residues of oxygen-containing compounds that are not deoxygenated and the carbon chain cracking loss caused by side reactions; The loss data obtained by evaluation is used to generate a raw material value dynamic evaluation mechanism through a blockchain smart contract, and the raw material is priced at a premium or discount based on the molecular level loss quantification results; Based on the association data of the substructure fingerprint and the graph structure fingerprint, the waste oil batch is matched with the process requirements of the refinery, including: high-acid-value waste oil is allocated to refineries with strong deacidification capacity, waste oil containing long carbon chains is allocated to refineries with high isomerization efficiency, and the minimum transportation cost is calculated in combination with real-time geographic location; wherein, waste oil with more than 15 carbon atoms in the carbon chain is determined as waste oil containing long carbon chains, and the matching result of the waste oil batch and the process requirements of the refinery and the transportation path data are written into the blockchain node; When the waste oil generating mechanism generates a fourth delivery behavior of waste oil, a delivery ledger is generated; When the collection and transportation company receives the third delivery behavior, the third delivery information is audited; after the audit is passed, the record of the third production treatment and the third delivery information are added; When the processing company receives the first delivery behavior, the first delivery information is audited; after the audit is passed, the first sampling report is uploaded, and the record of the first production treatment and the first delivery information are added; When the refinery receives the second delivery behavior, the second delivery information is audited; after the audit is passed, the second sampling report is uploaded, and the record of the second production treatment is added; Wherein, the blockchain node divides independent subchains according to the processing stage, wherein: The first production treatment subchain stores the substructure fingerprint, the pretreatment process parameter record and the first sampling report; The second production treatment subchain stores the graph structure fingerprint, the conversion process parameter record, the second sampling report and the detection result of the sustainable aviation fuel product, which includes the distillation range, freezing point and density parameters; When the detection result of the sustainable aviation fuel product does not meet the standard, the responsibility node of the first production treatment subchain or the second production treatment subchain is traced back through the cross-chain protocol, and a compensation mechanism is triggered based on the molecular fingerprint data; the cross-chain protocol is based on a notary mechanism, a cross-chain smart contract or an inter-chain communication protocol.

2. The method for tracking management of sustainable aviation fuel from kitchen waste oil according to claim 1, wherein, The first delivery information corresponds to the first delivery behavior; the third delivery information corresponds to the third delivery behavior; The first delivery behavior occurs based on the third delivery behavior or the fourth delivery behavior; the first delivery information is associated with the record of the first production treatment; the first delivery information is associated with the record of the first production treatment and the first sampling report; the first production treatment includes converting waste oil into intermediate products from waste; wherein, the intermediate products include industrial-grade mixed oil; The second warehousing behavior occurs based on the first warehousing behavior; the second warehousing information is associated with the record of the second production process and the second sampling report; and the second production process includes converting intermediate products into sustainable aviation fuel.

3. The method for tracing the source of the used cooking oil for sustainable aviation fuel according to claim 1, wherein the substructure fingerprint comprises MACCS keys and Atom-Pair fingerprints; the substructure fingerprint extracts specific substructures by a preset rule to match quality control indicators, and the specific substructures include double bonds, hydroxyl groups, and aromatic rings.

4. The method for tracing the source of the used cooking oil for sustainable aviation fuel according to claim 1, wherein the graph structure fingerprint comprises ECFP and Weisfeiler-Lehman fingerprints; the graph structure fingerprint regards a molecule as a graph model, records atom types, connection modes, and ring structure information, and distinguishes between isomers. The deployment of the graph neural network comprises: In the preprocessing stage, a lightweight graph neural network is embedded in a portable spectrometer, the substructure fingerprint is generated in real time by a local feature extraction algorithm, and the substructure fingerprint is transmitted to the cloud database based on a TensorFlowLite protocol; 5. The method for tracking management of sustainable aviation fuel from kitchen waste oil according to claim 1, wherein, In the conversion stage, the graph neural network is deployed in a gas chromatograph-mass spectrometer device to analyze the graph structure fingerprint and predict the freezing point and density of the sustainable aviation fuel product; the graph neural network shares node feature weights in the preprocessing stage and the conversion stage to establish a correlation model of the dynamic evolution of the molecular structure, and the correlation model includes feature mapping of the triglyceride cracking path and the alkane generation path. The generation of the graph structure fingerprint comprises: The multi-modal detection data of infrared spectroscopy, Raman spectroscopy, and gas chromatography-mass spectrometry data are fused, the dynamic evolution process of the molecular structure in the preprocessing stage and the conversion stage is modeled by a variational autoencoder combined with a graph convolution network, including: ester bond cracking, triglyceride hydrolysis molecular bond cracking events in the preprocessing stage; double bond hydrogenation saturation, deoxygenation of oxygen-containing compounds, and functional group elimination events in the conversion stage; the dynamic evolution process and the refining process parameters are bound in the space-time dimension to adjust the hydrogen consumption control strategy of the hydrogenation reactor, the catalyst regeneration cycle, and the isomerization reaction temperature range in real time, and the process adjustment instructions are encrypted and transmitted to the control center for execution, wherein the refining process parameters include distillation temperature, acid and base addition amount in the preprocessing stage, and hydrogen pressure, catalyst type in the conversion stage.

6. The method for tracking management of sustainable aviation fuel from kitchen waste oil according to claim 1, wherein, The method further comprises: based on the abnormal analysis results of the generated substructure fingerprint and the graph structure fingerprint, the control center dynamically adjusts the sampling strategy of each collaborative subject, and triggers a data return instruction when the sampling is unqualified to feedback the waste oil quality defects, specifically including: In the preprocessing stage, if the substructure fingerprint shows that the attenuation rate of the triglyceride peak intensity exceeds the threshold, the ester bond characteristic peak abnormally weakens, or the metal ion residue exceeds the standard, the ion exchange resin treatment effect is reviewed, the vacuum distillation temperature parameter is verified, or the impurity removal process is specially sampled.

7. The method for tracking management of sustainable aviation fuel from kitchen waste oil according to claim 1, wherein, ​ ​ In the conversion stage, if the graph structure fingerprint shows that the carbon chain distribution deviates from the target alkane chain length range, the residual amount of double bonds is excessive, or the proportion of oxygen-containing compound nodes is abnormal, the hydrogenation reaction pressure or temperature parameters are reviewed, the catalyst activity is detected, or the regeneration process instruction is triggered; wherein, if the proportion of carbon chain length <10 or >18 is >5%, it is determined that the carbon chain distribution deviates from the target alkane chain length range; Wherein, the sampling strategy needs to be cross-verified with the pretreatment process parameter record and the conversion process parameter record stored in the blockchain to ensure the relevance of the loss evaluation and the process operation; the pretreatment process parameter record includes distillation temperature, acid and base addition amount; the conversion process parameter record includes catalyst use time, hydrogen consumption.

8. A kitchen waste oil traceability management system for sustainable aviation fuel, characterized by, It includes: The cooperative network construction unit is used for constructing a cooperative network of waste oil generation mechanisms, collection and transportation companies, processing companies and oil refineries. A control center of the cooperative network synchronizes traceability data in real time to each cooperative subject. The traceability data is data generated in each link of waste oil collection and processing. Waste oil is regenerated into sustainable aviation fuel from waste. The traceability data is stored in a cloud database. The traceability data is encrypted by a hash algorithm and then connected in a private chain by a blockchain traceability strategy. Each link is a blockchain node, supports data verification, and binds a carbon emission reduction certificate to each blockchain node. During the storage process, if additional carbon emission reduction is generated due to environmental management, carbon credits for offsetting storage costs or carbon market transactions are automatically generated. Each link includes contract signing, invoicing, reverse invoicing, warehouse out, warehouse in, production processing, weighing, transportation and sampling. In each link, the quality parameters of waste oil are detected in real time by deploying Internet of Things detection equipment, and the detection results are synchronized to each cooperative subject through the control center. The Internet of Things detection equipment includes a spectrum analyzer and a weighing device, and the quality parameters include weight data and molecular fingerprint data of waste oil collected by the spectrum analyzer and the weighing device. The molecular fingerprint data is generated by stage detection, including: in the pretreatment stage, identifying key physicochemical parameters and functional group characteristics of waste oil based on a spectrum detection strategy, specifically including acid value, moisture content, metal ion content, triglyceride concentration and functional group characteristics of impurity molecules, generating a substructure fingerprint containing functional group existence, oil component ratio and impurity molecular structure identification; the spectrum detection strategy includes identifying by an infrared spectrum detector and identifying by a Raman spectrum detector; in the conversion stage, analyzing the molecular structure of waste oil based on a mass spectrum analysis strategy, mapping fatty acid composition, carbon chain length distribution, double bond position and branched chain structure into a graph structure with atoms as nodes and chemical bonds as edges, extracting molecular skeleton connection features and functional group spatial distribution features by a graph neural network, and generating a graph structure fingerprint; the mass spectrum analysis strategy includes analyzing by a gas chromatograph-mass spectrometer and analyzing by a liquid chromatograph-mass spectrometer; based on the generated substructure fingerprint and graph structure fingerprint, the reduction of raw material value is evaluated from the molecular level, including: in the pretreatment stage, evaluating the loss of effective molecules in the processing process by the triglyceride peak intensity attenuation rate, ester bond functional group retention rate and impurity characteristic peak increment in the substructure fingerprint; the loss of effective molecules includes volatilization of target components caused by distillation, ester bond breakage loss caused by acid-base treatment; in the conversion stage, evaluating the molecular level loss of incomplete conversion in the hydrodeoxygenation reaction by the carbon chain breakage degree, double bond residual proportion and oxygen-containing compound node abundance in the graph structure fingerprint; wherein, conversion is the conversion of the molecular structure of waste oil into hydrocarbon compounds meeting the performance requirements of aviation fuel.The molecular level loss of incomplete conversion in the hydrodeoxygenation reaction includes residual oxygen-containing compounds that are not deoxygenated and carbon chain cleavage loss caused by side reactions; the loss data obtained by evaluation is generated by a blockchain smart contract to generate a raw material value dynamic evaluation mechanism, and the raw material is priced at a premium or discount based on the molecular level loss quantification results; based on the association data of the substructure fingerprint and the graph structure fingerprint, the waste oil batch is matched with the process demand of the refinery, including: high-acid-value waste oil is allocated to refineries with strong deacidification capacity, waste oil containing long carbon chains is allocated to refineries with high isomerization efficiency, and the minimum transportation cost is calculated in combination with real-time geographic location; wherein, the waste oil with more than 15 carbon atoms in the carbon chain is determined as waste oil containing long carbon chains, the matching result of the waste oil batch and the process demand of the refinery and the transportation path data are written into the blockchain node; wherein, the blockchain node is divided into independent subchains according to the processing stage, wherein: the first production processing subchain stores the substructure fingerprint, the pretreatment process parameter record and the first sampling report; the second production processing subchain stores the graph structure fingerprint, the conversion process parameter record, the second sampling report and the detection result of the sustainable aviation fuel product, which includes the distillation range, freezing point and density parameters; when the detection result of the sustainable aviation fuel product does not meet the standard, the responsibility node is traced back to the first production processing subchain or the second production processing subchain through a cross-chain protocol, and a compensation mechanism is triggered based on the molecular fingerprint data; the cross-chain protocol is based on a notary mechanism, a cross-chain smart contract or an inter-chain communication protocol; The waste oil delivery and transfer unit is used to generate a delivery account when the waste oil generation mechanism generates a fourth delivery behavior of waste oil, to audit the third delivery information when the collection and transportation company receives the third delivery behavior, and to add a record of the third production treatment and the third delivery information after the audit is passed; The pretreatment unit is used to audit the first delivery information when the processing company receives the first delivery behavior, to upload the first sampling report after the audit is passed, and to add a record of the first production treatment and the first delivery information; The conversion unit is used to audit the second delivery information when the refinery receives the second delivery behavior, to upload the second sampling report after the audit is passed, and to add a record of the second production treatment.

Images