Aviation kerosene quality online detection and optimization system based on green hydrogen substitution technology
Through the online detection and optimization system of aviation kerosene quality based on green hydrogen replacement technology, the purity of sulfides, aromatics and hydrogen is monitored in real time, and combined with deep reinforcement learning algorithms to optimize process parameters, the real-time and optimization control problems of quality detection in aviation kerosene production are solved, effectively tracking and verification of carbon footprints, and production efficiency and environmental protection are improved.
Patent Information
- Application Number
- CN202510727045.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-03
- Publication Date
- 2025-08-29
AI Technical Summary
In the existing aviation kerosene production, there are insufficient real-time quality detection, insufficient optimization and regulation, and the difficulty in effectively tracking and verifying the carbon footprint of green hydrogen throughout the life cycle.
The online detection and optimization system of aviation kerosene quality based on green hydrogen replacement technology is adopted, including quality monitoring module, optimization control module and carbon footprint tracking and verification module. The quality monitoring module monitors the purity of sulfides, aromatics and hydrogen in real time through multi-spectral and electrochemical sensors. The optimization control module dynamically adjusts process parameters based on the deep reinforcement learning algorithm. The carbon footprint tracking module ensures data authenticity through life cycle evaluation and blockchain evidence storage.
Real-time monitoring and rapid response to aviation kerosene quality is achieved, the optimization control module has significantly improved the quality control level of the production process, and the carbon footprint tracking module provides reliable data support to help the green transformation of the aviation fuel industry.
Smart Images

Figure CN120562979A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of quality detection technology, and specifically to an online detection and optimization system for aviation kerosene quality based on green hydrogen substitution technology. Background Art
[0002] With the rapid development of the aviation transportation industry, the quality requirements for aviation kerosene are becoming increasingly stringent. In the traditional aviation kerosene production process, the control of impurity content such as sulfides and aromatics and the monitoring of hydrogen purity are key factors affecting product quality. However, existing quality detection methods are mostly offline analysis, which has problems such as long detection cycles and data lags, making it difficult to accurately reflect quality changes in the production process in real time. In addition, optimizing the process parameters in the aviation kerosene production process to improve sulfur removal efficiency, reduce energy consumption and carbon emissions is an urgent need to achieve green and sustainable development. Traditional optimization methods rely on experience or simple models, cannot adapt to complex changes in operating conditions, and have difficulty in achieving precise dynamic control.
[0003] In recent years, green hydrogen, as a clean energy carrier, has gradually been applied to the production of aviation kerosene. Replacing traditional hydrogen sources with green hydrogen not only reduces carbon emissions during the production process but also improves product quality. However, how to achieve real-time online detection and optimization of aviation kerosene quality in the context of green hydrogen substitution is an urgent problem to be solved. Existing technologies lack the means to track and verify the carbon footprint of green hydrogen during production, transportation, and use, and cannot provide reliable data support for carbon trading and certification. At the same time, during the green hydrogen substitution process, how to ensure the stability of hydrogen purity and optimize related process parameters to fully utilize the advantages of green hydrogen are also weak links in current technology.
[0004] While some technical solutions exist for jet fuel quality testing and process optimization, most suffer from the following limitations: First, insufficient detection accuracy and real-time performance prevent them from meeting the stringent quality control requirements of modern jet fuel production. Second, the optimization algorithms are not intelligent enough, making it difficult to achieve dynamic multi-objective optimization under complex operating conditions. Third, a lack of effective management and verification mechanisms for the carbon footprint of green hydrogen throughout its lifecycle prevents them from providing strong support for the green production of jet fuel. Therefore, developing an online jet fuel quality testing and optimization system based on green hydrogen substitution technology is crucial for improving jet fuel production quality, reducing energy consumption and carbon emissions, and promoting the sustainable development of the aviation fuel industry.
[0005] In order to solve the above defects, a technical solution is now provided. Summary of the Invention
[0006] The purpose of this invention is to solve the problems of insufficient real-time quality detection in existing aviation kerosene production, insufficient intelligence in optimization and control, and difficulty in effectively tracking and verifying the carbon footprint of green hydrogen throughout its life cycle, and to propose an online aviation kerosene quality detection and optimization system based on green hydrogen substitution technology.
[0007] The purpose of the present invention can be achieved through the following technical solutions: The online detection and optimization system for aviation kerosene quality based on green hydrogen substitution technology includes: The quality monitoring module uses multispectral and electrochemical sensors to monitor the purity of sulfides, aromatics, and hydrogen in real time. It combines fluid dynamics and chemical distribution analysis to determine the optimal installation location, ensuring data reliability. When the purity exceeds the standard, it triggers an alert and sends it to the optimization module. The optimization control module dynamically adjusts temperature, pressure, and green hydrogen flow based on a deep reinforcement learning algorithm. It also builds a multi-objective optimization model based on historical data to maximize sulfur removal efficiency while reducing energy consumption and carbon emissions, achieving closed-loop intelligent control. The carbon footprint tracking and verification module is used to integrate green hydrogen production, process energy consumption and transportation data, calculate the carbon emissions of the entire chain through life cycle assessment, use blockchain to store evidence to ensure that the data cannot be tampered with, generate visual reports and compliance certificates, and support carbon trading and certification.
[0008] Furthermore, the specific process of the quality monitoring module is as follows: Multispectral sensors are installed at optimal locations at the outlet of the hydrogenation reactor and downstream of the fractionator. They analyze the molecular characteristics of sulfides and aromatics using near-infrared spectroscopy and simultaneously collect hydrogen purity data detected by electrochemical sensors. Upload real-time data to the central processor, compare it with the preset aviation kerosene quality standards, and generate a quality deviation report;
[0009] If the detection value exceeds the threshold, an early warning signal is triggered and pushed to the intelligent optimization module.
[0010] Furthermore, the process of determining the optimal position in the quality monitoring module is as follows: Establish a three-dimensional CFD model of the hydrogenation reactor and fractionation tower, simulate the fluid velocity field and pressure field distribution, and calculate the flow uniformity index : ,in is the local flow velocity, is the average flow velocity, The closer it is to 1, the more uniform the flow is, and n is the number of flow velocity sampling points; Candidate location screening: Select and areas without eddy currents or dead zones; In situ sampling was performed at candidate locations and sulfide concentrations were analyzed by gas chromatography and aromatics concentration , construct the concentration gradient function: ,in Axial position The sulfide concentration at is the initial sulfide concentration, Axial position The aromatic hydrocarbon concentration at Initial aromatics concentration, is the axial distance, is the attenuation coefficient, which is determined by fitting the experimental data; choose The position with the largest rate of change ensures that the sensor is sensitive to composition fluctuations; Measure the temperature T and pressure P at the candidate locations and select the area that meets the sensor's operating range, including T∈[250℃,400℃] and P∈[2MPa,12MPa]; Evaluate sensor performance degradation rate through durability testing : ,in They are the initial sensitivity of the sensor and the sensitivity of the sensor after 100 hours of operation. position to ensure long-term stability; Construct the objective function F to comprehensively evaluate the candidate locations: , where F is the comprehensive score of the sensor installation location, is the sulfide concentration gradient, is the aromatic hydrocarbon concentration gradient, is the weight coefficient, which is set according to the process requirements; select the coordinate point that maximizes F.
[0011] Furthermore, the specific operation steps of the optimization control module are as follows: Real-time data integration: Receive the sulfur content (S real ), aromatic content (A real ) and hydrogen purity (H pure ), combined with the current temperature collected by the sensor (T now ), pressure (P now ), hydrogen-to-oil ratio (R now =F H2 / Q oil , where F H2 is the green hydrogen flow rate, Q oil is the crude oil flow rate), forming a real-time state vector ; Extract parameter combinations under similar working conditions from the historical process database The corresponding sulfur removal efficiency , energy consumption , carbon emissions , build a nonlinear regression model: ,in is the fitting coefficient, is the current reaction temperature, is the reference temperature, is the current reaction pressure, is the reference pressure, is the current hydrogen-to-oil ratio; Deep reinforcement learning algorithm optimization: state-action space definition, defining state space , action space ,in ; Through multi-objective reward function , balancing sulfur removal efficiency, energy consumption and carbon emissions: ,in is the current energy consumption, is the baseline energy consumption, is the current carbon emissions, is the baseline carbon emissions, , is the dynamic weight; Adopting a double-delayed deep deterministic policy gradient algorithm, the strategy is optimized through the Actor-Critic framework: Actor network, strategy function : Output action ; Critic network, value function : Evaluate action value ; The target function is to maximize the expected cumulative reward: , where D is the experience revisit pool, which stores the historical state-action-reward sequence; A dynamic learning rate adjustment strategy based on real-time performance feedback is introduced into the reinforcement learning algorithm. The learning rate is adaptively adjusted according to the change rate of sulfur removal efficiency and energy consumption fluctuations to improve the model convergence speed and stability. Real-time control instructions are issued and parameter adjustment calculations are performed: Actions output by the policy network , calculate the new process parameters:
[0012]
[0013]
[0014] in It is a limiting function to ensure that the parameters are within a safe range; Will Send to the hydrogenation reactor and fractionator controller, and collect a new round of real-time data at the same time , update the experience replay pool D to achieve closed-loop optimization.
[0015] Furthermore, the specific operation steps of introducing a dynamic learning rate adjustment strategy based on real-time performance feedback into the reinforcement learning algorithm in the optimization control module are as follows: Real-time monitoring of sulfur removal efficiency change rate and energy consumption fluctuations ; Dynamically adjust learning rate : ,in is the adjusted learning rate, is the initial learning rate, is the change rate of sulfur removal efficiency, For energy consumption fluctuations, is the baseline energy consumption; The updated learning rate is applied to the Actor-Critic network parameter update, prioritizing accelerating convergence under efficient working conditions while reducing the learning rate to maintain stability during abnormal fluctuations.
[0016] Furthermore, the specific operating steps of the carbon footprint tracking and verification module are as follows: Green hydrogen production source classification: Collect carbon emission factors based on green hydrogen source type: Wind power water electrolysis: carbon emission factors ,in is the carbon emission factor of green hydrogen generated by wind power and solar power water electrolysis, The power consumption for water electrolysis is is the grid emission factor, Producing carbon emissions for electrolyzers, is the hydrogen production; Biomass Gasification: Carbon Emission Factors ,in is the carbon emission factor of green hydrogen from biomass gasification, They are carbon emissions from material planting, transportation and chemical processes; Collect real-time energy consumption data E of hydrogenation reactor and fractionation tower reactor 、E fraction , combined with the equipment energy efficiency coefficient η equip , calculate the actual energy consumption: ; The CO2 concentration in the process exhaust gas is measured by infrared sensor, combined with the flow , calculate real-time emissions: ,in For real-time monitoring of carbon emissions, is the CO2 concentration in the exhaust gas, is the gas flow rate, is the CO2 density, t is the operating time; The full life cycle carbon footprint calculation, also known as the LCA model, has an LCA boundary definition: covering the entire chain from green hydrogen production → crude oil processing → aviation kerosene synthesis, and is calculated in stages: ; in For the full life cycle carbon footprint, Carbon emissions for green hydrogen production, ; For process energy consumption carbon emissions, , is the electricity emission factor; Carbon emissions from transporting raw materials, , are the transport distance and the emission factor per unit distance respectively; For a multi-product cogeneration system, carbon emissions are allocated based on mass ratio: ; The carbon footprint data of each batch of aviation kerosene is serialized into JSON format and a unique hash value is generated using the SHA-256 algorithm. , ,Will Written into a private blockchain network, with nodes including refineries, certification agencies, and airlines; Transaction verification is completed through the PBFT consensus mechanism to ensure that data cannot be tampered with; Use WebGL to build a 3D carbon footprint heat map and display it on a timeline Change trends, compared with international standards; generate PDF certificates that comply with the ISO 14064 standard, including the following fields: Carbon footprint value and confidence intervals; Blockchain transaction ID; Electronic signature from a third-party certification body; Use the verification mechanism to verify the carbon emission factor of the green hydrogen production process.
[0017] Furthermore, the carbon footprint tracking and verification module utilizes the verification mechanism to verify the carbon emission factor of the green hydrogen production process. The specific steps are as follows: Historical data storage and preprocessing: Stores historical green hydrogen production data, including daily renewable energy power generation G hist , electrolytic cell efficiency η hist , carbon emission factor , classified by timestamp and stored in the time series database; Outliers were removed and missing data were filled using linear interpolation; Multi-time window dynamic matching: real-time power generation of third parties Match with historical data of the same period and calculate similarity: ,in is the similarity between real-time and historical data, Real-time power generation for third parties, is the mean and standard deviation of historical data for the same period; Assign real-time data weight and historical data weight based on similarity : ; Multiple verification of carbon emission factors: Combine real-time and historical data to calculate verified carbon emission factors , ,in are the real-time grid emission factor and the historical grid emission factor respectively; if the deviation between the real-time data and the historical mean exceeds the threshold, that is , triggering the following checks: Check whether the third-party data interface is abnormal; Compare the equipment operation log to confirm whether there is a sudden change in the electrolyzer efficiency; If there are no abnormalities, the historical database is updated and marked as a verified new operating condition; Feedback and blockchain evidence: the verified carbon emission factor The weight distribution logic is written into the blockchain, and the hash value generation formula is expanded to: , add a new data verification level field in the digital certificate, based on the weight Label the confidence level.
[0018] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention uses multi-spectral and electrochemical sensors to monitor the purity of sulfides, aromatics, and hydrogen in real time, and combines fluid dynamics and chemical distribution analysis to determine the optimal installation position of the sensors, ensuring the accuracy and reliability of the data. When the detection value exceeds the threshold, the system can trigger an early warning and push it to the optimization module to achieve real-time monitoring and rapid response to the quality of aviation kerosene. The optimization control module dynamically adjusts the temperature, pressure, and green hydrogen flow rate based on a deep reinforcement learning algorithm, and builds a multi-objective optimization model based on historical data. It can maximize the sulfur removal efficiency while reducing energy consumption and carbon emissions, achieving closed-loop intelligent control, and significantly improving the quality control level of the aviation kerosene production process; (2) The carbon footprint tracking and verification module of the present invention can integrate green hydrogen production, process energy consumption and transportation data, calculate the carbon emissions of the entire chain through life cycle assessment, and use blockchain to store evidence to ensure that the data cannot be tampered with. The generated visual reports and compliance certificates support carbon trading and international certification, providing strong data support and certification basis for the green production of aviation kerosene. In addition, the carbon emission factors of the green hydrogen production link are multi-verified through the verification mechanism, further ensuring the authenticity and timeliness of the data source, and enhancing the reliability and credibility of the system; (3) This invention combines green hydrogen substitution technology with aviation kerosene quality detection and optimization, not only improving the quality and production efficiency of aviation kerosene, but also significantly reducing energy consumption and carbon emissions during the production process, contributing to the green transformation of the aviation fuel industry. Through intelligent detection and optimization methods, refined management of the aviation kerosene production process is achieved. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] In order to facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings; Figure 1 This is the overall system block diagram of the present invention. DETAILED DESCRIPTION
[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.
[0021] It should be understood that the terms “include” and “comprising” used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.
[0022] It should also be understood that the terminology used in this disclosure is for the purpose of describing specific embodiments only and is not intended to limit the disclosure. As used in this disclosure and the claims, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should further be understood that the term "and / or" as used in this disclosure and the claims refers to any and all possible combinations of one or more of the associated listed items, including and including these combinations.
[0023] like Figure 1As shown, the aviation kerosene quality online detection and optimization system based on green hydrogen replacement technology includes a quality monitoring module, an optimization control module, and a carbon footprint tracking and verification module; The quality monitoring module uses multispectral and electrochemical sensors to monitor the purity of sulfides, aromatics, and hydrogen in real time. It combines fluid dynamics and chemical distribution analysis to determine the optimal installation location, ensuring accurate and reliable data. When the data exceeds the standard, it triggers an alert and sends it to the optimization module. A multispectral sensor is installed at the optimal location at the outlet of the hydrogenation reactor and downstream of the fractionator. The molecular characteristics of sulfides and aromatics are analyzed using near-infrared spectroscopy, while hydrogen purity data detected by an electrochemical sensor is simultaneously collected. The process for determining the optimal location is as follows: Establish a three-dimensional CFD model of the hydrogenation reactor and fractionation tower, simulate the fluid velocity field and pressure field distribution, and calculate the flow uniformity index : ,in is the local flow velocity, is the average flow velocity, The closer it is to 1, the more uniform the flow is, and n is the number of flow velocity sampling points; Candidate location screening: Select and an area without eddy currents or dead zones (e.g. 3D to 5D downstream of the reactor outlet straight pipe section, where D is the pipe diameter); in-situ sampling is performed at the candidate location, and sulfide concentration is analyzed by gas chromatography. and aromatics concentration , construct the concentration gradient function: ,in Axial position The sulfide concentration at is the initial sulfide concentration, Axial position The aromatic concentration at Initial aromatics concentration, is the axial distance, is the attenuation coefficient, determined by fitting the experimental data; select The position with the largest rate of change ensures that the sensor is sensitive to composition fluctuations; Measure the temperature T and pressure P at the candidate locations and select areas that meet the sensor's operating range (e.g., T∈[250°C, 400°C], P∈[2MPa, 12MPa]); evaluate the sensor's performance attenuation rate through durability testing. : ,in They are the initial sensitivity of the sensor and the sensitivity of the sensor after 100 hours of operation. position to ensure long-term stability; Construct the objective function F to comprehensively evaluate the candidate locations: , where F is the comprehensive score of the sensor installation location, is the sulfide concentration gradient, is the aromatic hydrocarbon concentration gradient, is the weight coefficient, which is set according to the process requirements; select the coordinate point that maximizes F.
[0024] Real-time data is uploaded to the central processor and compared with the preset aviation kerosene quality standards (sulfur content <0.3%, aromatics <25%) to generate a quality deviation report; if the detection value exceeds the threshold, an early warning signal is triggered and pushed to the intelligent optimization module.
[0025] The optimization control module dynamically adjusts temperature, pressure, and green hydrogen flow based on a deep reinforcement learning algorithm (TD3). It also builds a multi-objective optimization model based on historical data to maximize sulfur removal efficiency while reducing energy consumption and carbon emissions, achieving closed-loop intelligent control. Real-time data integration: Receive the sulfur content (S real ), aromatic content (A real ) and hydrogen purity (H pure ), combined with the current temperature collected by the sensor (T now ), pressure (P now ), hydrogen-to-oil ratio (R now =F H2 / Q oil , where F H2 is the green hydrogen flow rate, Q oil is the crude oil flow rate), forming a real-time state vector ; Extract parameter combinations under similar working conditions from the historical process database The corresponding sulfur removal efficiency , energy consumption , carbon emissions , build a nonlinear regression model: ,in is the fitting coefficient, is the current reaction temperature, is the reference temperature (305°C), is the current reaction pressure, is the reference pressure (6.5 MPa), is the current hydrogen-to-oil ratio (kg / ton); Deep reinforcement learning algorithm optimization: state-action space definition, defining state space , action space ,in ; Through multi-objective reward function , balance sulfur removal efficiency, energy consumption and carbon emissions: ,in is the current energy consumption, is the baseline energy consumption, is the current carbon emissions, is the baseline carbon emissions, , are dynamic weights, with initial values of 0.6, 0.25, and 0.15; Adopting the double-delayed deep deterministic policy gradient (TD3) algorithm, the strategy is optimized through the Actor-Critic framework: Actor network (strategy function ): Output action ;Critic network (value function ): Evaluate the action value ; The objective function is to maximize the expected cumulative reward: , where D is the experience revisit pool, which stores the historical state-action-reward sequence.
[0026] A dynamic learning rate adjustment strategy based on real-time performance feedback is introduced into the reinforcement learning algorithm. According to the change rate of sulfur removal efficiency and energy consumption fluctuations, the learning rate is adaptively adjusted to improve the model convergence speed and stability. The specific process is as follows: Real-time monitoring of sulfur removal efficiency change rate and energy consumption fluctuations ; Dynamically adjust the learning rate : ,in is the adjusted learning rate, is the initial learning rate, is the change rate of sulfur removal efficiency, For energy consumption fluctuations, is the baseline energy consumption; The updated learning rate is applied to the Actor-Critic network parameter update, prioritizing accelerating convergence under efficient working conditions while reducing the learning rate to maintain stability during abnormal fluctuations.
[0027] Real-time control instructions are issued and parameter adjustment calculations are performed: Actions output by the policy network , calculate the new process parameters: ; ; ; in is a limiting function to ensure that the parameters are within a safe range; Send to the hydrogenation reactor and fractionator controller, and collect a new round of real-time data at the same time , update the experience replay pool D to achieve closed-loop optimization.
[0028] The carbon footprint tracking and verification module integrates green hydrogen production, process energy consumption, and transportation data, calculates carbon emissions across the entire supply chain through life cycle assessment (LCA), uses blockchain to store evidence to ensure data cannot be tampered with, generates visual reports and compliance certificates, and supports carbon trading and international certification. Green hydrogen production source classification: Collect carbon emission factors based on the green hydrogen source type (wind power water electrolysis or biomass gasification): Wind power water electrolysis: carbon emission factors ,in is the carbon emission factor of green hydrogen generated by wind power and solar power water electrolysis, is the power consumption for water electrolysis (kWh / kg H2), is the grid emission factor (kg CO2 / kWh, which is 0 if 100% renewable energy is used), Carbon emissions for electrolyzer production (kg CO2), is the hydrogen production (kg); Biomass Gasification: Carbon Emission Factors ,in is the carbon emission factor of green hydrogen from biomass gasification, They are carbon emissions from material cultivation (kg CO2), carbon emissions from transportation and carbon emissions from chemical processes.
[0029] Collect real-time energy consumption data of hydrogenation reactor and fractionation tower (E reactor 、E fraction , unit: kWh / ton), combined with the equipment energy efficiency coefficient η equip , calculate the actual energy consumption: ;Measure the CO2 concentration (CO2, unit: ppm) in the process exhaust gas through infrared sensor, combined with the flow (m³ / h), calculate real-time emissions: ,in For real-time monitoring of carbon emissions, is the CO2 concentration in the exhaust gas, is the gas flow rate, is the CO2 density, t is the operating time (h); Full life cycle carbon footprint calculation (LCA model), LCA boundary definition: covering the entire chain of "green hydrogen production → crude oil processing → aviation kerosene synthesis", calculated in stages: ,in For the full life cycle carbon footprint, Carbon emissions for green hydrogen production, ; For process energy consumption carbon emissions, (, is the electricity emission factor); Carbon emissions from transporting raw materials, ( are transport distance and emission factor per unit distance, respectively); For a multi-product co-production system (e.g. a refinery producing both jet fuel and diesel), carbon emissions are allocated based on the mass ratio: Serialize the carbon footprint data of each batch of aviation kerosene into JSON format and generate a unique hash value using the SHA-256 algorithm , ,Will Write to a private blockchain network (such as Hyperledger Fabric), with nodes including refineries, certification agencies, and airlines; transaction verification is completed through the PBFT consensus mechanism to ensure that the data cannot be tampered with; use WebGL to build a 3D carbon footprint heat map and display it on a timeline Change trends, compared with international standards; generate PDF certificates that comply with the ISO 14064 standard, including the following fields: Carbon footprint value and confidence interval (95% confidence level); blockchain transaction ID (which can be linked to a public browser to query the hash value); electronic signature of a third-party certification agency (based on RSA encryption).
[0030] The verification mechanism is used to verify the carbon emission factor of the green hydrogen production process to ensure the authenticity and timeliness of the data source. The specific process is as follows: Historical data storage and preprocessing: Stores historical green hydrogen production data, including daily renewable energy power generation G hist , electrolytic cell efficiency η hist , carbon emission factor etc., and store them in a time series database (such as InfluxDB) by timestamp classification; remove outliers (such as power generation that is negative or exceeds the equipment rating) and fill in missing data using linear interpolation; dynamic matching of multiple time windows: the real-time power generation of third parties is Match with historical data from the same period (such as data from the same period in the past 30 days) and calculate the similarity: ,in is the similarity between real-time and historical data, Real-time power generation for third parties, The mean and standard deviation of historical data for the same period; real-time data weight and historical data weight are assigned based on similarity : ; Multiple verification of carbon emission factors: Combine real-time and historical data to calculate verified carbon emission factors , ,in are the real-time grid emission factor and the historical grid emission factor respectively; if the deviation between the real-time data and the historical mean exceeds the threshold (such as ), triggering the following checks: Check whether the third-party data interface is abnormal; compare the equipment operation log to confirm whether the electrolyzer efficiency has changed suddenly; if there are no abnormalities, update the historical database and mark it as "verified new operating conditions"; Feedback and blockchain evidence: the verified carbon emission factor The weight distribution logic is written into the blockchain, and the hash value generation formula is expanded to: , add a new "Data Verification Level" field in the digital certificate, based on the weight Label the confidence level.
[0031] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.
Claims
1. The online detection and optimization system for aviation kerosene quality based on green hydrogen substitution technology is characterized by: include: The quality monitoring module uses multispectral and electrochemical sensors to monitor the purity of sulfides, aromatics, and hydrogen in real time. It combines fluid dynamics and chemical distribution analysis to determine the optimal installation location, ensuring data reliability. When the purity exceeds the standard, it triggers an alert and sends it to the optimization module. The optimization control module dynamically adjusts temperature, pressure, and green hydrogen flow based on a deep reinforcement learning algorithm. It also builds a multi-objective optimization model based on historical data to maximize sulfur removal efficiency while reducing energy consumption and carbon emissions, achieving closed-loop intelligent control. The carbon footprint tracking and verification module is used to integrate green hydrogen production, process energy consumption and transportation data, calculate the carbon emissions of the entire chain through life cycle assessment, use blockchain to store evidence to ensure that the data cannot be tampered with, generate visual reports and compliance certificates, and support carbon trading and certification.
2. The online detection and optimization system for aviation kerosene quality based on green hydrogen substitution technology according to claim 1 is characterized in that: The specific process of the quality monitoring module is as follows: Multispectral sensors are installed at optimal locations at the outlet of the hydrogenation reactor and downstream of the fractionator. They analyze the molecular characteristics of sulfides and aromatics using near-infrared spectroscopy and simultaneously collect hydrogen purity data detected by electrochemical sensors. Upload real-time data to the central processor, compare it with the preset aviation kerosene quality standards, and generate a quality deviation report; If the detection value exceeds the threshold, an early warning signal is triggered and pushed to the intelligent optimization module.
3. The online detection and optimization system for aviation kerosene quality based on green hydrogen substitution technology according to claim 2 is characterized in that: The process of determining the optimal position in the quality monitoring module is as follows: Establish a three-dimensional CFD model of the hydrogenation reactor and fractionation tower, simulate the fluid velocity field and pressure field distribution, and calculate the flow uniformity index : ,in is the local flow velocity, is the average flow velocity, The closer it is to 1, the more uniform the flow is, and n is the number of flow velocity sampling points; Candidate location screening: Select and areas without eddy currents or dead zones; In situ sampling was performed at candidate locations and sulfide concentrations were analyzed by gas chromatography and aromatics concentration , construct the concentration gradient function: ,in Axial position The sulfide concentration at is the initial sulfide concentration, Axial position The aromatic hydrocarbon concentration at Initial aromatics concentration, is the axial distance, is the attenuation coefficient, which is determined by fitting the experimental data; choose The position with the largest rate of change ensures that the sensor is sensitive to composition fluctuations; Measure the temperature T and pressure P at the candidate locations and select the area that meets the sensor's operating range, including T∈[250℃,400℃] and P∈[2MPa,12MPa]; Evaluate sensor performance degradation rate through durability testing : ,in They are the initial sensitivity of the sensor and the sensitivity of the sensor after 100 hours of operation. position to ensure long-term stability; Construct the objective function F to comprehensively evaluate the candidate locations: , where F is the comprehensive score of the sensor installation location, is the sulfide concentration gradient, is the aromatic hydrocarbon concentration gradient, is the weight coefficient, which is set according to the process requirements; select the coordinate point that maximizes F.
4. The online detection and optimization system for aviation kerosene quality based on green hydrogen substitution technology according to claim 1 is characterized in that: The specific operation steps of the optimization control module are as follows: Real-time data integration: Receive the sulfur content (S real ), aromatic content (A real ) and hydrogen purity (H pure ), combined with the current temperature collected by the sensor (T now ), pressure (P now ), hydrogen-to-oil ratio (R now =F H2 / Q oil , where F H2 is the green hydrogen flow rate, Q oil is the crude oil flow rate), forming a real-time state vector ; Extract parameter combinations under similar working conditions from the historical process database The corresponding sulfur removal efficiency , energy consumption , carbon emissions , build a nonlinear regression model: ,in is the fitting coefficient, is the current reaction temperature, is the reference temperature, is the current reaction pressure, is the reference pressure, is the current hydrogen-to-oil ratio; Deep reinforcement learning algorithm optimization: state-action space definition, defining state space , action space ,in ; Through multi-objective reward function , balancing sulfur removal efficiency, energy consumption and carbon emissions: ,in is the current energy consumption, is the baseline energy consumption, is the current carbon emissions, is the baseline carbon emissions, , is the dynamic weight; Adopting a double-delayed deep deterministic policy gradient algorithm, the strategy is optimized through the Actor-Critic framework: Actor network, policy function : Output action ; Critic network, value function : Evaluate action value ; The target function is to maximize the expected cumulative reward: , where D is the experience revisit pool, which stores the historical state-action-reward sequence; A dynamic learning rate adjustment strategy based on real-time performance feedback is introduced into the reinforcement learning algorithm. The learning rate is adaptively adjusted according to the change rate of sulfur removal efficiency and energy consumption fluctuations to improve the model convergence speed and stability. Real-time control instructions are issued and parameter adjustment calculations are performed: Actions output by the policy network , calculate the new process parameters: ; ; ; in It is a limiting function to ensure that the parameters are within a safe range; Will Send to the hydrogenation reactor and fractionator controller, and collect a new round of real-time data at the same time , update the experience replay pool D to achieve closed-loop optimization.
5. The online detection and optimization system for aviation kerosene quality based on green hydrogen substitution technology according to claim 4 is characterized in that: The specific steps of introducing a dynamic learning rate adjustment strategy based on real-time performance feedback into the reinforcement learning algorithm in the optimization control module are as follows: Real-time monitoring of sulfur removal efficiency change rate and energy consumption fluctuations ; Dynamically adjust learning rate : ,in is the adjusted learning rate, is the initial learning rate, is the change rate of sulfur removal efficiency, For energy consumption fluctuations, is the baseline energy consumption; The updated learning rate is applied to the Actor-Critic network parameter update, prioritizing accelerating convergence under efficient working conditions while reducing the learning rate to maintain stability during abnormal fluctuations.
6. The online detection and optimization system for aviation kerosene quality based on green hydrogen substitution technology according to claim 1 is characterized in that: The specific operating steps of the carbon footprint tracking and verification module are as follows: Green hydrogen production source classification: Collect carbon emission factors based on green hydrogen source type: Wind power water electrolysis: carbon emission factor ,in is the carbon emission factor of green hydrogen generated by wind power and solar power water electrolysis, The power consumption for water electrolysis is is the grid emission factor, Producing carbon emissions for electrolyzers, is the hydrogen production; Biomass Gasification: Carbon Emission Factors ,in is the carbon emission factor of green hydrogen from biomass gasification, They are carbon emissions from material planting, transportation and chemical processes; Collect real-time energy consumption data E of hydrogenation reactor and fractionation tower reactor 、E fraction , combined with the equipment energy efficiency coefficient η equip , calculate the actual energy consumption: ; The CO2 concentration in the process exhaust gas is measured by infrared sensor, combined with the flow , calculate real-time emissions: ,in For real-time monitoring of carbon emissions, is the CO2 concentration in the exhaust gas, is the gas flow rate, is the CO2 density, t is the operating time; The full life cycle carbon footprint calculation, also known as the LCA model, has an LCA boundary definition: covering the entire chain from green hydrogen production → crude oil processing → aviation kerosene synthesis, and is calculated in stages: ; in For the full life cycle carbon footprint, Carbon emissions for green hydrogen production, ; For process energy consumption carbon emissions, , is the electricity emission factor; Carbon emissions from transporting raw materials, , are the transport distance and the emission factor per unit distance respectively; For a multi-product cogeneration system, carbon emissions are allocated based on mass ratio: ; Serialize the carbon footprint data of each batch of aviation kerosene into JSON format and generate a unique hash value using the SHA-256 algorithm , ,Will Written into a private blockchain network, with nodes including refineries, certification agencies, and airlines; Transaction verification is completed through the PBFT consensus mechanism to ensure that data cannot be tampered with; Use WebGL to build a 3D carbon footprint heat map and display it on a timeline Change trends, compared with international standards; generate PDF certificates that comply with the ISO 14064 standard, including the following fields: Carbon footprint value and confidence intervals; Blockchain transaction ID; Electronic signature from a third-party certification body; Use the verification mechanism to verify the carbon emission factor of the green hydrogen production process.
7. The aviation kerosene quality online detection and optimization system based on green hydrogen substitution technology according to claim 6 is characterized in that: The specific steps for verifying the carbon emission factor of green hydrogen production using the verification mechanism in the carbon footprint tracking and verification module are as follows: Historical data storage and preprocessing: Stores historical green hydrogen production data, including daily renewable energy power generation G hist , electrolytic cell efficiency η hist , carbon emission factor , stored in the time series database according to timestamp classification; Outliers were removed and missing data were filled using linear interpolation; Multi-time window dynamic matching: real-time power generation of third parties Match with historical data of the same period and calculate similarity: ,in is the similarity between real-time and historical data, Real-time power generation for third parties, is the mean and standard deviation of historical data for the same period; Assign real-time data weight and historical data weight based on similarity : ; Multiple verification of carbon emission factors: Combine real-time and historical data to calculate verified carbon emission factors , ,in are the real-time grid emission factor and the historical grid emission factor respectively; if the deviation between the real-time data and the historical mean exceeds the threshold, that is , triggering the following checks: Check whether the third-party data interface is abnormal; Compare the equipment operation log to confirm whether there is a sudden change in the electrolyzer efficiency; If there are no abnormalities, the historical database is updated and marked as a verified new operating condition; Feedback and blockchain evidence: the verified carbon emission factor The weight distribution logic is written into the blockchain, and the hash value generation formula is expanded to: , add a new data verification level field in the digital certificate, based on the weight Label the confidence level.
Citation Information
Cited By
Tall oil production batch quality tracing management system based on big data
CN122311972A
