Big data operation management service system of methanol hydrogen production machine

Abstract

This invention discloses a big data operation and management service system for a methanol-to-hydrogen (MHT) generator. The MHT generator includes a methanol reforming reactor, a catalyst preparation and performance testing unit, a hydrogen purification unit, and a methanol combustion heating unit. Parameters from the sensing layer are transmitted to the LaaS (Layer as a Service) and PaaS (PaaS) layers via a transmission layer. The PaaS layer contains a methanol-to-hydrogen database. The SaaS layer performs terminal visualization monitoring, manual and intelligent dual-mode control, graded fault warning, and multi-site clustered management and maintenance based on the data output from the PaaS layer. The beneficial effects are: collecting and uploading data from the MHT generator, integrating data from the entire process (reforming reaction, purification, heating, etc.) using the PaaS layer, generating collaborative control commands, improving hydrogen purity and stability, enabling full-process monitoring in the methanol-to-hydrogen process, resulting in more stable hydrogen production, and significantly reducing manual debugging costs by assisting in optimization and adjustments.

Description

Technical Field

[0001] This invention relates to the field of methanol-to-hydrogen, and more specifically to a big data operation and management service system for a methanol-to-hydrogen machine. Background Technology

[0002] With the development of clean energy technologies, hydrogen energy, as a clean and efficient energy source, has received widespread attention globally. Among various hydrogen production technologies, methanol-to-hydrogen (MTO) machines primarily utilize the reforming reaction of methanol and water under specific temperature, pressure, and catalyst conditions to release hydrogen from the methanol, ultimately yielding hydrogen. Methanol has high energy density and is easy to store, enabling the provision of high-purity hydrogen to fuel cell vehicle refueling stations, chemical plants, and the medical industry.

[0003] Comparing this to Chinese patent CN213802914U, which discloses a hydrogen generator for methanol cracking to produce hydrogen, the device is characterized by comprising a heat exchanger, a vaporizer, a superheater, and a reactor. The heat exchanger has a first inlet connected to a methanol storage tank and a second inlet connected to the outlet of the reactor. The heat exchanger is used to preheat the methanol by exchanging heat with the high-temperature mixed gas generated by the reactor, while simultaneously cooling the high-temperature mixed gas. The vaporizer heats and vaporizes the methanol after passing through the heat exchanger to form vaporized methanol. The superheater heats the vaporized methanol to the temperature required for the reaction. The reactor promotes a catalytic cracking reaction of the vaporized methanol heated by the superheater to generate a mixed gas containing hydrogen.

[0004] In the structure of the hydrogen generator for methanol cracking hydrogen production described above, the heat exchanger, vaporizer, and reactor are independent modules connected only by material pipelines, with no data interaction. For example, when the reactor's catalytic efficiency decreases, it cannot provide feedback to the vaporizer to adjust the methanol vaporization rate. Manual adjustment based on parameters is required in real time. As a result, the equipment is not easy to maintain stable hydrogen production during use, and it is also impossible to link data according to actual conditions.

[0005] Therefore, a big data operation and management service system for methanol-to-hydrogen machines is needed to solve the above problems. This system can automatically adjust the process parameters of methanol-to-hydrogen based on market demand, stabilize hydrogen production, and perform full-process monitoring of the methanol-to-hydrogen process. Summary of the Invention

[0006] The purpose of this invention is to provide a big data operation and management service system for methanol-to-hydrogen machines, so as to solve the problems of lagging hydrogen production process adjustment and unstable hydrogen production in the existing technology.

[0007] The objective of this invention is achieved as follows: The present invention provides a big data operation and management service system for a methanol-to-hydrogen machine, including a methanol-to-hydrogen machine, wherein the methanol-to-hydrogen machine is equipped with a methanol reforming reactor, a catalyst preparation and performance testing unit, a hydrogen purification unit and a methanol combustion heating unit. The sensing layer includes a parameter acquisition component for the methanol-to-hydrogen machine; The parameters of the perception layer are transmitted through the transmission layer to the LaaS layer for cloud data storage and communication, and to the PaaS layer for data processing, algorithm calculation and instruction generation. The PaaS layer contains a methanol-to-hydrogen database. The algorithm analyzes the real-time data of the sensing layer and the methanol-to-hydrogen database to generate dynamic control commands, graded early warning signals, and maintenance plans for the methanol-to-hydrogen machine. The SaaS layer performs terminal visualization monitoring, manual and intelligent dual-mode control, hierarchical fault warning, and multi-site clustered management and maintenance based on the output data of the PaaS layer.

[0008] As a preferred embodiment, the transmission layer includes an industrial Ethernet module, a 5G communication module, and an edge computing gateway. The industrial Ethernet module is used to connect the various parameter acquisition components of the sensing layer. After preprocessing the acquired raw data, the edge computing gateway transmits it to the LaaS layer through the 5G communication module, whereby the LaaS layer provides a cloud storage resource pool and communication forwarding services.

[0009] As a preferred embodiment, the LaaS layer includes cloud redundant storage and communication forwarding services. The cloud redundant storage stores data through a disk array and sets up a local data backup unit at the edge node near the methanol-to-hydrogen machine. The communication forwarding service synchronizes the real-time received sensing layer data to the PaaS layer.

[0010] As a preferred embodiment, the data acquisition components include temperature sensors, pressure sensors, flow sensors, and concentration sensors installed on each unit of the methanol-to-hydrogen machine, as well as a catalyst performance detector installed on the catalyst preparation and performance testing unit and a component analyzer installed at the inlet and outlet of the hydrogen purification unit.

[0011] As a preferred embodiment, the acquisition components employ tiered sampling, with the temperature and pressure sensors acquiring data at a frequency of As / time, the flow and concentration sensors acquiring data at a frequency of Bs / time, and the catalyst performance analyzer and component analyzer acquiring data at a frequency of Cs / time, where B < A < C.

[0012] As a preferred embodiment, the algorithm operation includes at least one of the following: multiple linear regression regulation, threshold grading determination, process parameter linkage compensation, real-time load calculation algorithm, catalyst activity decay compensation algorithm, and reaction temperature gradient control algorithm.

[0013] As a preferred embodiment, the PaaS layer monitors the temperature data uploaded by the sensing layer in real time through data analysis, and uses a threshold grading judgment algorithm to determine whether two adjacent temperature fluctuations exceed ±5℃. When two consecutive temperature fluctuations exceed ±5℃, an instruction is generated, which sends a command to the sensing layer to increase the sampling frequency of the flow and concentration sensors to four times the original frequency, until two consecutive temperature fluctuations are within ±5℃.

[0014] As a preferred embodiment, the PaaS layer includes a methanol-to-hydrogen process optimization module. This module is communicatively connected to the algorithm and the methanol-to-hydrogen database. Through the algorithm, it fuses real-time data and historical data from the perception layer to generate AI dynamic control commands.

[0015] As a preferred embodiment, the manual and intelligent dual-mode control is normally AI intelligent control, and the AI ​​intelligent control calls the AI ​​dynamic control command. When switching to manual control, the AI ​​dynamic control command is paused and the control parameters input by the authorized operator are received. When switching to AI intelligent control or manual control times out, it will revert to AI intelligent control and re-invoke the AI ​​dynamic control command.

[0016] As a preferred option, methanol feed flow rate, reforming reaction temperature, catalyst activity parameters, hydrogen purity, and pressure data are defined as sensitive data. It also includes a security encryption layer, which encrypts the sensitive data using an encryption algorithm. The encrypted sensitive data must be decrypted correctly during transmission. At the same time, it performs identity authentication and permission monitoring on user access to the SaaS layer.

[0017] Positive and beneficial effects: Data from the methanol-to-hydrogen machine is collected through the sensing layer and uploaded through the transmission layer, solving the data silo problem of traditional equipment. The PaaS layer is used to integrate data from the entire process, including reforming reaction, purification, and heating, to generate collaborative control commands, thereby improving the purity and stability of hydrogen. The entire process of methanol-to-hydrogen technology can be monitored, resulting in more stable hydrogen production. This system can be used to assist in optimizing and adjusting process parameters during the operation of a methanol-to-hydrogen generator, thereby reducing methanol loss and significantly lowering manual debugging costs. Attached Figure Description

[0018] Figure 1 This is a system architecture diagram of the present invention; The diagram shows: Perception Layer 1; Transmission Layer 2; IaaS Layer 3; PaaS Layer 4; SaaS Layer 5. Detailed Implementation

[0019] The present invention will be further described below with reference to the accompanying drawings and embodiments. First embodiment:

[0020] See Figure 1 As shown, the present invention provides a big data operation and management service system for a methanol-to-hydrogen machine, including a methanol-to-hydrogen machine, which is equipped with a methanol reforming reactor, a catalyst preparation and performance testing unit, a hydrogen purification unit and a methanol combustion heating unit. Specifically, the methanol reforming reactor has a vertical cylindrical structure, with a raw material distribution zone, a catalyst bed zone, and a reaction product collection zone arranged sequentially from top to bottom. The catalyst bed zone is filled with a copper-based catalyst, and multiple temperature sensors are embedded inside the bed to monitor the reaction temperature distribution in real time. The top of the methanol reforming reactor has a methanol-water solution inlet and an inert gas purging port, which are connected to an external inert gas cylinder. The bottom has a reaction product outlet, and the side is connected to the catalyst preparation and performance testing unit. A component analyzer is installed at the reaction product outlet at the bottom of the methanol reforming reactor. The component analyzer is used to monitor the proportion of hydrogen, carbon dioxide, unreacted methanol, and other components in the reaction product in real time, so as to judge the efficiency and stability of the reforming reaction in a timely manner. In addition, it can also monitor the initial state of the mixed gas entering the hydrogen purification unit. The methanol reforming reactor allows methanol and water to undergo a reforming reaction under high temperature and with the help of a catalyst to produce a hydrogen-rich mixture. During its use, it is often necessary to precisely control the reaction temperature, feed ratio and catalyst bed distribution to improve reaction efficiency and hydrogen yield, while suppressing the occurrence of side reactions. The catalyst preparation and performance testing unit enables the entire process of preparing copper-based catalysts, from raw material mixing and high-temperature calcination to particle forming. Equipped with a catalyst data acquisition instrument, this unit can collect activity data during catalyst preparation and testing. For example, during the catalyst preparation stage, the instrument collects activity data under different calcination temperatures and forming pressures. During the performance testing stage, it continuously tracks the catalyst's activity decay over long-term reactions and its selectivity to the target product. This data helps determine the catalyst's quality and is also fed back to the methanol reforming reactor's control system to optimize overall reaction efficiency. The hydrogen purification unit is located at the bottom of the methanol reforming reactor, at the end of the reaction product outlet. It contains a molecular sieve adsorbent. When the mixed gas from the reforming reactor enters the tower group, the adsorbent will preferentially adsorb impurities such as carbon dioxide, water, and unreacted methanol, allowing only hydrogen to pass through, thereby obtaining hydrogen with higher purity. A component analyzer is also installed at the outlet of the hydrogen purification unit. The component analyzer is used to directly detect whether the purity of the purified hydrogen meets the standard. If the purity is insufficient, the system can quickly trigger the adjustment mechanism. The methanol combustion heating unit is located close to the methanol reforming reactor. It includes a burner, a heat exchanger, and a fuel delivery pipeline. A portion of the methanol feedstock is delivered to the burner through the pipeline. Here, the methanol is fully mixed with air and combusted to produce high-temperature flue gas. When the high-temperature flue gas flows through the heat exchanger, it transfers heat to the heating medium required by the reforming reactor. After its own temperature decreases, it is discharged, thus ensuring a stable supply of the high-temperature environment required for the reforming reaction.

[0021] Sensing layer 1 includes parameter acquisition components for establishing a methanol-to-hydrogen generator. Further, the acquisition components include temperature sensors, pressure sensors, flow sensors, and concentration sensors installed on each unit of the methanol-to-hydrogen generator, as well as a catalyst performance detector installed on the catalyst preparation and performance testing unit and a component analyzer installed at the inlet and outlet of the hydrogen purification unit. Inside the methanol reforming reactor, temperature sensors are embedded at different heights in the catalyst bed to monitor the temperature distribution in the most active reaction area in real time, preventing local overheating that could lead to catalyst sintering and deactivation. Pressure sensors are installed on the reaction product outlet pipe at the bottom of the reactor, close to the reactor body, to monitor pressure changes throughout the reaction system and prevent excessive pressure from causing safety risks or excessively low pressure from affecting reaction efficiency. Flow sensors are located downstream of the methanol-water solution inlet at the top of the reactor to measure the amount of raw material entering the reactor per unit time. Concentration sensors are placed on the connecting pipe at the bottom of the methanol reforming reactor, after the reaction product outlet and before the component analyzer, to detect the concentration of key components such as hydrogen and carbon monoxide in the mixed gas coming out of the reactor in real time, providing rapid feedback on the reaction effect. In the catalyst preparation and performance testing unit, temperature sensors are embedded in the heating chamber wall and the internal feed liquid of the catalyst synthesis reactor to monitor the temperature gradient during the preparation process and ensure stable growth of the catalyst crystal form; pressure sensors are installed on the inert gas inlet pipe at the top of the reactor to control the pressure environment inside the catalyst preparation reactor and prevent catalyst particle breakage due to high pressure; flow sensors are arranged at the outlet of the metering pump in the catalyst active component feed pipeline to clarify the catalyst feed data; and concentration sensors are installed at the outlet of the reaction evaluation device in the performance testing module to detect the conversion rate and product selectivity of the catalyst for methanol reforming reaction in real time and quickly evaluate its catalytic performance. Catalyst performance testing instruments are typically installed on the testing unit. One end is connected to the outlet of the catalyst reaction evaluation device via a sampling pipeline, and the other end is connected to the data processing system. Catalyst performance testing instruments can comprehensively analyze various key performance indicators of catalysts. For example, they can calculate catalytic activity and selectivity by detecting the concentration of each component in the reaction products, and evaluate the stability and anti-poisoning ability of catalysts by combining reaction temperature and pressure data. In the hydrogen purification unit, a temperature sensor is installed in the adsorbent bed of the pressure swing adsorption tower to monitor temperature changes during the adsorption / desorption process; pressure sensors are installed at the inlet, outlet, and equalization pipeline of the adsorption tower; flow sensors are arranged on the feed gas inlet main and product hydrogen outlet pipelines of the purification unit to measure the feed gas throughput and hydrogen production; a concentration sensor is installed after the buffer tank at the product hydrogen outlet to detect the purity of the purified hydrogen in real time; and a component analyzer is installed at both the inlet and outlet of the hydrogen purification unit to detect the purity of the hydrogen before and after purification. In the methanol combustion heating unit, temperature sensors are installed in the burner flame zone, on the outer wall of the heat exchange coil, and at the flue gas outlet to monitor combustion temperature, heat exchange efficiency, and exhaust gas temperature, preventing local overheating and equipment damage. Pressure sensors are installed at the methanol fuel feed pipe and the combustion air inlet to obtain fuel and air ratio pressure data. Flow sensors are arranged on the outlet pipe of the methanol fuel pump and the outlet duct of the combustion air fan to assist in obtaining and adjusting the fuel and air flow ratio. Concentration sensors are installed on the flue gas emission pipe to detect the concentration of carbon monoxide and nitrogen oxides in the flue gas, monitoring whether combustion is complete and environmental indicators. The data acquisition components are deployed in each unit of the entire process, including the methanol reforming reactor, catalyst preparation and performance testing unit, hydrogen purification unit, and methanol combustion heating unit. Each sensor and detection device is precisely deployed according to the process characteristics of each unit, so as to realize the real-time and accurate acquisition of various process parameters and media parameters during the operation of the hydrogen production system, and provide basic data support for system regulation. The collected parameters from Sensing Layer 1 are processed and transmitted via Transmission Layer 2. Transmission Layer 2 is equipped with HINET data gateway and HINET smart gateway. Through communication methods such as Ethernet, 3 / 4 / 5G mobile networks, and NB / WIFI networks, and based on transmission protocols such as UDP, HTTP, TCP, MOTIT, and WEBSERVICES, the data is stably transmitted to the cloud level. The data is first transmitted to LaaS Layer 3, where LaaS Layer 3, relying on the data center core bus and distributed data communication system, completes the secure cloud storage and distributed data communication of all the sensing data, providing core support for data storage and transmission for the entire IoT system.

[0022] The parameters of the methanol-to-hydrogen machine collected by the aforementioned acquisition components are locally accessed through the industrial Ethernet module of transmission layer 2, directly connecting to each parameter acquisition component to achieve data aggregation. The aggregated data is then transmitted to the edge computing gateway, where it performs preprocessing operations such as filtering and noise reduction, outlier removal, and data format conversion to reduce the pressure of invalid data transmission. The preprocessed data is then transmitted to the LaaS layer 3 through the 5G communication module. The LaaS layer 3 provides a cloud storage resource pool to persistently save the data and synchronizes the real-time data that needs further processing to the PaaS layer 4 through the built-in communication forwarding service. PaaS layer 4 contains a methanol-to-hydrogen database that stores real-time and historical data, equipment parameters, and process standard thresholds. Algorithms analyze the real-time data from perception layer 1 and the methanol-to-hydrogen database to generate dynamic control commands, tiered warning signals, and maintenance plans for the methanol-to-hydrogen generator. The built-in methanol-to-hydrogen database in PaaS layer 4 systematically stores real-time production data, historical operating records, equipment parameters, and process standard thresholds uploaded from perception layer 1. The algorithm unit periodically retrieves this data and uses a multiple linear regression control algorithm to analyze the deviation between real-time data and historical optimal operating conditions, generating precise AI dynamic control commands. Simultaneously, a threshold grading algorithm compares real-time data with preset safety thresholds, generating four levels of warning signals (blue, yellow, orange, and red) based on the degree of deviation. Furthermore, combining a catalyst activity decay compensation algorithm to analyze catalyst usage time and performance data, and a reaction temperature gradient control algorithm to assess equipment heat loss, a personalized maintenance plan is automatically generated, including spare parts replacement cycles, equipment maintenance time windows, and catalyst regeneration timing. These instructions, signals, and plans will be synchronized to SaaS layer 5 in real time and distributed to the corresponding execution components or operation and maintenance systems through the methanol-to-hydrogen communication library.

[0023] The algorithm operation includes at least one of the following: multiple linear regression control, threshold grading determination, process parameter linkage compensation, real-time load calculation algorithm, catalyst activity decay compensation algorithm, and reaction temperature gradient control algorithm. Different industries have vastly different requirements for hydrogen concentration. For example, the electronics industry, also known as the semiconductor industry, often requires hydrogen with a purity of over 99.99%, while the metallurgical industry requires a hydrogen concentration of 99% purity when reducing metals with hydrogen. Specific values ​​often need to be adjusted for different industries and even for different scenarios within the same industry. Therefore, the appropriate industry and hydrogen standard can be selected on the SaaS layer 5, and the PaaS layer 4 can configure various algorithm parameters according to requirements, adjusting parameters such as the judgment threshold and calculation weight of each algorithm. Specifically, the PaaS layer 4 receives data such as the temperature and feed flow rate of the methanol reforming reactor uploaded by the sensing layer 1 in real time and automatically triggers the threshold grading judgment algorithm. This algorithm presets a temperature fluctuation threshold of ±5℃ for the characteristics of the hydrogen production industry. When the temperature difference between two consecutive catalyst bed measurements exceeds this threshold, it immediately triggers a multivariate linear regression control algorithm. This algorithm calls the raw material ratio mathematical model specific to methanol-to-hydrogen production and combines it with a real-time load calculation algorithm to calculate the current optimal feed rate. Finally, through a process parameter linkage compensation algorithm, the adjustment command is sent to the execution layer to complete the dynamic correction of the flow parameters. Throughout the process, the catalyst activity decay compensation algorithm runs synchronously in the background, fine-tuning the temperature judgment benchmark based on the cumulative running time to adapt to the changes in process characteristics caused by catalyst aging.

[0024] The SaaS layer 5 uses the output data from the PaaS layer 4 to perform terminal visualization monitoring, manual and intelligent dual-mode control, hierarchical fault warning, and multi-site clustered management and maintenance.

[0025] LaaS Layer 3 includes cloud-based redundant storage and communication forwarding services. Cloud-based redundant storage uses RAID disk arrays to construct cloud-distributed storage sets, storing multiple copies of information such as methanol-to-hydrogen machine operating data, equipment logs, and maintenance records. This results in longer storage time, and even if a single disk fails, the data can still be accessed and recovered normally, enabling stable storage of industrial data. Furthermore, a local backup unit is deployed on the edge computing server near the methanol-to-hydrogen machine, which can use local hard drives or embedded storage modules for local data storage. The local data backup unit automatically performs incremental backups of the raw data uploaded from Sensing Layer 1 at fixed times every day. The communication forwarding service synchronizes the data received from Sensing Layer 1 in real time to PaaS Layer 4, supporting subsequent data processing and analysis. Industrial IoT protocols such as MQTT / CoAP / OPCUA are typically used for data forwarding.

[0026] The second embodiment differs from the first embodiment in that: Considering only the operating conditions within the methanol reforming reactor, this document describes the specific implementation details regarding the deployment of data acquisition components, sampling frequency, anomaly triggering mechanisms, and edge data backup. Temperature sensors in all data acquisition components are installed outside the reaction chamber and inlet / outlet pipelines of the methanol reforming reactor to collect the process temperature during the methanol reforming reaction, which is the temperature parameter of the methanol-to-hydrogen reaction. Pressure sensors are installed in the feed and product output pipelines of the methanol reforming reactor to collect pressure data within the pipelines during the reaction process, accurately capturing pressure changes during raw material transport and product discharge to provide reaction conditions. Flow sensors are installed in the methanol feed pipeline and hydrogen product output pipeline to collect raw material input flow and hydrogen output flow data, monitoring the dynamic changes in methanol feed supply and hydrogen production. Concentration sensors are installed at the product outlet of the methanol reforming reactor to collect parameters such as hydrogen purity and methanol residual concentration, enabling monitoring of reaction products. The component analyzer is installed on the inlet and outlet pipelines of the hydrogen purification unit. The component analyzer on the inlet pipeline of the hydrogen purification unit is also the outlet pipeline of the methanol reforming reactor. The component analyzer on the outlet pipeline of the methanol reforming reactor detects and analyzes the composition of the hydrogen product to help determine reaction efficiency and product purity. The collected data is directly related to the methanol-to-hydrogen process. In the tiered sampling of data, the temperature parameter is collected every 2 seconds, with temperature sensors acquiring the reaction temperature of the methanol reforming reactor in real time. The focus is on monitoring the temperature stability of the reaction system, providing a data basis for subsequent anomaly detection. Pressure sensors in the methanol reforming pipelines and reactor are collected every 2 seconds, synchronously with the temperature parameters, to monitor pressure fluctuations during the reaction process in real time, preventing pressure anomalies from affecting reaction efficiency. Simultaneous sampling of temperature and pressure also allows for comparison. Flow sensors on the methanol feed and hydrogen production pipelines collect flow data every 1 second, capturing real-time changes in feedstock and product output flow rates, providing accurate data for process control. Concentration data at the product outlet of the methanol reforming reactor is collected every 1 second. The component analyzer collects data every 90 seconds, focusing on monitoring core process indicators such as catalyst activity and hydrogen purity, balancing monitoring accuracy with rational resource utilization. In the temperature data of the methanol reforming reactor, when the temperature difference between two adjacent measurements is ≥5℃, meaning the temperature fluctuation of the methanol reforming reactor exceeds the reasonable range of the process, a sudden temperature change triggers a sampling frequency adjustment command. This does not involve abnormal triggering of other irrelevant parameters. It only targets core parameters related to flow rate and concentration. For example, when the temperature of the methanol reforming reactor is detected to fluctuate from 280℃ to 288℃, which is an 8℃ temperature difference between the two data points and exceeds the ±5℃ threshold, the threshold classification judgment in the algorithm calculation of PaaS layer 4 automatically triggers a command to be sent to sensing layer 1 to increase the sampling frequency of the flow sensor and concentration sensor to 4 times the original frequency, which is 0.25 seconds / time. This allows for real-time capture of subtle changes in raw material supply and product concentration, ensuring timely response to the process impact caused by temperature fluctuations. When the temperature fluctuation of the methanol reforming reactor returns to the ±5℃ range, the system automatically issues a command to restore the sampling frequency of the flow sensor and concentration sensor to the original 1 second / time, avoiding resource waste caused by ineffective high-frequency sampling.

[0027] The third embodiment differs from the first embodiment in that: The algorithm operation includes at least one of the following: multiple linear regression control, threshold grading determination, process parameter linkage compensation, real-time load calculation algorithm, catalyst activity decay compensation algorithm, and reaction temperature gradient control algorithm.

[0028] PaaS layer 4 contains a methanol-to-hydrogen process optimization module. This module is connected to the algorithm computation and methanol-to-hydrogen database to form a two-way communication. The methanol-to-hydrogen process optimization module uses algorithm computation to perform data fusion on the real-time data and historical stored data uploaded by the perception layer 1. Specifically, it can select data for cleaning, remove outliers, and align the cleaned real-time data with the historical data retrieved from the communication database according to the time series. The data fusion algorithm summarized by the algorithm computation unit adds dynamic weights to the data, and finally generates a fusion data set that takes into account both real-time operating condition fluctuations and historical best experience, i.e., AI dynamic control instructions. AI dynamic control commands are directly sent to the actuators in the methanol-to-hydrogen machine through PaaS layer 4, such as circulating pumps, electronically controlled valves, and agitators. By controlling the start-up, shutdown, opening degree, and speed of the actuators, the dynamic adjustment of the hydrogen production process is achieved.

[0029] In the SaaS layer 5 of this system, a dual-mode control working mode of manual and intelligent is adopted on the terminal device. Under normal circumstances, the manual and intelligent dual-mode control is AI intelligent control, which calls AI dynamic control commands. When switching to manual control, the AI ​​dynamic control command is paused and the control parameters input by the authorized operator are received, and the operation of the execution component is directly controlled. When switching to AI intelligent control or manual control times out, it will revert to AI intelligent control and re-invoke the AI ​​dynamic control command.

[0030] The fourth embodiment differs from the first embodiment in that: The methanol feed flow rate, reforming reaction temperature, catalyst activity parameters, hydrogen purity, and pressure data are defined as sensitive data. It also includes a security encryption layer, which encrypts sensitive data using encryption algorithms. After collecting production data, the sensors and edge nodes in the acquisition layer need to desensitize sensitive fields before packaging the data using encryption algorithms. When data is uploaded to the IaS layer 3, the underlying cloud servers, storage devices, and network links are protected by hardware-level encryption. The received encrypted data is overlaid with AES-256 storage encryption, and data synchronization between storage nodes is transmitted through a dedicated encrypted tunnel. The methanol-to-hydrogen communication library in the PaaS layer 4 encrypts the generated AI dynamic control commands with AES-256 before issuing them. At the same time, it transmits data and control commands with the edge nodes in the acquisition layer and the SaaS layer 5 through an SSL / TLS encrypted channel. The SaaS layer 5 uses multi-factor authentication for operators, and all cross-layer encryption keys are rotated periodically through a dedicated key management server to achieve end-to-end data security protection. The encrypted sensitive data needs to be decrypted through a corresponding decryption mechanism during transmission to be read correctly. At the same time, user access to the SaaS layer 5 is subject to identity authentication and permission monitoring.

[0031] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A big data operation and management service system for a methanol-to-hydrogen generator, comprising a methanol-to-hydrogen generator, wherein the methanol-to-hydrogen generator is equipped with a methanol reforming reactor, a catalyst preparation and performance testing unit, a hydrogen purification unit, and a methanol combustion heating unit, characterized in that: The sensing layer (1) includes a parameter acquisition component for establishing the methanol-to-hydrogen machine; The parameters of the perception layer (1) are transmitted to the LaaS layer (3) via the transmission layer (2) for cloud data storage and communication, and the PaaS layer (4) performs data processing, algorithm calculation and instruction generation. The PaaS layer (4) contains a methanol-to-hydrogen database. The algorithm analyzes the real-time data of the sensing layer (1) and the methanol-to-hydrogen database to generate dynamic control commands, graded early warning signals and maintenance plans for controlling the methanol-to-hydrogen machine. The SaaS layer (5) performs terminal visualization monitoring, manual and intelligent dual-mode control, hierarchical fault warning, and multi-site clustered management and maintenance based on the output data of the PaaS layer (4).

2. The big data operation management service system for a methanol-to-hydrogen machine according to claim 1, characterized in that, The transmission layer (2) includes an industrial Ethernet module, a 5G communication module and an edge computing gateway. The industrial Ethernet module is used to connect the parameter acquisition components of the sensing layer (1). The edge computing gateway preprocesses the acquired raw data and transmits it to the LaaS layer (3) through the 5G communication module. The LaaS layer (3) provides cloud storage resource pool and communication forwarding services.

3. The big data operation management service system for a methanol-to-hydrogen machine according to claim 2, characterized in that, The LaaS layer (3) includes cloud redundant storage and communication forwarding services. The cloud redundant storage stores data through a disk array and sets up a local data backup unit at the edge node near the location of the methanol-to-hydrogen machine. The communication forwarding service synchronizes the data received in real time from the perception layer (1) to the PaaS layer (4).

4. A big data operation management service system for a methanol-to-hydrogen machine according to any one of claims 1-3, characterized in that, The data acquisition components include temperature sensors, pressure sensors, flow sensors, and concentration sensors installed on each unit of the methanol-to-hydrogen machine, as well as a catalyst performance detector installed on the catalyst preparation and performance testing unit and a component analyzer installed at the inlet and outlet of the hydrogen purification unit.

5. The big data operation management service system for a methanol-to-hydrogen machine according to claim 4, characterized in that, The acquisition components employ tiered sampling. The acquisition frequency of the temperature sensor and pressure sensor is As / time, the acquisition frequency of the flow rate and concentration sensor is Bs / time, and the acquisition frequency of the catalyst performance tester and component analyzer is Cs / time, where B < A < C.

6. The big data operation management service system for a methanol-to-hydrogen machine according to claim 5, characterized in that, The algorithm operation includes at least one of the following: multiple linear regression regulation, threshold grading determination, process parameter linkage compensation, real-time load calculation algorithm, catalyst activity decay compensation algorithm, and reaction temperature gradient control algorithm.

7. The big data operation management service system for a methanol-to-hydrogen machine according to claim 6, characterized in that, The PaaS layer (4) monitors the temperature data uploaded by the sensing layer (1) in real time through data analysis, and uses a threshold grading judgment algorithm to determine whether the temperature fluctuations of two adjacent times exceed ±5℃. When two adjacent temperature fluctuations exceed ±5℃, an instruction is triggered to generate and send an instruction to the sensing layer (1) to increase the sampling frequency of the flow and concentration sensors to 4 times the original frequency until two adjacent temperature fluctuations are within ±5℃.

8. The big data operation management service system for a methanol-to-hydrogen machine according to claim 4, characterized in that, The PaaS layer (4) is equipped with a methanol-to-hydrogen process optimization module. This module is connected to the algorithm operation and the methanol-to-hydrogen database. The algorithm operation is used to fuse real-time data and historical data of the perception layer (1) to generate AI dynamic control instructions.

9. The big data operation management service system for a methanol-to-hydrogen machine according to claim 8, characterized in that, The manual and intelligent dual-mode control is normally AI intelligent control, which calls the AI ​​dynamic control command. When switching to manual control, the AI ​​dynamic control command is paused and the control parameters input by the authorized operator are received. When switching to AI intelligent control or manual control times out, it will revert to AI intelligent control and re-invoke the AI ​​dynamic control command.

10. A big data operation management service system for a methanol-to-hydrogen machine according to any one of claims 4-8, characterized in that, The methanol feed flow rate, reforming reaction temperature, catalyst activity parameters, hydrogen purity, and pressure data are defined as sensitive data. It also includes a security encryption layer, which encrypts the sensitive data using an encryption algorithm. The encrypted sensitive data must be decrypted correctly during transmission. At the same time, it performs identity authentication and permission monitoring on the user's access to the SaaS layer (5).

Images