Intelligent prediction management system for service life of methanol synthesis catalyst
By developing an intelligent prediction management system for the service life of methanol synthesis catalysts in chemical enterprises, combining big data and artificial intelligence, problems such as unstable performance and difficulty in preventing deterioration in catalyst management have been solved, intelligent management of catalysts has been realized, and service life has been extended. And reduce operating costs.
Patent Information
- Application Number
- CN202411165096.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-11-12
AI Technical Summary
Chemical companies lack real-time online monitoring and perception methods in catalyst management, resulting in unstable catalyst performance, difficult to prevent deterioration, difficult to share experience, no predictions in the future, and difficult to pre-manage and control maintenance, resulting in low production efficiency and high cost.
Develop an intelligent prediction management system for the service life of methanol synthesis catalysts, combining big data, artificial intelligence and Internet of Things technologies to realize catalyst performance perception, early warning, health management and life prediction, and provide intelligent management throughout the life cycle.
Through real-time monitoring and data analysis, the catalyst operation status can be visualized and optimized, extend the catalyst service life, reduce operating costs, and improve production efficiency and product quality.
Smart Images

Figure CN118942574A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of auxiliary operation in the chemical industry, and in particular to an intelligent prediction and management system for the service life of a methanol synthesis catalyst. Background Art
[0002] Catalysts are an important part of the chemical production process, with many application scenarios in chemical companies and a great impact on production operations. However, due to the lack of corresponding data accumulation and technical support in the past, the operation and maintenance management of catalysts has made it impossible for companies to effectively monitor and perceive the health and performance status of catalysts, and it is even more difficult to establish a virtuous cycle between management, production decision-making and operation. As a result, there is a lack of effective means to ensure the stability of catalyst working conditions, healthy operation and performance optimization, making it difficult to break through bottlenecks such as increasing product output, reducing operating energy consumption and stabilizing product quality.
[0003] The existing technology of methanol synthesis catalyst management mainly has the following problems:
[0004] 1. The performance of the catalyst is unstable, fluctuates greatly, and there is no perception of the detailed status.
[0005] The performance evaluation of catalysts lacks real-time online monitoring and perception means, and there is no means to intuitively present the operating status and performance changes. The operating status of the catalyst is to evaluate the catalyst status and performance through the overall macroscopic performance of the catalyst in a cycle. The evaluation only evaluates the average performance of the catalyst over a period of time. It cannot control the unstable performance perception and fluctuation in real time, and cannot meet the needs of real-time performance evaluation and perception of the catalyst during use.
[0006] 2. Deterioration and harm are difficult to prevent, and healthy operation is not guaranteed.
[0007] The on-site management of accidents and incidents of catalysts is post-event management. For example, in the case of temperature spikes in the catalyst, only when temperature spikes occur will the disposal operation be carried out according to the established plan. There are no early warning and prevention management methods. Catalyst hazards also include catalyst breakage, overheating, coking, poisoning, temporary deactivation, etc. There is no analysis, management and evaluation of catalyst performance degradation, and there is a lack of effective means to implement a "physical examination" of the catalyst, analyze, diagnose, evaluate and manage the health status of the catalyst to ensure the stable operation of the catalyst and healthy operation, thereby extending the catalyst life cycle.
[0008] 3. It is difficult to share experience and optimize performance.
[0009] The activity optimization of the catalyst is only based on the experience of the operator or engineer. It is not only difficult to adjust the operation quantitatively, but also difficult to summarize the experience and share the optimization plan. The activity of the catalyst and whether the operation is stable will vary depending on the operator. Even if the same person is operating, it is difficult to make fine adjustments and optimizations to ensure the continuous high-performance operation of the catalyst due to the lack of decision-making and optimization basis. It is difficult to optimize the activity and performance status of the catalyst under the existing conditions and break through the operational bottleneck of reducing costs and increasing efficiency.
[0010] 4. There is no prediction for the future and no basis for maintenance.
[0011] The life of catalysts is almost always based on the manufacturer's recommendations. For example, if the service life of a catalyst is 3 years, it should be replaced upon expiration. Without predicting the performance indicators at the end of the catalyst's service life and the remaining duration of use, it is difficult to make overall arrangements for catalyst operation, maintenance, and replacement in advance. If the catalyst deteriorates severely prematurely and is replaced only when it expires, production will be seriously affected, resulting in production losses. Otherwise, if the catalyst performance is good, it will result in premature replacement of the catalyst, wasting expensive catalyst costs.
[0012] 5. Post-management, operation and maintenance are difficult to manage and control in advance.
[0013] During the operation and management of the catalyst, damages such as overheating, sintering, breakage, carbon deposition, poisoning, and deactivation may occur, which will cause the performance of the catalyst to deteriorate sharply or even become scrapped. However, the current management methods can only sense, judge, and deal with the damage after it occurs. It belongs to post-event management. There is no way to achieve the benefits of predicting degradation, intervening in advance, and preventing and controlling management in advance. In addition, the mismatch between the physical properties of raw materials and the process control parameters and performance status during the catalytic reaction will lead to the inability to fully exert the performance of the catalyst, and indicators such as conversion rate, selectivity, and reaction activity intensity cannot be optimized, which will lead to increased raw material consumption and decreased product yield. In the existing management process, performance accounting is only performed by work group or natural day, and problems cannot be dealt with in time to avoid or reduce losses.
[0014] In view of this, we conducted in-depth research on the above issues, which led to the present case. Summary of the invention
[0015] The purpose of the present invention is to provide an intelligent prediction and management system for the service life of a methanol synthesis catalyst, so as to solve the problems existing in the management of the methanol synthesis catalyst proposed in the above background technology.
[0016] To achieve the above object, the present invention provides the following technical solutions: an intelligent prediction and management system for the service life of a methanol synthesis catalyst;
[0017] The catalyst intelligent management system is divided into front-end display and back-end management;
[0018] The front-end display mainly includes:
[0019] The core homepage collects the main information of each module; including health index monitoring, early warning status, benefit status, performance monitoring, conversion rate analysis, performance optimization, life prediction, quality and yield prediction and other functions, analyzing and displaying the operation status and overall health status of the catalyst from a global perspective;
[0020] Data monitoring and early warning, mainly including physical properties, processes, performance indicators monitoring, synthesis tower temperature field digital twin and early warning statistical analysis system; the data monitoring and early warning is to monitor the operating status of key indicators such as materials, processes, performance of the methanol synthesis tower, and analyze and display them according to different dimensions; alarm indicators are alarmed, alarm information is displayed, and alarm distribution is analyzed; at the same time, the completion of each team of key indicators is ranked and evaluated; the system includes key indicator card module, key indicator trend analysis module, key indicator team ranking module, hot spot temperature monitoring and analysis module, key performance indicator monitoring and analysis module, indicator alarm monitoring module, and alarm distribution statistics module;
[0021] Status analysis and evaluation, including core systems such as correlation analysis, health management, performance evaluation, and operating condition optimization, and prediction and decision-making assistance modules including life prediction, three-year performance prediction, and comprehensive performance evaluation; the status analysis and evaluation also includes catalyst life prediction and decision-making, which is used to comprehensively predict the operation of the synthesis tower from multiple angles, predict the catalyst operation time, total processing capacity, and total output, and display actual operation data, prediction data, and estimated remaining capacity; at the same time, it displays the load situation and the health of the catalyst; according to the three-year life cycle of the catalyst, it predicts the core indicators such as conversion rate, activity, yield, and reaction pressure difference when the three-year period is reached; and provides assistance for users to optimize process indicators and make catalyst replacement decisions;
[0022] The back-end management mainly includes system management, security management and audit management.
[0023] Preferably, the key indicators of the data monitoring and early warning include real-time display of key indicators such as total methanol production of the reaction system, total synthesis gas unit consumption, load rate, conversion rate, and team ranking;
[0024] Basic indicators include the temperature trend of the methanol synthesis tower, temperature change rate, product yield, conversion rate and other basic indicators are displayed in real time;
[0025] The alarm information includes the alarm trend of the methanol synthesis tower, which is displayed according to the dimensions of materials, processes, and performance, and the proportion is statistically analyzed;
[0026] The main data source of the system is DCS, which mainly includes design parameters, process parameters, physical parameters, performance parameters, and state parameters; it includes not only real-time parameters of temperature, pressure, flow, composition, and liquid level, but also derivative, analysis, and judgment parameters, including:
[0027] Design parameters: filling quantity, form, physical properties and process design parameters;
[0028] Process parameters: reactor inlet temperature, pressure, temperature at each point of the bed, temperature rise, reaction pressure, pressure drop, liquid level, outlet temperature, outlet pressure, reaction heat, drum pressure, temperature, steam temperature, water supply temperature; raw gas (circulating gas, mixed gas) pressure and temperature;
[0029] Physical parameters: raw gas volume, composition, components; circulating gas volume, temperature, pressure, components, components; tower inlet gas volume, temperature, components; tower outlet temperature, pressure, composition, components; product output, by-product output;
[0030] Performance parameters: various raw material conversion rates, selectivity, yield, activity, health index, and space velocity performance indicators derived from the original point through the mechanism model;
[0031] Status parameters: instrument running status, stop status, alarm status;
[0032] Set parameters: warning threshold, warning upper limit, alarm lower limit, encoding method, and range.
[0033] Preferably, the key indicator card module mainly displays the real-time operation data of multiple synthesis tower public indicators, analyzes the growth and decline compared with the average of the previous working day and the average of yesterday, and pops up a window to view the historical trend chart of the indicator;
[0034] The key indicator trend analysis module performs trend analysis on multiple key public indicators of the synthesis tower according to different time dimensions, including indicator values and mean values;
[0035] The key indicator team ranking module counts the indicator completion ranking of each team in the corresponding time dimension;
[0036] The hot spot temperature monitoring and analysis module monitors the real-time temperature and change rate of the hot spot bed of each synthesis tower in real time, and performs trend analysis and comparison of the hot spot temperature according to different time dimensions;
[0037] The key performance indicator monitoring and analysis module monitors the operation and change rate of each indicator in different time dimensions in the form of indicator cards or circular progress bars for each key performance indicator of each synthesis tower;
[0038] The indicator alarm monitoring module scrolls and displays all the synthesizer indicator alarm information, summarizes and displays the alarm times by indicator type, checks the historical alarm trends of various indicators, and statistically analyzes the ranking of the alarm times of each indicator;
[0039] The alarm distribution statistics module ranks the indicators by the number of alarms, causing users to focus on the indicators of alarm frequency, and ranks them by indicator type.
[0040] Preferably, the data monitoring and early warning also includes:
[0041] Material data monitoring: monitor the operating status of key material indicators of the methanol synthesis tower, analyze and display them in different dimensions; issue alarms for alarm indicators, display alarm information, and count alarm processing status; the system includes material component monitoring and analysis module, key material monitoring module, indicator alarm statistics module, and alarm information module;
[0042] Process data monitoring: monitor the operating status of key process indicators of the methanol synthesis tower, analyze and display them in different dimensions; issue alarms for alarm indicators, display alarm information, and count alarm processing status; the system includes key process indicator monitoring and analysis modules, related process indicator monitoring modules, indicator alarm statistics modules, and alarm information modules;
[0043] Performance data monitoring: monitor the operating status of key performance indicators of the methanol synthesis tower, analyze and display them in different dimensions; issue alarms for alarm indicators, display alarm information, and count alarm processing status; the system includes key performance indicator monitoring module, performance status monitoring module, alarm ranking module, alarm statistics module, and alarm information module;
[0044] Monitoring of key indicators of the synthesis tower: monitor the operating status of key indicators of the methanol synthesis tower and conduct multi-dimensional analysis of the temperature of each bed layer of the synthesis tower; including key indicator monitoring module and layer temperature monitoring and analysis module.
[0045] Preferably, the status analysis and evaluation specifically includes:
[0046] Indicator correlation analysis: Monitor the operating data of each synthesis tower and main system in the form of DCS screen; select a key indicator, conduct correlation analysis on the indicator, calculate the correlation coefficient between the target indicator and other indicators in the screen, and find out the strongly correlated indicators;
[0047] Health assessment and fault diagnosis: Evaluate the health of key indicators of each synthesis tower, and evaluate the matching and deviation between the actual operating status of the indicators and the health curve; monitor the operating status of indicators in real time, calculate the health value in real time, and compare the real-time parameters with the health threshold; issue early warnings on the operating status of indicators in real time, display early warning information, and count the number of various early warnings;
[0048] Health indicator correlation information query: used to query the correlation degree between a certain indicator and other indicators within a certain working condition range, and provide users with the optimal values of each related indicator under the selected conditions to guide user operations;
[0049] Performance status evaluation and impact: Based on the indicator health curve, the actual operation status of the key indicators of the synthesis tower is evaluated. The deviation of the actual operation data of the indicators from the health curve is evaluated from multiple time dimensions. The total operation time, efficient operation time, inefficient operation time, parking time and the corresponding proportion of the synthesis tower in different time dimensions are counted. The operation status of the main related indicators is analyzed, so as to evaluate the performance status of a specific indicator.
[0050] Real-time optimization of performance indicators: real-time display of current working condition information, real-time display of current value and recommended value of given target indicators, real-time display of current value and recommended value of related parameters, and guidance of user operation;
[0051] Query of optimal index interval: mainly used for users to query the operating interval of the specified index under each time dimension, obtain relevant influencing factors through correlation analysis, calculate the optimal operating interval of the target index and the healthy interval and reference value of the relevant factors by specifying the operating interval of the influencing factors, form a production plan under the specified working conditions and constraints, and guide user operations;
[0052] Catalyst life prediction: Comprehensively predict the operation of the synthesis tower from multiple angles to predict the catalyst operation time, total processing capacity, and total output, and display actual operation data, forecast data, and estimated remaining capacity; also display the load situation and the health of the catalyst; predict the core indicators such as conversion rate, activity, yield, and reaction pressure difference when the three-year life cycle of the catalyst is reached; provide assistance for users to optimize process indicators and make catalyst replacement decisions.
[0053] Preferably, the catalyst life prediction and decision-making specifically include:
[0054] Quality and output forecast: predict the crude methanol output and water content in the next hour, as well as the impact time and overall change trend, and provide correlation indicators that affect the change of indicators;
[0055] Life and performance prediction: including life prediction based on core indicators such as conversion rate, defining the worst degradation boundary value, and predicting the service life, remaining processing volume, and product volume through prediction models based on historical data and real-time data; in addition, according to the actual management standard requirements, develop a prediction of the final performance indicators based on the end of the three-year use period of the catalyst to provide operational decisions for catalyst maintenance, replacement, and adjustment;
[0056] Performance management: Based on the performance indicators formed by each module, performance management characteristic values are formed, and performance ranking and performance analysis are performed to combine with offline management to improve the overall management closed loop.
[0057] Preferably, the system management of the back-end management includes three systems: organization management, user management, and indicator management. Organization management mainly includes functions such as adding, modifying, and deleting organizations; user management mainly includes functions such as adding users, associating users with organizations, modifying user information, freezing and unfreezing users; indicator management mainly includes health indicator management and optimization indicator management, which can define health thresholds and optimization target thresholds online, and provide definition and modification functions for process indicator management;
[0058] The security management includes three menus: menu management, role management, and user log management. Menu management is mainly used to display all function menus of the system; role management mainly includes functions such as role addition, role modification, role deletion, user role association, and role authorization; user logs are mainly used to audit user operation behaviors;
[0059] The audit management mainly audits the operational behaviors of system administrators and security administrators.
[0060] Preferably, the data service flow shows that after the data of each module is collected, it enters the data platform for aggregation and management, and the required performance and other data are calculated in real time through the mechanism formula and stored together in the InfluxDB real-time database;
[0061] The stored data is processed into derived data using the mechanism formula for monitoring and sensing module output. Through correlation analysis, the relationship between input, output, materials, process and performance data is established. The correlation relationship is embodied in data, such as positive and negative correlation, correlation coefficient, etc., and the correlation is displayed in combination with the actual process.
[0062] At the same time, the data correlation relationship, field experience and reaction mechanism are used to divide the actual working conditions and data sets under various working conditions; the input data is trained through the advanced prediction model to simulate the health values under various subdivided working conditions, and the health curve showing the current working conditions is predicted in the time dimension. According to the on-site engineers and process card management standards, the upper and lower thresholds based on the health curve are defined; for the values of the health indicator fluctuations of concern, the main cause analysis is carried out, and the top 6 most manageable factors are presented in real time to form data samples, which enter the knowledge and experience module and solidify into experience knowledge. At the same time, the parameters are used as the boundary data of the optimization module; for abnormal indicator fluctuations, which cause the catalyst performance to decline and cause catalyst degradation failure data, early warning prompts are issued, and the changes in the main related factors are analyzed to find out the changes, and the formed data samples enter the health management to optimize the health management model and become the data boundary for working condition optimization; according to the early warning indicators, the early warning statistics module is included, and the data is used as a performance indicator and displayed in the monitoring module at the same time;
[0063] For the normal operating conditions and index data, the performance optimization algorithm model is used for processing, and the data control plan with stable optimization is output for reference and optimization by operators; the optimized results are evaluated for performance, and the optimized samples and experience are solidified as knowledge data; at the same time, auxiliary optimization is performed on the data results of degradation analysis, and the optimized performance evaluation is also performed; the data after performance evaluation is used as input data for the performance management module;
[0064] The input and output data of the above modules are used as input data of the prediction module. The multi-algorithm prediction model is used to output relatively accurate prediction data to make long-term predictions on service life and performance; at the same time, short-term predictions are made on output and product quality to assist management and operators in making decisions and judgments; the core result data of the above modules are displayed as the content of the comprehensive homepage, so that decision-makers and managers can quickly understand the operation status of methanol synthesis conditions, catalyst performance, performance optimization and the performance of the operating team;
[0065] The performance indicators in management will be included in the performance ranking and comprehensive performance evaluation system, which will conduct comprehensive analysis and evaluation on the management effectiveness of control parameters, the optimization performance of performance indicators, and the data implemented to economic benefits, provide support for the closed-loop data cycle of system data recycling and re-optimization, and ensure a virtuous cycle of seeking excellence in stability and selecting the best from the best.
[0066] Preferably, the underlying technical architecture of the system is based on the Hadoop architecture, and uses MapReduce and Spark computing frameworks to achieve batch storage and calculation of data; on the IOT side, the data on the network gate side is taken over based on the MQTT protocol; the Kafka message subscription and publishing system is used to manage messages; the flink batch stream combined processing framework is used to ensure the system's millisecond / second processing capabilities for massive data; on the platform service side, the overall microservice architecture system is adopted, the algorithm model design is implemented through the Cloudiip-lab component, the digital twin model design is implemented through the HT component, and the registration and verification of the algorithm model are uniformly managed through the Zookeeper distributed collaboration framework.
[0067] Preferably, the system covers three types of industrial models: data algorithm model, industrial mechanism model, and business process model, specifically including:
[0068] a. Data algorithm model
[0069] Advanced warning model: Based on the health warning management of the advanced warning model application, the data of the normal working condition of the catalyst is selected to build the advanced warning model to realize the prediction of the health value; the difference between the actual value and the predicted value is evaluated based on the approximation evaluation algorithm to realize the early warning of abnormal indicators; to achieve the purpose of early intervention in process operation and ensure the long-term healthy operation of the catalyst;
[0070] Optimal parameter sample parameter model: The optimal process sample library is constructed based on the partitioning algorithm. The partitioning algorithm is used to cluster the on-site working conditions to achieve the division of working conditions. The working condition identification model is established based on the working condition division results to calculate the optimal process parameters for each working condition. The production data is collected in real time to identify the working condition, and the optimal process parameters under the working condition are found and recommended to the operator for production guidance.
[0071] Catalyst service life and performance prediction model: Catalyst service life prediction based on machine learning, building a regression model and establishing a machine learning mechanism based on indicators such as conversion rate, output, activity, moisture content, and usage time, using a supervised algorithm to predict the remaining life of the catalyst; effectively guiding production scheduling, catalyst procurement, and catalyst replacement business;
[0072] Synthesis tower heat distribution model: Based on the digital twin, the synthesis tower heat distribution model refers to the actual equipment structure, measurement point distribution, and synthesis tower temperature data of the synthesis tower to achieve intuitive real-time monitoring of the temperature field distribution of each bed layer of the synthesis tower, providing a strong decision-making basis for reactor operation status evaluation, abnormal status diagnosis, and reaction trend prediction;
[0073] b. Industrial Mechanism Model
[0074] Series and parallel process flow accounting model: The mechanism-based series and parallel process flow accounting model uses part of the flow, components, and physical property balance data of the feed gas, circulating gas, and mixed gas of towers 1 and 2, combined with the actual series and parallel process flow of towers 1 and 2, and uses the ideal gas state equation, Dalton's law of partial pressure and other mechanism methods to convert mass and volume flow. In addition, according to the mechanism principle of element balance, the flow data of the reactors without series and parallel connection is calculated, and the mixed gas data is intelligently predicted; the technical transformation of adding flow meters and component measurement points is avoided;
[0075] Catalyst performance characteristics visualization model: Using the mechanism formula, the performance index value is calculated in real time, combined with the circulation ratio, hydrogen-carbon ratio and other working condition division indicators, the operating performance change characteristics and relationships are established and displayed in multiple dimensions, and the performance spatial change model is established for visualization and perception;
[0076] c. Management process model
[0077] Through data collection from various modules of the system, health thresholds and optimization range thresholds can be customized in combination with business needs. Combined with the management of these thresholds, performance evaluation indicators for practice and operation processes are formed. Based on the performance evaluation indicators, performance rankings are conducted on each management object, and comprehensive performance evaluation and management are conducted at different time granularities to form a closed-loop management model for perception, alarm, intervention, health protection, working condition optimization, performance evaluation, and prediction.
[0078] Compared with the prior art, the invention has the following beneficial effects: the intelligent prediction and management system for the service life of the methanol synthesis catalyst is
[0079] Based on the management business of the methanol synthesis catalytic reaction section, the catalyst operation management business is closely integrated with advanced technologies such as big data, artificial intelligence, and the Internet of Things to create an intelligent catalyst management system. This system is the first of its kind in the industry and a successful practice to promote the digital transformation of enterprises and technological innovation. It is an innovative achievement that uses the deep integration of IT, DT, and OT technologies. It is mainly used for the full life cycle management of the catalyst use process, and mainly serves the core catalytic reaction sections of various forms such as fixed bed, slurry bed, fluidized bed, etc. on the main process route in the petrochemical, coal chemical, fine chemical, and basic chemical industries.
[0080] The system is a catalyst intelligent management system developed based on the full-cycle operation data of the reaction catalyst in the methanol synthesis section, combined with on-site management standards and business needs. It provides core functions such as performance perception, early warning, main cause diagnosis, health management, auxiliary optimization, and life performance prediction, aiming to assist and upgrade catalyst operation and maintenance management, achieve stable reaction conditions, reduce catalyst damage, optimize catalyst performance, extend catalyst service life, improve production efficiency and reduce production costs.
[0081] The system gathers, manages, counts and analyzes related data on physical properties, processes, performance and environment, and uses core algorithms to establish data models, forming core functions such as operation cockpit, data monitoring, working condition warning, situation awareness, correlation analysis, health management, performance evaluation, working condition optimization, product output and life prediction. Through the catalyst intelligent management system, auxiliary operation, stable operation, hazard reduction, health protection, life extension, activity optimization and efficiency improvement can be achieved during the operation of the catalyst, thereby improving decision-making level, reducing operating costs and increasing profitability.
[0082] The catalyst intelligent management system improves the stability of catalyst operation through the overall solution of real-time monitoring → early warning → fault diagnosis → health assessment → auxiliary optimization, and uses big data value mining to improve corporate decision-making level, reduce production and operation costs, and improve corporate profitability.
[0083] Real-time visualization of catalyst performance enables visible, clear, and predictable perception of catalyst operating status and the use of research-level auxiliary tools in practice, which can quickly and intuitively determine performance status, operating condition matching, and optimization direction determination.
[0084] Catalyst online health management defines health standard lines through the system, and based on the standard lines, performs online physical examination, diagnosis and management of the catalyst operation monitoring status to ensure the healthy operation of the catalyst.
[0085] Seek the best in stable working conditions, maintain high-activity operation, take performance data as the goal, and health data as the basis to select high-performance corresponding status and operation data. After verification of design and process card data, combined with empirical rules, push optimization control and operation plans. The plan combines mechanism, experience, history and current situation factors, and has the characteristics of easy operation, easy achievement, stable regulation, and reliable optimization.
[0086] Catalyst status prediction, pre-management, and pre-control. The system issues an early warning before parameter degradation, intervenes in advance, and assists in correcting operating parameters to avoid damage and degradation. At the same time, based on changes in input-side data, predict performance, output, and quality trend changes in advance, push optimization plans, and maintain high-performance operation. Transform post-analysis and management into pre-warning, intervention, and management. Improve the predictability and timeliness of problem perception, and assist in pre-management of catalyst operation.
[0087] By using this system, the effect of "state perception provides a basis for performance analysis, state analysis provides direction for health management, health management provides scope for operation optimization, and operation optimization improves the overall operation performance of the catalyst" can be achieved, thereby realizing intelligent management of the entire life cycle of the catalyst.
[0088] Through system construction, benefits have been improved in management, economy and society, which are specifically reflected in:
[0089] 1. Management Benefits
[0090] State perception: Achieve detailed and situational awareness through monitoring perception. Add and improve 38% of real-time online high-value derivative data.
[0091] Early warning: Real-time perception, evaluation, and early warning of catalyst status through physical properties and process data. The warning time is shortened by more than 50% compared to the DCS alarm time.
[0092] Health management: Real-time "physical examination" and online health management. The stability rate increased by 2.2%-4.1% and the damage was reduced by 80%.
[0093] Performance management: Integrates process management with catalyst system management, and realizes closed-loop management modes such as catalyst management index definition, real-time control, effect evaluation, real-time optimization, and performance evaluation. Real-time control of on-site catalyst operation, abnormal conditions, and optimization and performance achievement, full digital control, efficient, transparent, and automated management.
[0094] Improve quality and increase efficiency: Through system-assisted management, the output and quality of the product are taken as optimization goals, and an operation plan is formed, including the control parameters of the raw material indicators and the control parameters of the reaction process, so as to give full play to the activity of the catalyst and achieve the goal of improving the conversion rate or selectivity, ultimately reducing the production of water and by-products in the product, increasing the yield of methanol, and reducing the processing difficulty of the stable section.
[0095] Operational improvement: Through the online process card indicators and operational target optimization indicator threshold definition and dynamic indicator matching control, more lean indicator management and process control are achieved, making stability easier to achieve, optimization easier to achieve, reducing the analysis process and difficulty, directly reaching the control plan, improving operational management efficiency, and making operational management more accurate and efficient.
[0096] (II) Economic Benefits
[0097] Performance optimization: Through performance optimization, the CO single-pass conversion rate increased by 0.51%, and the CO2 single-pass conversion rate increased by 1.74%.
[0098] Advanced prediction: predict life, performance, output, and make predictions on real-time operational changes with an accuracy rate of over 90%.
[0099] Benefit calculation: Through health management and optimization of working conditions, the single-pass conversion rate of CO and CO2 can be increased by 1.5%-4.5%, the selectivity of methanol can be increased by 0.8%-1.2%, and the yield can be increased by 0.76% by conservative estimation. Based on the designed annual production capacity of methanol of 1.8 million tons, the annual production can be increased by 13,680 tons based on the increase in yield alone, with a value of approximately RMB 27.36 million.
[0100] Energy saving: Through performance visualization, it can assist in rational energy use, reasonably reduce the circulating gas volume, and indirectly reduce the power load of the compressor by about 3.8%.
[0101] 3. Social Benefits
[0102] Environmentally friendly carbon emissions: While improving performance, the CO2 carbon conversion rate is improved, that is, the carbon utilization rate and carbon efficiency are improved, while the CO2 conversion rate is improved, directly reducing CO2 carbon emissions by about 2%.
[0103] Innovation Demonstration: Catalyst Management Cloud is the first of its kind in the industry, a product of the deep integration of IT, DT, and OT technologies, and a concrete and in-depth practice of the national strategy of integrating industrialization and informationization. It can be summarized as four new things: new tools for catalyst intelligent management; new help in achieving cost reduction and efficiency improvement; new ways to go from digitalization to intelligence; and big data empowering new industrial fields. BRIEF DESCRIPTION OF THE DRAWINGS
[0104] Figure 1 This is a schematic diagram of the data monitoring home page of the present invention;
[0105] Figure 2 This is a schematic diagram of material data monitoring of the present invention;
[0106] Figure 3 This is a schematic diagram of process data monitoring of the present invention;
[0107] Figure 4 This is a schematic diagram of performance data monitoring of the present invention;
[0108] Figure 5 This is a schematic diagram of monitoring key indicators of the synthesis tower of the present invention;
[0109] Figure 6 It is a schematic diagram of the state analysis and evaluation of the present invention;
[0110] Figure 7 This is a schematic diagram of the correlation analysis of indicators of the present invention;
[0111] Figure 8 A schematic diagram of health assessment and fault diagnosis of the present invention;
[0112] Fig. 9 This is a schematic diagram of the optimal operating range of the indicators of the present invention;
[0113] Fig.10This is a schematic diagram of the core indicator association information of the present invention;
[0114] Fig.11 This is a schematic diagram of catalyst life prediction and decision-making in the present invention;
[0115] Fig.12 This is a management flow chart of the catalyst intelligent management system of the present invention;
[0116] Fig.13 The catalyst intelligent management system data and business process diagram of the present invention;
[0117] Fig.14 This is a technical architecture diagram of the catalyst intelligent management system of the present invention;
[0118] Fig.15 This is a functional architecture diagram of the catalyst intelligent management system of the present invention;
[0119] Fig.16 It is a schematic diagram of the performance state distribution of the present invention;
[0120] Fig.17 This is a schematic diagram of the digital twin of the catalyst temperature field of the present invention;
[0121] Fig.18 This is a schematic diagram of the health curve of the present invention;
[0122] Fig.19 This is a schematic diagram of the degradation risk warning of the present invention;
[0123] Fig. 20 It is a schematic diagram of the performance evaluation of the present invention;
[0124] Fig.21 This is a schematic diagram of real-time optimization of the performance of the present invention;
[0125] Fig. 22 This is a schematic diagram of the optimal interval query of the present invention;
[0126] Fig.23 This is a schematic diagram of the life and performance prediction of the present invention;
[0127] Fig.24 This is a schematic diagram of short-term prediction of output and quality of the present invention;
[0128] Fig.25 It is a schematic diagram of the performance evaluation of the present invention;
[0129] Fig.26 Schematic diagram of the performance ranking of the present invention. DETAILED DESCRIPTION
[0130] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0131] See also Figure 1-26 ,The present invention provides a technical solution: an intelligent prediction and management system for the service life of a methanol synthesis catalyst;
[0132] like Fig.15 As shown, the catalyst intelligent management system is divided into front-end display and back-end management;
[0133] The front-end display mainly includes:
[0134] The core homepage collects the main information of each module; including health index monitoring, early warning status, benefit status, performance monitoring, conversion rate analysis, performance optimization, life prediction, quality and yield prediction and other functions, analyzing and displaying the operation status and overall health status of the catalyst from a global perspective;
[0135] Data monitoring and early warning, such as Figure 1 As shown, it mainly includes physical properties, processes, performance index monitoring, digital twin of synthesis tower temperature field and early warning statistical analysis system; data monitoring and early warning monitors the operating status of key indicators such as materials, processes, and performance of the methanol synthesis tower, and analyzes and displays them according to different dimensions; it warns of alarm indicators, displays alarm information, and analyzes alarm distribution; at the same time, it ranks and evaluates the completion of each team of key indicators; the system includes key indicator card module, key indicator trend analysis module, key indicator team ranking module, hot spot temperature monitoring and analysis module, key performance indicator monitoring and analysis module, indicator alarm monitoring module, and alarm distribution statistics module;
[0136] The key indicator card module mainly displays the real-time operation data of multiple synthetic tower public indicators, analyzes the growth and decline compared with the average of the previous working day and the average of yesterday, and pops up a window to view the historical trend chart of the indicator;
[0137] The key indicator trend analysis module performs trend analysis on multiple key public indicators of the synthesis tower according to different time dimensions, including indicator values and average values;
[0138] The key indicator team ranking module counts the indicator completion ranking of each team in the corresponding time dimension;
[0139] The hot spot temperature monitoring and analysis module monitors the real-time temperature and change rate of the hot spot bed of each synthesis tower in real time, and performs trend analysis and comparison of each hot spot temperature according to different time dimensions;
[0140] The key performance indicator monitoring and analysis module monitors the key performance indicators of each synthesis tower in the form of indicator cards or circular progress bars to monitor the operation status and change rate of each indicator in different time dimensions;
[0141] The indicator alarm monitoring module scrolls and displays all the indicator alarm information of the synthesis tower, summarizes and displays the alarm times by indicator type, checks the historical alarm trends of various indicators, and statistically analyzes the ranking of the alarm times of each indicator;
[0142] The alarm distribution statistics module ranks indicators by the number of alarms, drawing users' attention to indicators of alarm frequency, and ranks them by indicator type.
[0143] Data monitoring and early warning also include:
[0144] Material data monitoring: such as Figure 2 As shown, the operating status of key material indicators of the methanol synthesis tower is monitored and analyzed and displayed according to different dimensions; alarm indicators are alarmed, alarm information is displayed, and alarm processing status is statistically analyzed; the system includes material component monitoring and analysis module, key material monitoring module, indicator alarm statistics module, and alarm information module;
[0145] Process data monitoring: such as Figure 3 As shown, the operating status of the key process indicators of the methanol synthesis tower is monitored and analyzed and displayed according to different dimensions; alarm indicators are alarmed, alarm information is displayed, and alarm processing status is counted; the system includes key process indicator monitoring and analysis module, related process indicator monitoring module, indicator alarm statistics module, and alarm information module;
[0146] Performance data monitoring: Figure 4 As shown, the operating status of the key performance indicators of the methanol synthesis tower is monitored and analyzed and displayed according to different dimensions; alarm indicators are alarmed, alarm information is displayed, and alarm processing status is counted; the system includes a key performance indicator monitoring module, a performance status monitoring module, an alarm ranking module, an alarm statistics module, and an alarm information module;
[0147] Monitoring of key indicators of synthesis tower: Figure 5 As shown, the operating status of key indicators of the methanol synthesis tower is monitored, and the temperature of each bed layer of the synthesis tower is analyzed in multiple dimensions; including key indicator monitoring module and layer temperature monitoring and analysis module.
[0148] Status analysis and evaluation, such as Figure 6As shown, it includes core systems such as correlation analysis, health management, performance evaluation, and operating condition optimization. The prediction and auxiliary decision-making modules include life prediction, three-year performance prediction, and comprehensive performance evaluation. The status analysis and evaluation also include catalyst life prediction and decision-making, which are used to comprehensively predict the operation of the synthesis tower from multiple angles. The catalyst operation time, total processing capacity, and total output are predicted, and the actual operation data, prediction data, and estimated remaining capacity are displayed; at the same time, the load situation and the health of the catalyst are displayed; according to the three-year life cycle of the catalyst, the core indicators such as conversion rate, activity, yield, and reaction pressure difference are predicted when the three-year period is reached; and assistance is provided for users to optimize process indicators and make catalyst replacement decisions;
[0149] Status analysis and evaluation specifically include:
[0150] Indicator correlation analysis: Figure 7 As shown, the operation data of each synthesis tower and the main system are monitored in the form of DCS screen; a key indicator can be selected, the correlation analysis of the indicator can be performed, the correlation coefficient between the target indicator and other indicators in the screen can be calculated, and the strongly correlated indicator can be found;
[0151] Health assessment and fault diagnosis: Figure 8 As shown, the health of key indicators of each synthesis tower is evaluated, and the actual operating status of the indicators is evaluated. The matching and deviation between the health curve can be evaluated; the operating status of the indicators can be monitored in real time, the health value can be calculated in real time, and the real-time parameters can be compared with the health threshold; the indicator operation status can be warned in real time, the warning information can be displayed, and the number of various warnings can be counted;
[0152] Query the optimal interval of indicators: Fig. 9 As shown, it is mainly used for users to query the operating range of the specified indicator under each time dimension, obtain relevant influencing factors through correlation analysis, and calculate the optimal operating range of the target indicator and the healthy range and reference value of the relevant factors by specifying the operating range of the influencing factors of concern, so as to form a production plan under the specified working conditions and constraints to guide user operations;
[0153] Health indicator related information query: Fig.10 As shown, it is used to query the correlation between a certain indicator and other indicators within a certain working condition range, and provide users with the optimal values of each related indicator under the selected conditions to guide user operations;
[0154] Performance status evaluation and impact: Based on the indicator health curve, the actual operation status of the key indicators of the synthesis tower is evaluated. The deviation of the actual operation data of the indicators from the health curve is evaluated from multiple time dimensions. The total operation time, efficient operation time, inefficient operation time, parking time and the corresponding proportion of the synthesis tower in different time dimensions are counted. The operation status of the main related indicators is analyzed, so as to evaluate the performance status of a specific indicator.
[0155] Real-time optimization of performance indicators: real-time display of current working condition information, real-time display of current value and recommended value of given target indicators, real-time display of current value and recommended value of related parameters, and guidance of user operation;
[0156] Catalyst life prediction: Fig.11 As shown, the operation of the synthesis tower is comprehensively predicted from multiple angles, and the catalyst operation time, total processing capacity, and total output are predicted. The actual operation data, forecast data, and estimated remaining capacity are displayed; at the same time, the load situation and the health of the catalyst are displayed; according to the three-year life cycle of the catalyst, the core indicators such as conversion rate, activity, yield, and reaction pressure difference are predicted when the three-year period is reached; and assistance is provided for users to optimize process indicators and make catalyst replacement decisions.
[0157] Catalyst life prediction and decision-making specifically include:
[0158] Quality and output forecast: predict the crude methanol output and water content in the next hour, as well as the impact time and overall change trend, and provide correlation indicators that affect the change of indicators;
[0159] Life and performance prediction: including life prediction based on core indicators such as conversion rate, defining the worst degradation boundary value, and predicting the service life, remaining processing volume, and product volume through prediction models based on historical data and real-time data; in addition, according to the actual management standard requirements, develop a prediction of the final performance indicators based on the end of the three-year use period of the catalyst to provide operational decisions for catalyst maintenance, replacement, and adjustment;
[0160] Performance management: Based on the performance indicators formed by each module, performance management characteristic values are formed, and performance ranking and performance analysis are performed to combine with offline management to improve the overall management closed loop.
[0161] Back-end management mainly includes system management, security management and audit management.
[0162] The system management of back-end management includes three systems: organization management, user management, and indicator management. Organization management mainly includes functions such as adding, modifying, and deleting organizations; user management mainly includes functions such as adding users, associating users with organizations, modifying user information, and freezing and unfreezing users; indicator management mainly includes health indicator management and optimization indicator management. It can define health thresholds and optimization target thresholds online, and provide definition and modification functions for process indicator management;
[0163] Security management includes three menus: menu management, role management, and user log management. Menu management is mainly used to display all function menus of the system; role management mainly includes functions such as role addition, role modification, role deletion, user role association, and role authorization; user logs are mainly used to audit user operation behaviors;
[0164] Audit management mainly audits the operational behaviors of system administrators and security administrators.
[0165] This system is a management system that serves actual business and assists the operation of catalysts. If you want to achieve a truly closed-loop management of a series of links from monitoring perception, problem discovery, problem push, problem solving, effect testing, and continuous improvement, you need to combine online management with offline management to achieve soft closed-loop management from software to control to business optimization. The main management processes are as follows: Fig.12 shown.
[0166] The key indicators of data monitoring and early warning include real-time display of key indicators such as total methanol production of the reaction system, total synthesis gas consumption, load rate, conversion rate, and team ranking;
[0167] Basic indicators include the temperature trend of the methanol synthesis tower, temperature change rate, product yield, conversion rate and other basic indicators are displayed in real time;
[0168] The alarm information includes the alarm trend of the methanol synthesis tower, which is displayed according to the dimensions of materials, processes, and performance, and the proportion is statistically analyzed;
[0169] The main data source of the system is DCS, which mainly includes design parameters, process parameters, physical parameters, performance parameters, and state parameters; it includes not only real-time parameters of temperature, pressure, flow, composition, and liquid level, but also derivative, analysis, and judgment parameters, including:
[0170] Design parameters: filling quantity, form, physical properties and process design parameters;
[0171] Process parameters: reactor inlet temperature, pressure, temperature at each point of the bed, temperature rise, reaction pressure, pressure drop, liquid level, outlet temperature, outlet pressure, reaction heat, drum pressure, temperature, steam temperature, water supply temperature; raw gas (circulating gas, mixed gas) pressure and temperature;
[0172] Physical parameters: raw gas volume, composition, components; circulating gas volume, temperature, pressure, components, components; tower inlet gas volume, temperature, components; tower outlet temperature, pressure, composition, components; product output, by-product output;
[0173] Performance parameters: various raw material conversion rates, selectivity, yield, activity, health index, and space velocity performance indicators derived from the original point through the mechanism model;
[0174] Status parameters: instrument running status, stop status, alarm status;
[0175] Set parameters: warning threshold, warning upper limit, alarm lower limit, encoding method, and range.
[0176] Combination Fig.13 As shown in the figure, the data business flow shows that after the data of each module is collected, it enters the data platform for aggregation and management. The required performance and other data are calculated in real time through the mechanism formula and stored in the InfluxDB real-time database.
[0177] The stored data is processed into derived data using the mechanism formula for monitoring and sensing module output. Through correlation analysis, the relationship between input, output, materials, process and performance data is established. The correlation relationship is embodied in data, such as positive and negative correlation, correlation coefficient, etc., and the correlation is displayed in combination with the actual process.
[0178] At the same time, the data correlation relationship, field experience and reaction mechanism are used to divide the actual working conditions and data sets under various working conditions; the input data is trained through the advanced prediction model to simulate the health values under various subdivided working conditions, and the health curve showing the current working conditions is predicted in the time dimension. According to the on-site engineers and process card management standards, the upper and lower thresholds based on the health curve are defined; for the values of the health indicator fluctuations of concern, the main cause analysis is carried out, and the top 6 most manageable factors are presented in real time to form data samples, which enter the knowledge and experience module and solidify into experience knowledge. At the same time, the parameters are used as the boundary data of the optimization module; for abnormal indicator fluctuations, which cause the catalyst performance to decline and cause catalyst degradation failure data, early warning prompts are issued, and the changes in the main related factors are analyzed to find out the changes, and the formed data samples enter the health management to optimize the health management model and become the data boundary for working condition optimization; according to the early warning indicators, the early warning statistics module is included, and the data is used as a performance indicator and displayed in the monitoring module at the same time;
[0179] For the normal operating conditions and index data, the performance optimization algorithm model is used for processing, and the data control plan with stable optimization is output for reference and optimization by operators; the optimized results are evaluated for performance, and the optimized samples and experience are solidified as knowledge data; at the same time, auxiliary optimization is performed on the data results of degradation analysis, and the optimized performance evaluation is also performed; the data after performance evaluation is used as input data for the performance management module;
[0180] The input and output data of the above modules are used as input data of the prediction module. The multi-algorithm prediction model is used to output relatively accurate prediction data to make long-term predictions on service life and performance; at the same time, short-term predictions are made on output and product quality to assist management and operators in making decisions and judgments; the core result data of the above modules are displayed as the content of the comprehensive homepage, so that decision-makers and managers can quickly understand the operation status of methanol synthesis conditions, catalyst performance, performance optimization and the performance of the operating team;
[0181] The performance indicators in management will be included in the performance ranking and comprehensive performance evaluation system, which will conduct comprehensive analysis and evaluation on the management effectiveness of control parameters, the optimization performance of performance indicators, and the data implemented to economic benefits, provide support for the closed-loop data cycle of system data recycling and re-optimization, and ensure a virtuous cycle of seeking excellence in stability and selecting the best from the best.
[0182] Combination Fig.14 As shown in the figure, the underlying technical architecture of the system is based on the Hadoop architecture, and uses the MapReduce and Spark computing frameworks to achieve batch storage and computing of data; on the IOT side, the MQTT protocol is used to take over the data on the network gate side; the Kafka message subscription and publishing system is used to manage messages; the flink batch stream combined processing framework is used to ensure the system's millisecond / second processing capability for massive data; on the platform service side, the overall microservice architecture system is adopted, the algorithm model design is implemented through the Cloudiip-lab component, the digital twin model design is implemented through the HT component, and the registration and verification of the algorithm model are uniformly managed through the Zookeeper distributed collaboration framework.
[0183] The system covers three types of industrial models: data algorithm model, industrial mechanism model, and business process model, including:
[0184] a. Data algorithm model
[0185] Advanced warning model: Based on the health warning management of the advanced warning model application, the data of the normal working condition of the catalyst is selected to build the advanced warning model to realize the prediction of the health value; the difference between the actual value and the predicted value is evaluated based on the approximation evaluation algorithm to realize the early warning of abnormal indicators; to achieve the purpose of early intervention in process operation and ensure the long-term healthy operation of the catalyst;
[0186] Optimal parameter sample parameter model: The optimal process sample library is constructed based on the partitioning algorithm. The partitioning algorithm is used to cluster the on-site working conditions to achieve the division of working conditions. The working condition identification model is established based on the working condition division results to calculate the optimal process parameters for each working condition. The production data is collected in real time to identify the working condition, and the optimal process parameters under the working condition are found and recommended to the operator for production guidance.
[0187] Catalyst service life and performance prediction model: Catalyst service life prediction based on machine learning, building a regression model and establishing a machine learning mechanism based on indicators such as conversion rate, output, activity, moisture content, and usage time, using a supervised algorithm to predict the remaining life of the catalyst; effectively guiding production scheduling, catalyst procurement, and catalyst replacement business;
[0188] Synthesis tower heat distribution model: Based on the digital twin, the synthesis tower heat distribution model refers to the actual equipment structure, measurement point distribution, and synthesis tower temperature data of the synthesis tower to achieve intuitive real-time monitoring of the temperature field distribution of each bed layer of the synthesis tower, providing a strong decision-making basis for reactor operation status evaluation, abnormal status diagnosis, and reaction trend prediction;
[0189] b. Industrial Mechanism Model
[0190] Series and parallel process flow accounting model: The mechanism-based series and parallel process flow accounting model uses part of the flow, components, and physical property balance data of the feed gas, circulating gas, and mixed gas of towers 1 and 2, combined with the actual series and parallel process flow of towers 1 and 2, and uses the ideal gas state equation, Dalton's law of partial pressure and other mechanism methods to convert mass and volume flow. In addition, according to the mechanism principle of element balance, the flow data of the reactors without series and parallel connection is calculated, and the mixed gas data is intelligently predicted; the technical transformation of adding flow meters and component measurement points is avoided;
[0191] Catalyst performance characteristics visualization model: Using the mechanism formula, the performance index value is calculated in real time, combined with the circulation ratio, hydrogen-carbon ratio and other working condition division indicators, the operating performance change characteristics and relationships are established and displayed in multiple dimensions, and the performance spatial change model is established for visualization and perception;
[0192] c. Management process model
[0193] Through data collection from various modules of the system, health thresholds and optimization range thresholds can be customized in combination with business needs. Combined with the management of these thresholds, performance evaluation indicators for practice and operation processes are formed. Based on the performance evaluation indicators, performance rankings are conducted on each management object, and comprehensive performance evaluation and management are conducted at different time granularities to form a closed-loop management model for perception, alarm, intervention, health protection, working condition optimization, performance evaluation, and prediction.
[0194] The following are specific application cases:
[0195] 1. To address the issues of unstable catalyst performance, large fluctuations, and lack of awareness of the current status.
[0196] The monitoring and perception module of the catalyst operation management system realizes the real-time visualization of catalyst performance, and the mechanism formula is used to realize data characterization. Combined with the digital twin technology, it automatically draws multi-parameter performance state diagrams including space velocity, hydrogen-carbon ratio, conversion rate, selectivity, temperature rise, etc., temperature field diagrams, and performance condition scatter distribution diagrams (see Fig.16 ) and other visual display methods, which realize the use of research-level auxiliary tools in practice, and can quickly and intuitively determine the performance status, working condition matching, optimization direction determination, etc. The catalyst operation status is realized from seeing, seeing clearly to foreseeing details and situation awareness. Among them, the three-dimensional modeling of the reactor catalyst in this module (see Fig.17 ), realize the digital twin of the temperature field, realize real-time monitoring and abnormal alarm. Visualize the relationship between temperature field and performance, and provide decision-making basis for operation and research.
[0197] 2. It is difficult to prevent the degradation and damage of the catalyst during operation, and its healthy operation cannot be guaranteed.
[0198] By combining multiple algorithms and deeply integrating business operations to establish a health management model, we can extract health standard lines applicable to different working conditions from historical data and engineer experience (see Fig.18 ), based on the standard line, online physical examination, diagnosis and management of the catalyst operation monitoring status are carried out to ensure the healthy operation of the catalyst.
[0199] The health management module can be used to evaluate the current operating status and performance of the catalyst from multiple perspectives and angles. The advanced prediction model is used to train the historical data to find the catalyst operating health guidance value and the corresponding threshold value, and the real-time value is used to compare them. If the threshold value is exceeded, it is considered to be deviating from the healthy trend. Through degradation analysis, the abnormal indicators in the historical data are modeled, the degradation sample data is found, and the degradation model and fault model are established. At the same time, the degree of catalyst degradation, the influencing factors, the main factors of degradation, etc. are compared, analyzed and displayed in real time to perceive the degradation trend in advance. This can provide early warning for abnormal indicators, catalyst damage, degradation prediction and early warning (see Fig.19 ), reduce catalyst damage and keep the catalyst running healthy and stable.
[0200] Through performance evaluation, the current status, historical status, and concurrent status are analyzed and evaluated, compared, and control parameters and process parameters that need to be improved are pushed, providing decision-making basis for operators (see Fig. 20 ).
[0201] 3. It is difficult to share experience in management, operation, and maintenance during the operation of the catalyst, and it is difficult to optimize performance.
[0202] The module for finding the best solution in stable working condition (see Fig.21 , Fig. 22), maintain high-activity operation, take performance data as the optimization target, and select high-performance corresponding state data from healthy operation history data through optimization model. After verification of design data and process card data, combined with empirical rules, push single or multi-objective optimization control and operation plans. The plan combines mechanism, experience, history and current situation factors, and has the characteristics of easy operation, easy achievement, stable regulation, and reliable optimization.
[0203] The performance optimization module uses the operating condition optimization algorithm and sample data under healthy conditions to find the optimal performance and operating conditions, push operation plans, and guide production, while maintaining the healthy and stable operation of the catalyst. It provides necessary data query and analysis functions to facilitate technical research, accumulation, and experience inheritance.
[0204] 4. There is no means to predict the remaining service life and performance changes of the catalyst in the future, and there is no basis for maintenance.
[0205] The performance prediction module of the system can be used to build a regression model based on indicators such as conversion rate, output, activity, moisture content, and usage time. The established machine learning mechanism uses a supervised composite prediction algorithm model to predict the remaining life of the catalyst and the performance indicators at the end of the operation period (see Fig.23 ). It can achieve qualitative and quantitative prediction of catalyst performance and life, and assist in guiding production scheduling, maintenance, and replacement planning. At the same time, it can also predict the output and quality indicators of products in the short term (see Fig.24 ), prompting operators to make corresponding adjustments in advance to maintain high-performance operation of the catalyst.
[0206] 5. Post-operation management of catalysts makes operation and maintenance difficult to manage and control in advance.
[0207] The catalyst operation management system can use the health curve drawn by the advanced prediction model, combined with the customized control thresholds in the process, to perform health control on the operation data, issue early warnings before parameter deterioration, intervene in advance, and assist operators to adjust deviating parameters to a healthy range to avoid degradation. At the same time, the system will predict the changing trends of performance, output, and quality in advance according to the changes in input data, and remind operators of the downward trend of the predicted value of the target indicator to optimize performance. Maintain continuous high-performance operation of the catalyst. Transform time analysis and management into advance warning, intervention, and management. Improve the predictability and timeliness of problem perception, and assist in the realization of advance management of catalyst operation (see Fig.25 , Fig.26 ).
[0208] Contents not described in detail in this specification belong to the prior art known to professional and technical personnel in the field. Although the embodiments of the present invention have been shown and described, it is understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the attached claims and their equivalents.
Claims
1. An intelligent prediction and management system for the service life of a methanol synthesis catalyst, characterized in that: The catalyst intelligent management system is divided into front-end display and back-end management; The front-end display mainly includes: The core homepage collects the main information of each module; including health index monitoring, early warning status, benefit status, performance monitoring, conversion rate analysis, performance optimization, life prediction, quality and yield prediction and other functions, analyzing and displaying the operation status and overall health status of the catalyst from a global perspective; Data monitoring and early warning, mainly including physical properties, processes, performance indicators monitoring, synthesis tower temperature field digital twin and early warning statistical analysis system; the data monitoring and early warning is to monitor the operating status of key indicators such as materials, processes, performance of the methanol synthesis tower, and analyze and display them according to different dimensions; alarm indicators are alarmed, alarm information is displayed, and alarm distribution is analyzed; at the same time, the completion of each team of key indicators is ranked and evaluated; the system includes key indicator card module, key indicator trend analysis module, key indicator team ranking module, hot spot temperature monitoring and analysis module, key performance indicator monitoring and analysis module, indicator alarm monitoring module, and alarm distribution statistics module; Status analysis and evaluation, including core systems such as correlation analysis, health management, performance evaluation, and operating condition optimization, and prediction and decision-making assistance modules including life prediction, three-year performance prediction, and comprehensive performance evaluation; the status analysis and evaluation also includes catalyst life prediction and decision-making, which is used to comprehensively predict the operation of the synthesis tower from multiple angles, predict the catalyst operation time, total processing capacity, and total output, and display actual operation data, prediction data, and estimated remaining capacity; at the same time, it displays the load situation and the health of the catalyst; according to the three-year life cycle of the catalyst, it predicts the core indicators such as conversion rate, activity, yield, and reaction pressure difference when the three-year period is reached; and provides assistance for users to optimize process indicators and make catalyst replacement decisions; The back-end management mainly includes system management, security management and audit management.
2. The intelligent prediction and management system for service life of a methanol synthesis catalyst according to claim 1, characterized in that: The key indicators of the data monitoring and early warning include real-time display of key indicators such as total methanol production of the reaction system, total synthesis gas unit consumption, load rate, conversion rate, and team ranking; Basic indicators include the temperature trend of the methanol synthesis tower, temperature change rate, product yield, conversion rate and other basic indicators are displayed in real time; The alarm information includes the alarm trend of the methanol synthesis tower, which is displayed according to the dimensions of materials, processes, and performance, and the proportion is statistically analyzed; The main data source of the system is DCS, which mainly includes design parameters, process parameters, physical parameters, performance parameters, and state parameters; it includes not only real-time parameters of temperature, pressure, flow, composition, and liquid level, but also derivative, analysis, and judgment parameters, including: Design parameters: filling quantity, form, physical properties and process design parameters; Process parameters: reactor inlet temperature, pressure, temperature at each point of the bed, temperature rise, reaction pressure, pressure drop, liquid level, outlet temperature, outlet pressure, reaction heat, drum pressure, temperature, steam temperature, water supply temperature; raw gas (circulating gas, mixed gas) pressure and temperature; Physical parameters: raw gas volume, composition, components; circulating gas volume, temperature, pressure, components, components; tower inlet gas volume, temperature, components; tower outlet temperature, pressure, composition, components; product output, by-product output; Performance parameters: various raw material conversion rates, selectivity, yield, activity, health index, and space velocity performance indicators derived from the original point through the mechanism model; Status parameters: instrument running status, stop status, alarm status; Set parameters: warning threshold, warning upper limit, alarm lower limit, encoding method, and range.
3. The intelligent prediction and management system for service life of a methanol synthesis catalyst according to claim 1, characterized in that: The key indicator card module mainly displays the real-time operation data of multiple synthetic tower public indicators, analyzes the growth and decline compared with the working average and yesterday's average, and pops up a window to view the historical trend chart of the indicator; The key indicator trend analysis module performs trend analysis on multiple key public indicators of the synthesis tower according to different time dimensions, including indicator values and mean values; The key indicator team ranking module counts the indicator completion ranking of each team in the corresponding time dimension; The hot spot temperature monitoring and analysis module monitors the real-time temperature and change rate of the hot spot bed of each synthesis tower in real time, and performs trend analysis and comparison of the hot spot temperature according to different time dimensions; The key performance indicator monitoring and analysis module monitors the operation and change rate of each indicator in different time dimensions in the form of indicator cards or circular progress bars for each key performance indicator of each synthesis tower; The indicator alarm monitoring module scrolls and displays all the synthesizer indicator alarm information, summarizes and displays the alarm times by indicator type, checks the historical alarm trends of various indicators, and statistically analyzes the ranking of the alarm times of each indicator; The alarm distribution statistics module ranks the indicators by the number of alarms, causing users to focus on the indicators of alarm frequency, and ranks them by indicator type.
4. The intelligent prediction and management system for service life of a methanol synthesis catalyst according to claim 1, characterized in that: The data monitoring and early warning also includes: Material data monitoring: monitor the operating status of key material indicators of the methanol synthesis tower, analyze and display them in different dimensions; issue alarms for alarm indicators, display alarm information, and count alarm processing status; the system includes material component monitoring and analysis module, key material monitoring module, indicator alarm statistics module, and alarm information module; Process data monitoring: monitor the operating status of key process indicators of the methanol synthesis tower, analyze and display them in different dimensions; issue alarms for alarm indicators, display alarm information, and count alarm processing status; the system includes key process indicator monitoring and analysis modules, related process indicator monitoring modules, indicator alarm statistics modules, and alarm information modules; Performance data monitoring: monitor the operating status of key performance indicators of the methanol synthesis tower, analyze and display them in different dimensions; issue alarms for alarm indicators, display alarm information, and count alarm processing status; the system includes key performance indicator monitoring module, performance status monitoring module, alarm ranking module, alarm statistics module, and alarm information module; Monitoring of key indicators of the synthesis tower: monitor the operating status of key indicators of the methanol synthesis tower and conduct multi-dimensional analysis of the temperature of each bed layer of the synthesis tower; including key indicator monitoring module and layer temperature monitoring and analysis module.
5. The intelligent prediction and management system for service life of methanol synthesis catalyst according to claim 1, characterized in that: The status analysis and evaluation specifically include: Indicator correlation analysis: Monitor the operating data of each synthesis tower and main system in the form of DCS screen; select a key indicator, conduct correlation analysis on the indicator, calculate the correlation coefficient between the target indicator and other indicators in the screen, and find out the strongly correlated indicators; Health assessment and fault diagnosis: Evaluate the health of key indicators of each synthesis tower, and evaluate the matching and deviation between the actual operating status of the indicators and the health curve; monitor the operating status of indicators in real time, calculate the health value in real time, and compare the real-time parameters with the health threshold; issue early warnings on the operating status of indicators in real time, display early warning information, and count the number of various early warnings; Health indicator correlation information query: used to query the correlation degree between a certain indicator and other indicators within a certain working condition range, and provide users with the optimal values of each related indicator under the selected conditions to guide user operations; Performance status evaluation and impact: Based on the indicator health curve, the actual operation status of the key indicators of the synthesis tower is evaluated. The deviation of the actual operation data of the indicators from the health curve is evaluated from multiple time dimensions. The total operation time, efficient operation time, inefficient operation time, parking time and the corresponding proportion of the synthesis tower in different time dimensions are counted. The operation status of the main related indicators is analyzed, so as to evaluate the performance status of a specific indicator. Real-time optimization of performance indicators: real-time display of current working condition information, real-time display of current value and recommended value of given target indicators, real-time display of current value and recommended value of related parameters, and guidance of user operation; Query of optimal index interval: mainly used for users to query the operating interval of the specified index under each time dimension, obtain relevant influencing factors through correlation analysis, calculate the optimal operating interval of the target index and the healthy interval and reference value of the relevant factors by specifying the operating interval of the influencing factors, form a production plan under the specified working conditions and constraints, and guide user operations; Catalyst life prediction: Comprehensively predict the operation of the synthesis tower from multiple angles to predict the catalyst operation time, total processing capacity, and total output, and display actual operation data, forecast data, and estimated remaining capacity; also display the load situation and the health of the catalyst; predict the core indicators such as conversion rate, activity, yield, and reaction pressure difference when the three-year life cycle of the catalyst is reached; provide assistance for users to optimize process indicators and make catalyst replacement decisions.
6. The intelligent prediction and management system for service life of a methanol synthesis catalyst according to claim 1, characterized in that: The catalyst life prediction and decision-making specifically include: Quality and output forecast: predict the crude methanol output and water content in the next hour, as well as the impact time and overall change trend, and provide correlation indicators that affect the change of indicators; Life and performance prediction: including life prediction based on core indicators such as conversion rate, defining the worst degradation boundary value, and predicting the service life, remaining processing volume, and product volume through prediction models based on historical data and real-time data; in addition, according to the actual management standard requirements, develop a prediction of the final performance indicators based on the end of the three-year use period of the catalyst to provide operational decisions for catalyst maintenance, replacement, and adjustment; Performance management: Based on the performance indicators formed by each module, performance management characteristic values are formed, and performance ranking and performance analysis are performed to combine with offline management to improve the overall management closed loop.
7. The intelligent prediction and management system for service life of methanol synthesis catalyst according to claim 1, characterized in that: The system management of the back-end management includes three systems: organization management, user management, and indicator management. Organization management mainly includes functions such as adding, modifying, and deleting organizations; user management mainly includes functions such as adding users, associating users with organizations, modifying user information, freezing and unfreezing users; indicator management mainly includes health indicator management and optimization indicator management, which can define health thresholds and optimization target thresholds online, and provide definition and modification functions for process indicator management; The security management includes three menus: menu management, role management, and user log management. Menu management is mainly used to display all function menus of the system; role management mainly includes functions such as role addition, role modification, role deletion, user role association, and role authorization; user logs are mainly used to audit user operation behaviors; The audit management mainly audits the operational behaviors of system administrators and security administrators.
8. The intelligent prediction and management system for service life of methanol synthesis catalyst according to claim 1, characterized in that: The data service flow shows that after the data of each module is collected, it enters the data platform for aggregation and management, and the required performance and other data are calculated in real time through the mechanism formula and stored in the InfluxDB real-time database; The stored data is processed into derived data using the mechanism formula for monitoring and sensing module output. Through correlation analysis, the relationship between input, output, materials, process and performance data is established. The correlation relationship is embodied in data, such as positive and negative correlation, correlation coefficient, etc., and the correlation is displayed in combination with the actual process. At the same time, the data correlation, field experience and reaction mechanism are used to divide the actual working conditions and data sets under various working conditions; The input data is trained through the advanced prediction model to simulate the health values under various subdivided working conditions. The health curve of the current working condition is predicted in the time dimension. The upper and lower thresholds based on the health curve are defined according to the on-site engineers and process card management standards. For the values of the health indicator fluctuations of concern, the main cause analysis is carried out, and the top 6 most manageable factors are presented in real time to form data samples, which enter the knowledge and experience module and solidify into experience knowledge. At the same time, the parameters are used as the boundary data of the optimization module. For abnormal indicator fluctuations, which cause the catalyst performance to decline and cause catalyst degradation failure, early warning prompts are issued, and the changes in the main related factors are analyzed to find out. The formed data samples enter the health management, optimize the health management model, and become the data boundary for working condition optimization. According to the early warning indicators, they are included in the early warning statistics module, and the data are used as performance indicators and displayed in the monitoring module at the same time. For the normal operating conditions and index data, the performance optimization algorithm model is used for processing, and the data control plan with stable optimization is output for reference and optimization by operators; the optimized results are evaluated for performance, and the optimized samples and experience are solidified as knowledge data; at the same time, auxiliary optimization is performed on the data results of degradation analysis, and the optimized performance evaluation is also performed; the data after performance evaluation is used as input data for the performance management module; The input and output data of the above modules are used as input data of the prediction module. The multi-algorithm prediction model is used to output relatively accurate prediction data to make long-term predictions on service life and performance; at the same time, short-term predictions are made on output and product quality to assist management and operators in making decisions and judgments; the core result data of the above modules are displayed as the content of the comprehensive homepage, so that decision-makers and managers can quickly understand the operation status of methanol synthesis conditions, catalyst performance, performance optimization and the performance of the operating team; The performance indicators in management will be included in the performance ranking and comprehensive performance evaluation system, which will conduct comprehensive analysis and evaluation on the management effectiveness of control parameters, the optimization performance of performance indicators, and the data implemented to economic benefits, provide support for the closed-loop data cycle of system data recycling and re-optimization, and ensure a virtuous cycle of seeking excellence in stability and selecting the best from the best.
9. The intelligent prediction and management system for service life of methanol synthesis catalyst according to claim 1, characterized in that: The underlying technical architecture of the system is based on the Hadoop architecture, and uses the MapReduce and Spark computing frameworks to achieve batch storage and computing of data; on the IOT side, the MQTT protocol is used to take over the data on the network gate side; the Kafka message subscription and publishing system is used to manage messages; the flink batch stream combined processing framework is used to ensure the system's millisecond / second processing capabilities for massive data; on the platform service side, the overall microservice architecture system is adopted, the algorithm model design is implemented through the Cloudiip-lab component, the digital twin model design is implemented through the HT component, and the registration and verification of the algorithm model are uniformly managed through the Zookeeper distributed collaboration framework.
10. The intelligent prediction and management system for service life of methanol synthesis catalyst according to claim 1, characterized in that: The system covers three types of industrial models: data algorithm model, industrial mechanism model, and business process model, specifically including: a. Data algorithm model Advanced warning model: Based on the health warning management of the advanced warning model application, the data of the normal working condition of the catalyst is selected to build the advanced warning model to realize the prediction of the health value; the difference between the actual value and the predicted value is evaluated based on the approximation evaluation algorithm to realize the early warning of abnormal indicators; to achieve the purpose of early intervention in process operation and ensure the long-term healthy operation of the catalyst; Optimal parameter sample parameter model: The optimal process sample library is constructed based on the partitioning algorithm. The partitioning algorithm is used to cluster the on-site working conditions to achieve the division of working conditions. The working condition identification model is established based on the working condition division results to calculate the optimal process parameters for each working condition. The production data is collected in real time to identify the working condition, and the optimal process parameters under the working condition are found and recommended to the operator for production guidance. Catalyst service life and performance prediction model: Catalyst service life prediction based on machine learning, building a regression model and establishing a machine learning mechanism based on indicators such as conversion rate, output, activity, moisture content, and usage time, using a supervised algorithm to predict the remaining life of the catalyst; effectively guiding production scheduling, catalyst procurement, and catalyst replacement business; Synthesis tower heat distribution model: Based on the digital twin, the synthesis tower heat distribution model refers to the actual equipment structure, measurement point distribution, and synthesis tower temperature data of the synthesis tower to achieve intuitive real-time monitoring of the temperature field distribution of each bed layer of the synthesis tower, providing a strong decision-making basis for reactor operation status evaluation, abnormal status diagnosis, and reaction trend prediction; b. Industrial Mechanism Model Series and parallel process flow accounting model: The mechanism-based series and parallel process flow accounting model uses part of the flow, components, and physical property balance data of the feed gas, circulating gas, and mixed gas of towers 1 and 2, combined with the actual series and parallel process flow of towers 1 and 2, and uses the ideal gas state equation, Dalton's law of partial pressure and other mechanism methods to convert mass and volume flow. In addition, according to the mechanism principle of element balance, the flow data of the reactors without series and parallel connection is calculated, and the mixed gas data is intelligently predicted; the technical transformation of adding flow meters and component measurement points is avoided; Catalyst performance characteristics visualization model: Using the mechanism formula, the performance index value is calculated in real time. Combined with the circulation ratio, hydrogen-carbon ratio and other working condition division indicators, the operating performance change characteristics and relationships are established and displayed in multiple dimensions, and the performance spatial change model is established for visual display and perception; c. Management process model Through data collection from various modules of the system, health thresholds and optimization range thresholds can be customized in combination with business needs. Combined with the management of these thresholds, performance evaluation indicators for practice and operation processes are formed. Based on the performance evaluation indicators, performance rankings are conducted on each management object, and comprehensive performance evaluation and management are conducted at different time granularities to form a closed-loop management model for perception, alarm, intervention, health protection, working condition optimization, performance evaluation, and prediction.
Citation Information
Cited By
Denitration catalyst performance index intelligent query system based on dynamic data fusion and query method thereof
CN120670461A
