Function analysis-based coal gasification device safety analysis and pre-judgment method

Abstract

The invention discloses a coal gasification device safety analysis and pre-judgment method based on function analysis, and relates to the field of coal gasification device safety, and the method comprises the following operation steps: S1, constructing a digital function list and a three-dimensional digital model; s2, constructing a functional network topological graph of the coal gasification device; s3, analysis and diagnosis: analyzing a fault mode by integrating the data of the coal gasification device; s4, performing risk assessment on the analyzed fault mode; and S5, performing safety pre-judgment based on the principle of a biological immune system. According to the coal gasification device safety analysis and pre-judgment method based on function analysis, through the constructed function network topological graph, the complex mutual relation between all the function units in the device can be recognized, and how a local fault affects other parts of the whole device through the mutual relation can be clearly seen; therefore, a comprehensive risk view is provided, and the missing of potential safety hazards is avoided.

Description

Technical Field

[0001] The present invention relates to the safety field of coal gasification devices, and in particular to a safety analysis and prediction method for coal gasification devices based on functional analysis. Background Art

[0002] A coal gasification unit is an industrial equipment system that converts solid coal into gaseous synthesis gas, which includes a feeding system, a gasifier, a synthesis gas cooling system, and an ash treatment system.

[0003] When conducting safety analysis and prediction of coal gasification units, most of the focus is on functional analysis of individual equipment and local subsystems in the coal gasification unit, and they cannot be fully integrated and analyzed, so that the safety analysis cannot cover the entire coal gasification unit, which will lead to safety loopholes. At the same time, there is less in-depth research on the dynamic characteristics of the coal gasification unit, and it is impossible to capture the real-time changes in the working status of each equipment in the coal gasification unit under different working conditions and the mutual influence relationship, resulting in the inability to predict the subsequent operation of the coal gasification unit.

[0004] Therefore, it is necessary to propose a safety analysis and prediction method for coal gasification units based on functional analysis to solve the above problems. Summary of the invention

[0005] The main purpose of the present invention is to provide a safety analysis and prediction method for a coal gasification device based on functional analysis, which can effectively solve the problems in the background technology.

[0006] To achieve the above object, the technical solution adopted by the present invention is:

[0007] The safety analysis and prediction method of coal gasification unit based on functional analysis includes the following steps:

[0008] S1: Build a digital function list and a three-dimensional digital model, disassemble the coal gasification unit into different subsystems according to the process flow, define and number the functions of each subsystem, and build a digital function list;

[0009] Construct a high-precision three-dimensional digital model of the coal gasification unit based on 3D laser scanning and digital modeling;

[0010] S2: Construct the functional network topology of the coal gasification unit. Based on the digital function list and the three-dimensional digital model, the functional units are used as nodes, and the material flow, energy flow and information flow between functions are used as connecting edges. The structural characteristics of the functional network topology are analyzed and dynamically simulated.

[0011] S3: Analysis and diagnosis: By integrating the real-time sensor data, historical operation data, equipment maintenance records, inspection reports and external environmental data of the coal gasification unit, a comprehensive data collection platform is built. The convolutional neural network-long short-term memory network fusion architecture is used to analyze the real-time data and the fault mode.

[0012] S4: Conduct risk assessment on the analyzed failure modes and formulate corresponding risk response strategies and suggestions according to different assessment levels;

[0013] S5: Safety prediction is performed based on the principle of biological immune system. The normal operating status parameters of the coal gasification device are defined as self-mode, and parameter changes, abnormal fluctuations and equipment failure signals beyond the normal range in the coal gasification device are regarded as non-self-mode, and safety prediction is performed on the detected non-self-mode.

[0014] Preferably, the step S1 includes the following steps:

[0015] S101: Collect detailed process flow of coal gasification unit, including coal feeding, pretreatment, gasification reaction, synthesis gas purification and cooling, ash treatment;

[0016] S102: Determine the principle of functional unit division. According to the physical structure and process function of the coal gasification unit, determine the principle of functional unit division, including division according to equipment type and process stage. Based on the independence and integrity of the functional units, each functional unit can only independently complete a specific function, and its input and output are clearly defined. For equipment and process steps that need to work together, they are combined into one functional unit;

[0017] S103: Preliminary division of functional units: According to the determined division principle, the coal gasification device is preliminarily divided into functional units, starting from coal feeding to ash treatment, and each link is gradually disassembled into functional units, and each functional unit is numbered and named using a hierarchical coding method;

[0018] S104: Functional definition: for each divided functional unit, clarify its main function, including the role of the functional unit in the entire coal gasification unit, and its processing methods for materials, energy and information;

[0019] S105: Determine the input parameters of each functional unit, including the properties, flow, temperature, pressure of the material, and the input form of energy and control signals;

[0020] S106: Functional details and feature description, describing the working characteristics of the functional unit, including but not limited to working cycle, operation mode, start and stop conditions;

[0021] S107: Integration and recording of digital function lists. Use spreadsheet software to create a digital function list table. The column headings of the table include but are not limited to the function unit number, name, subsystem, main function, input parameters, output parameters, working characteristics, and special requirements. In the digital function list, set the association relationship between the function units.

[0022] Preferably, the step S2 specifically includes the following steps:

[0023] S201: Based on the system dynamics simulation software, combined with the actual operation data of the coal gasification unit and the process mechanism model, the functional network topology diagram is dynamically simulated to simulate the state changes and mutual influence relationships of each node in the functional network topology diagram under different working conditions, and the functional state dynamic transfer matrix is ​​constructed;

[0024] S202: In the functional network topology diagram, key nodes and vulnerable links are determined through node degree distribution and clustering coefficient. Key nodes are core function executors, high flow and high energy transmission points, and control hubs in the coal gasification device; vulnerable links are high-risk material transmission links in the coal gasification device, connection components under complex working conditions, and backup links that rely on a single device;

[0025] S203: using the function state dynamic transfer matrix, observing the state changes of key nodes and vulnerable links under different working conditions;

[0026] S204: Fault propagation path analysis: sort out the fault propagation path according to the dynamic simulation results of the functional network.

[0027] Preferably, in S3, the construction of the convolutional neural network-long short-term memory network fusion architecture includes the following steps:

[0028] S301: Clean and normalize the collected sensor data, and design the CCN module based on it, including the design of the convolution layer and the pooling layer, and input the designed feature data into the LSTM layer in chronological order;

[0029] S302: A fully connected layer is connected after the LSTM layer as the output layer. The number of neurons in the output layer is the number of failures in the coal gasification unit. The Softmax activation function is used to normalize the results of the output layer so that the output results are represented as the probability distribution of each failure mode.

[0030] S303: Train and optimize the convolutional neural network-long short-term memory network fusion architecture.

[0031] Preferably, the step S4 includes the following steps:

[0032] S401: Based on the functional analysis and historical fault data of the coal gasification unit, a fault mode library is established. The fault modes include equipment failure, process failure, and external interference failure. Each fault mode is described in detail, including the manifestation characteristics of the fault, the cause of the fault, and the location of the fault.

[0033] S402: Determine the risk quantification index and the risk value. The formula is:

[0034] R k =P k +S k +C k ;

[0035] Where k is the failure mode, P k is the probability of occurrence of fault k, S k is the influence range coefficient of the failure mode, C k is the consequence severity factor of the failure mode, Rv k is the risk value of fault k, and the impact range of the fault mode is obtained based on the constructed coal gasification unit functional network topology diagram;

[0036] S403: Calculate the total risk value based on the risk value, the formula is:

[0037] TR v =∑ k R k ;

[0038] Where TRv is the total risk value;

[0039] S404: Fault probability calculation, calculating the probability of fault occurrence based on the Bayesian network method;

[0040] S405: Output of risk assessment results. Based on the calculation results of risk quantification indicators, a risk assessment report of the coal gasification unit is output. The report includes the risk value of each functional unit, the main failure mode and its probability of occurrence, and the total risk value of the entire unit, and the risk assessment results are graded.

[0041] Preferably, in S405, the classification is based on classification factors, wherein the classification factors include: probability of failure, scope of failure impact, and severity of failure consequence, wherein the probability of failure is recorded as P, and based on the result of failure probability calculation, it is divided into extremely low probability: P < 0.01; low probability: 0.01 ≤ P < 0.1; medium probability: 0.1 ≤ P < 0.5; high probability: 0.5 ≤ P;

[0042] Based on the constructed coal gasification unit functional network topology, the fault impact range is obtained, which is denoted as S and is divided into local impact, where the fault only affects a single non-critical functional unit: S < 0.2; subsystem impact, where the fault affects multiple functional units within a subsystem: 0.2 ≤ S < 0.6; system-wide impact, where the fault affects multiple subsystems of the entire coal gasification unit: 0.6 ≤ S;

[0043] The severity of the consequences of the failure is recorded as C, which is divided into minor consequences. The failure has a short-term and quickly recoverable impact on production, and basically no harm to safety and the environment: C = 1; medium consequences: The failure causes production interruption and requires certain repairs and adjustments to recover, which has potential risks to safety and the environment,

[0044] C = 3; Severe consequences: The failure causes long-term production stagnation, serious damage to equipment, and great harm to safety and the environment: C = 5; Catastrophic consequences: The failure causes an irreversible disaster:

[0045] C=7

[0046] Preferably, the definition of risk classification standards based on the risk factors includes low risk, which satisfies one of the following conditions: the probability of failure is extremely low, the scope of failure impact is local impact, and the severity of failure consequence is minor consequence;

[0047] The probability of failure is low, the scope of failure is local, and the severity of failure consequence is minor, and Rv = P × S × C < 0.2;

[0048] Medium risk: one of the following conditions is met: the probability of failure is medium, the scope of failure impact is local, the severity of failure consequence is minor, and 0.2≤Rv<0.6;

[0049] The probability of failure is low, the scope of failure impact is subsystem impact, the severity of failure consequence is medium, and 0.2≤Rv<0.6;

[0050] High risk, meeting one of the following conditions: the probability of failure is high, the scope of failure impact is local, the severity of failure consequence is minor, and 0.6≤Rv;

[0051] The probability of failure is medium, the scope of failure impact is subsystem impact, the severity of failure consequence is medium, and 0.6≤Rv;

[0052] The probability of failure is low, the scope of failure impact is system-wide impact, the severity of failure consequence is severe, and 0.6≤Rv;

[0053] Very high risk, meeting one of the following conditions: the probability of failure is high, the scope of failure impact is subsystem impact, the severity of failure consequence is severe, and Rv ≥ 1.0;

[0054] The probability of failure is medium, the scope of failure impact is system-wide impact, the severity of failure consequence is severe, and Rv≥1.0;

[0055] In any case, the severity of the consequences of the failure is catastrophic.

[0056] Preferably, S5 includes the following steps:

[0057] S501: Anomaly detection: develop an anomaly detection model based on immune algorithm, adopt negative selection algorithm, generate a set of detectors representing self-mode, which is used to identify non-self-mode that is different from self-mode, establish abnormal thresholds for each fault mode, and generate detectors corresponding to the mean, variance and change trend of historical normal operation data of the gasification unit. When the matching degree between the actual data of the gasification unit and the detector is lower than the set threshold, it is judged as abnormal, and the working state of the gasification unit is predicted according to the abnormal signal;

[0058] S502: Use immune network theory to build a dynamic monitoring network, so that different detectors form a network relationship and dynamically adjust the detection sensitivity through mutual stimulation and inhibition. When multiple detectors in a local area send out abnormal signals at the same time, the interaction between them can enhance the overall prediction signal strength, which is used to avoid ignoring potential safety hazards due to minor false alarms of a single detector;

[0059] S503: Predict signal propagation. After an abnormality is detected, the prediction signal is quickly transmitted to relevant functional units and operators according to the scope of fault impact;

[0060] S504: Establish a self-repair strategy library, and formulate corresponding repair strategies for different types of faults based on the prediction signals. The strategies include equipment redundancy switching, automatic parameter adjustment, intelligent control strategy adjustment, and starting emergency repair equipment. Comprehensively consider the repair time, repair cost, and factors affecting the overall operation of the device to select the most appropriate repair plan.

[0061] Preferably, the S502 includes the following steps:

[0062] S50201: Treat each detector as a node in the network and establish a connection relationship between them;

[0063] S50202: Setting stimulation and inhibition rules between detectors. When a detector detects a slight abnormal signal, it will send stimulation signals to other detectors connected to it. When at least two detectors detect abnormal signals at the same time and the signal strength has not exceeded the set threshold, inhibition occurs to avoid overreaction.

[0064] S50203: Use sensors to collect the operating data of the gasification device in real time and input the data into each detector. The detector determines whether the data is abnormal based on its own set threshold and the stimulation and inhibition signals received from other detectors, and generates corresponding monitoring signals. The abnormal monitoring signals are transmitted and integrated in the dynamic monitoring network according to their network relationships.

[0065] Compared with the prior art, the present invention provides a safety analysis and prediction method for a coal gasification device based on functional analysis, which has the following beneficial effects:

[0066] 1. The safety analysis and prediction method of the coal gasification unit based on functional analysis can regard the coal gasification unit as a whole through the constructed functional network topology diagram, rather than looking at individual equipment and subsystems separately. It can help identify the complex relationships between the various functional units in the unit, including the interaction of material flow, energy flow and information flow. When analyzing the fault propagation path, it can be clearly seen how a local fault affects other parts of the entire unit through these interactive relationships, thereby providing a comprehensive risk perspective to avoid missing potential safety hazards.

[0067] 2. This coal gasification unit safety analysis and prediction method based on functional analysis does not need to analyze the impact of the fault from scratch when a fault occurs. Instead, it can draw on historical data of similar faults in the past. For common unit faults, by referring to the scope and degree of the impact of the fault on other equipment and process parameters in historical data, it can quickly and preliminarily estimate the severity of the current fault, thereby buying time for subsequent detailed analysis and formulation of countermeasures.

[0068] 3. The safety analysis and prediction method of the coal gasification device based on functional analysis introduces the principle of the biological immune system for safety prediction, and compares the coal gasification device to a living organism with immune function. It can quickly discover potential safety hazards just like the biological immune system recognizes antigens. Once a fault is detected, it can draw on the self-repair process of biological tissues and use pre-designed redundant structures and spare equipment to repair and replace faulty components and abnormal functions. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0070] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0071] like Figure 1 As shown in the figure, the safety analysis and prediction method of coal gasification unit based on functional analysis includes the following steps:

[0072] S1: Build a digital function list and a three-dimensional digital model, disassemble the coal gasification unit into different subsystems according to the process flow, define and number the functions of each subsystem, and build a digital function list;

[0073] The high-precision three-dimensional digital model of the coal gasification unit is constructed based on 3D laser scanning and digital modeling, including the following steps:

[0074] S101: Collect detailed process flow of coal gasification unit, including coal feeding, pretreatment, gasification reaction, synthesis gas purification and cooling, ash treatment;

[0075] S102: Determine the principle of functional unit division. According to the physical structure and process function of the coal gasification device, determine the principle of functional unit division, including division according to equipment type and process stage. Based on the independence and integrity of the functional units, each functional unit can only independently complete a specific function, and its input and output are clearly defined. For equipment and process steps that need to work together, they are combined into a functional unit. For example, the gasifier and its supporting feeding system and combustion control system can be divided into a gasification reaction functional unit.

[0076] S103: Preliminary division of functional units: According to the determined division principle, the coal gasification device is preliminarily divided into functional units, starting from coal feeding to ash treatment, and each link is gradually disassembled into functional units. The coal feeding is divided into coal bunker, feeder, and conveyor belt; the pretreatment part can be divided into crushing unit, screening unit, and drying unit. Each functional unit is numbered and named using a hierarchical coding method;

[0077] S104: Functional definition. For each divided functional unit, its main function is clearly defined, including the role of the functional unit in the entire coal gasification device, and its processing methods for materials, energy and information. For the crushing unit, its main function is to crush large pieces of coal into particle sizes that meet the gasification requirements. Its input is raw coal of different particle sizes, and its output is crushed coal with a smaller particle size. Describe the function from the perspectives of physical process and chemical process. For functional units involving chemical reactions, explain in detail the type of reaction, reactants, products, and reaction conditions. For physical processes, describe the transportation method and flow control method.

[0078] S105: Determine the input parameters of each functional unit, including the nature, flow, temperature, pressure of the material, and the input form and control signal of energy. For the feeder functional unit, the input parameters include the coal level signal of the coal bunker, the set feeding speed signal, and the power supply. Similarly, clarify the output parameters of the functional unit. The output parameters should correspond to the input parameters of the next functional unit, and can also correspond to the relevant parameters of the final product and waste of the device;

[0079] S106: Functional details and characteristic description, describing the working characteristics of the functional unit, including but not limited to the working cycle, operation mode, start and stop conditions. Some pumps may run continuously, while some mixing equipment may run intermittently. Their start and stop may be related to the liquid level and reaction time of the material. The special requirements and restrictions of the functional unit shall be recorded. When some equipment may have requirements for ambient temperature and humidity, and have certain tolerance limits for the corrosiveness and abrasiveness of the material, for functional units involving flammable and explosive gases, special notes shall be given on explosion-proof requirements and safety measures;

[0080] S107: Integration and recording of digital function lists: Use spreadsheet software to create a digital function list table, the column headings of which include but are not limited to the function unit number, name, subsystem, main function, input parameter, output parameter, working characteristics, and special requirements. In the digital function list, set the association relationship between the function units;

[0081] S2: Construct the functional network topology of the coal gasification unit. Based on the digital function list and the three-dimensional digital model, the functional units are used as nodes, and the material flow, energy flow and information flow between functions are used as connecting edges. The structural characteristics of the functional network topology are analyzed and dynamically simulated. The specific steps include the following:

[0082] S201: Based on the system dynamics simulation software, combined with the actual operation data of the coal gasification unit and the process mechanism model, the functional network topology diagram is dynamically simulated to simulate the state changes and mutual influence relationships of each node in the functional network topology diagram under different working conditions, and the functional state dynamic transfer matrix is ​​constructed;

[0083] S202: In the functional network topology diagram, key nodes and vulnerable links are determined through node degree distribution and clustering coefficient. Key nodes are core function executors, high-flow and high-energy transmission points, and control hubs in the coal gasification device; vulnerable links are high-risk material transmission links, connection components under complex working conditions, and backup links that rely on a single device in the coal gasification device. The gasification reactor is the core node of the entire coal gasification device, and it connects many material flows and energy flows. When the node fails, it is confirmed that it will have an impact on the entire system. Analyze the vulnerable links. When the pipeline connecting the gasification reactor and the synthesis gas cooling system is the key path for energy and material transmission, when the pipeline is blocked or leaking, the synthesis gas cannot be cooled normally, which will lead to a series of subsequent safety issues;

[0084] S203: Use the function state dynamic transfer matrix to observe the state changes of key nodes and vulnerable links under different working conditions. When the coal type changes and the gasification reaction rate changes, the simulation can show how the temperature and pressure parameters in the gasification reactor change, and how these changes propagate to other subsystems through the vulnerable links.

[0085] S204: Fault propagation path analysis. According to the dynamic simulation results of the functional network, the fault propagation path is sorted out. Taking the local overheating fault in the gasifier of the coal gasification device as an example, overheating will cause damage to the furnace wall material, and then the synthesis gas will leak into the surrounding environment. The leaked synthesis gas will come into contact with the surrounding high-temperature equipment and fire sources, causing fires and explosions; consider different types of faults, including but not limited to mechanical faults, chemical faults, and control faults in the functional network. Mechanical faults will cause damage to equipment parts and affect material transportation, thereby propagating along the material flow path; chemical faults will change the reaction process and affect other related functional units through energy flow and information flow;

[0086] S3: Analysis and diagnosis. By integrating the real-time sensor data, historical operation data, equipment maintenance records, inspection reports and external environment data of the coal gasification unit, a comprehensive data collection platform is built. The convolutional neural network-long short-term memory network fusion architecture is used to analyze the real-time data and the fault mode. When the temperature sensor data fluctuates abnormally, and the historical operation data shows that similar fluctuations have been related to equipment aging, the model can comprehensively judge that the temperature control failure may be caused by equipment aging. For complex faults, in-depth diagnosis is performed by integrating data from different sources. When the composition of the syngas is abnormal, the real-time composition detection data, coal quality data, and equipment operation parameters are combined to analyze whether the fault is caused by coal quality changes, equipment failure, or improper operating parameters.

[0087] According to the process requirements and safety monitoring requirements of the coal gasification unit, appropriate types of sensors are selected, including temperature sensors, pressure sensors, and gas concentration sensors, and the collected data are transmitted through wireless transmission. For temperature sensors, high-temperature resistant thermal resistance sensors with an accuracy of ±1°C are selected in the high-temperature area inside the coal gasification unit, and ordinary temperature sensors are selected in the normal temperature and low temperature areas of the coal gasification unit; for pressure sensors, high-precision pressure sensors with an accuracy of ±0.1%FS are selected in the high-pressure pipelines and gasifiers in the coal gasification unit; for gas concentration sensors, the detection lower limit of the gas concentration sensor reaches the ppm level, and the response time is less than 30 seconds. The construction of the convolutional neural network-long short-term memory network fusion architecture includes the following steps:

[0088] S301: Clean and normalize the collected sensor data, and design the CCN module based on it, including the design of the convolution layer and the pooling layer, and input the designed feature data into the LSTM layer in time sequence;

[0089] S302: A fully connected layer is connected after the LSTM layer as the output layer. The number of neurons in the output layer is the number of failures in the coal gasification unit. The Softmax activation function is used to normalize the results of the output layer so that the output results are represented as the probability distribution of each failure mode.

[0090] S303: Train and optimize the convolutional neural network-long short-term memory network fusion architecture;

[0091] S4: Conduct risk assessment on the analyzed failure modes and formulate corresponding risk response strategies and suggestions according to the different levels of assessment, including the following steps:

[0092] S401: Based on the functional analysis and historical fault data of the coal gasification unit, a fault mode library is established. The fault modes include equipment failure, process failure, and external interference failure. Each fault mode is described in detail, including the manifestation characteristics of the fault, the cause of the fault, and the location of the fault.

[0093] S402: Determine the risk quantification index and the risk value. The formula is:

[0094] R k =P k +S k +C k ;

[0095] Where k is the failure mode, P k is the probability of occurrence of fault k, S k is the influence range coefficient of the failure mode, C k is the consequence severity factor of the failure mode, Rv kis the risk value of fault k, and the impact range of the fault mode is obtained based on the constructed coal gasification unit functional network topology diagram;

[0096] S403: Calculate the total risk value based on the risk value, the formula is:

[0097] TRv=∑ k R k ;

[0098] Where TRv is the total risk value;

[0099] S404: Fault probability calculation, calculating the probability of fault occurrence based on the Bayesian network method;

[0100] S405: Output of risk assessment results. According to the calculation results of risk quantification indicators, a risk assessment report of the coal gasification unit is output. The report includes the risk value of each functional unit, the main failure mode and its probability of occurrence, and the total risk value of the entire unit. The risk assessment results are graded based on the grading factors, where the grading factors include: failure probability, failure impact range, and failure consequence severity. The failure probability is recorded as P. Based on the calculation results of the failure probability, it is divided into extremely low probability: P < 0.01; low probability: 0.01 ≤ P < 0.1; medium probability: 0.1 ≤ P < 0.5; high probability: 0.5 ≤ P;

[0101] Based on the constructed coal gasification unit functional network topology, the fault impact range is obtained, which is denoted as S and is divided into local impact, where the fault only affects a single non-critical functional unit: S < 0.2; subsystem impact, where the fault affects multiple functional units within a subsystem: 0.2 ≤ S < 0.6; system-wide impact, where the fault affects multiple subsystems of the entire coal gasification unit: 0.6 ≤ S;

[0102] The severity of the consequences of the failure is recorded as C, which is divided into minor consequences. The failure has a short-term and quickly recoverable impact on production, and basically no harm to safety and the environment: C = 1; medium consequences: The failure causes production interruption and requires certain repairs and adjustments to recover, which has potential risks to safety and the environment,

[0103] C = 3; Severe consequences: The failure causes long-term production stagnation, serious damage to equipment, and great harm to safety and the environment: C = 5; Catastrophic consequences: The failure causes an irreversible disaster:

[0104] C = 7;

[0105] The definition of risk classification standards based on risk factors includes low risk, which meets one of the following conditions: the probability of failure is extremely low, and the scope of failure impact is local impact, and the severity of failure consequences is minor consequences;

[0106] The probability of failure is low, the scope of failure is local, and the severity of failure consequence is minor, and Rv = P × S × C < 0.2;

[0107] For low risk: Maintain regular monitoring frequency, use the existing sensors of the device to collect status, the collection time interval is 15-30 minutes, record the monitoring data in detail, establish data archives for subsequent trend analysis, and observe whether there are small fluctuations by recording the coal feed flow data for a long time, so as to find potential problems early, and perform maintenance according to the normal equipment maintenance plan, including regular equipment inspections, component lubrication, and cleaning;

[0108] Medium risk: one of the following conditions is met: the probability of failure is medium, the scope of failure impact is local, the severity of failure consequence is minor, and 0.2≤Rv<0.6;

[0109] The probability of failure is low, the scope of failure impact is subsystem impact, the severity of failure consequence is medium, and 0.2≤Rv<0.6;

[0110] For medium risks: Increase the monitoring frequency, shorten the collection time interval to 10-15 minutes, increase vibration monitoring and noise monitoring of key equipment to obtain more information about the status of the device, establish a real-time data monitoring platform, use data visualization technology to display and analyze the monitoring data in real time, and issue an alarm in time to notify relevant personnel once abnormal data is found. For failure risks that affect subsystems, conduct comprehensive functional testing and performance evaluation of the entire subsystem, including checking whether the pipeline connection is tight and whether the valve is opened and closed normally. According to the evaluation results, formulate targeted maintenance and improvement plans, including local repair and upgrade of equipment; optimize the process flow;

[0111] High risk, meeting one of the following conditions: the probability of failure is high, the scope of failure impact is local, the severity of failure consequence is minor, and 0.6≤Rv;

[0112] The probability of failure is medium, the scope of failure impact is subsystem impact, the severity of failure consequence is medium, and 0.6≤Rv;

[0113] The probability of failure is low, the scope of failure impact is system-wide impact, the severity of failure consequence is severe, and 0.6≤Rv;

[0114] For high risks: enter the emergency monitoring state, increase the monitoring frequency of key parameters to once every 5-10 minutes, comprehensively consider real-time monitoring data, historical fault data and expert experience, conduct a comprehensive assessment of the safety status of the gasification unit, immediately activate the emergency plan, clarify the responsibilities of each department and personnel in emergency situations, establish an emergency command team, including the production department, maintenance department, and safety management department, responsible for coordinating emergency handling work, and take temporary risk control measures, including reducing the operating load of the gasification unit to reduce the working pressure of the equipment and reduce the possibility of failure. Isolate and back up the fault source that may cause serious consequences. When a key pump has a high-risk failure, activate the backup pump, and urgently repair and replace the faulty pump. Organize experts to conduct in-depth analysis of high-risk failures and formulate long-term risk reduction strategies;

[0115] Very high risk, meeting one of the following conditions: the probability of failure is high, the scope of failure impact is subsystem impact, the severity of failure consequence is severe, and Rv ≥ 1.0;

[0116] The probability of failure is medium, the scope of failure impact is system-wide impact, the severity of failure consequence is severe, and Rv≥1.0;

[0117] In any case, the severity of the consequences of the failure is catastrophic;

[0118] In case of extremely high risk, emergency shutdown measures should be taken immediately to stop the operation of the coal gasification unit. During the shutdown process, it is necessary to ensure that the operation is carried out in accordance with the safety operating procedures to prevent secondary disasters caused by the emergency shutdown. A comprehensive crisis management procedure should be initiated, including notifying the superior competent department, surrounding communities and relevant emergency rescue agencies, organizing internal emergency rescue teams, and taking emergency measures for possible casualties and environmental pollution. An in-depth root cause analysis should be conducted on the causes of extremely high risk. The root cause of the failure should be determined through the functional network topology diagram of the coal gasification unit. Based on the results of the root cause analysis, comprehensive rectification measures should be formulated, including redesigning and renovating the entire coal gasification unit, replacing equipment that does not meet safety standards, and strengthening safety training and emergency drills for personnel, so as to ensure that the unit can operate safely and stably after restarting and completely eliminate extremely high risk hazards.

[0119] S5: Based on the principle of biological immune system, safety prediction is performed. The normal operating state parameters of the coal gasification device are defined as self-mode, and parameter changes, abnormal fluctuations and equipment failure signals beyond the normal range in the coal gasification device are regarded as non-self-mode. Safety prediction is performed on the detected non-self-mode, including the following steps:

[0120] S501: Anomaly detection: develop an anomaly detection model based on immune algorithm, adopt negative selection algorithm, generate a set of detectors representing self-mode, which is used to identify non-self-mode that is different from self-mode, establish abnormal thresholds for each fault mode, and generate corresponding detectors according to the mean, variance and change trend of historical normal operation data of the gasification unit. When the matching degree between the actual data of the gasification unit and the detector is lower than the set threshold, it is judged as abnormal. Predict the working state of the gasification unit according to the abnormal signal. For temperature data, generate temperature detectors according to the mean, variance and change trend of historical normal operation temperature. When the matching degree between the actual temperature data and the detector is lower than the set threshold, it is judged as abnormal.

[0121] S502: Use immune network theory to build a dynamic monitoring network, so that different detectors form a network relationship, and dynamically adjust the detection sensitivity through mutual stimulation and inhibition. When multiple detectors in a local area send out abnormal signals at the same time, the interaction between them can enhance the overall prediction signal strength, which is used to avoid ignoring potential safety hazards due to minor false alarms of a single detector. The operation steps include the following:

[0122] S50201: Treat each detector as a node in the network and establish connections between them. For example, there is a connection between the detectors that monitor the temperature and pressure of the gasifier in the coal gasification device, because temperature changes often affect pressure changes, and vice versa;

[0123] S50202: Set stimulation and inhibition rules between detectors. When a detector detects a slight abnormal signal, it will send stimulation signals to other detectors connected to it. When a detector monitoring the temperature of a certain area finds that the temperature has a slight upward trend and has not exceeded the set threshold, it sends stimulation signals to related pressure and flow detectors to make them lower their own abnormal detection thresholds and enter the prejudgment state in advance. When at least two detectors detect abnormal signals at the same time and the signal strength has not exceeded the set threshold, an inhibition effect is generated to avoid overreaction. At least two temperature detectors in adjacent areas detect abnormal high temperature signals, and they inhibit each other and adjust their respective signal strengths to prevent misjudgment of the entire network due to local hot spots, so that the prejudgment signal can better reflect the overall safety status of the device.

[0124] S50203: Use sensors to collect the operation data of the coal gasification device in real time, and input the data into each detector. The detector determines whether the data is abnormal based on its own threshold value and the stimulation and inhibition signals received from other detectors, and generates corresponding monitoring signals. The abnormal monitoring signals are transmitted and integrated in the dynamic monitoring network according to their network relationships;

[0125] S503: Predict signal propagation. After an abnormality is detected, the prediction signal is quickly transmitted to relevant functional units and operators according to the scope of fault impact;

[0126] S504: Establish a self-repair strategy library. Based on the prediction signal, formulate corresponding repair strategies for different types of faults. The strategies include equipment redundancy switching, such as automatically switching to the backup pump when the main pump fails, automatic parameter adjustment, such as automatically increasing the cooling medium flow rate according to the abnormal temperature increase, intelligent control strategy adjustment, such as changing the reaction control algorithm to adapt to the abnormalities caused by raw material changes, and starting emergency repair equipment, such as dispatching robots to repair pipeline leaks and replace faulty valves. Comprehensively consider the repair time, repair cost, and factors affecting the overall operation of the device to select the most appropriate repair plan.

[0127] It should be noted that the present invention is a safety analysis and prediction method for a coal gasification device based on functional analysis, when used.

[0128] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A safety analysis and prediction method for a coal gasification unit based on functional analysis, characterized by: The steps include: S1: Build a digital function list and a three-dimensional digital model, disassemble the coal gasification unit into different subsystems according to the process flow, define and number the functions of each subsystem, and build a digital function list; Construct a high-precision three-dimensional digital model of the coal gasification unit based on 3D laser scanning and digital modeling; S2: Construct the functional network topology of the coal gasification unit. Based on the digital function list and the three-dimensional digital model, the functional units are used as nodes, and the material flow, energy flow and information flow between functions are used as connecting edges. The structural characteristics of the functional network topology are analyzed and dynamically simulated. S3: Analysis and diagnosis: By integrating the real-time sensor data, historical operation data, equipment maintenance records, inspection reports and external environmental data of the coal gasification unit, a comprehensive data collection platform is built. The convolutional neural network-long short-term memory network fusion architecture is used to analyze the real-time data and the fault mode. S4: Conduct risk assessment on the analyzed failure modes and formulate corresponding risk response strategies and suggestions according to different assessment levels; S5: Safety prediction is performed based on the principle of biological immune system. The normal operating status parameters of the coal gasification device are defined as self-mode, and parameter changes, abnormal fluctuations and equipment failure signals beyond the normal range in the coal gasification device are regarded as non-self-mode, and safety prediction is performed on the detected non-self-mode.

2. The method for safety analysis and prediction of a coal gasification device based on functional analysis according to claim 1 is characterized in that: The S1 includes the following steps: S101: Collect detailed process flow of coal gasification unit, including coal feeding, pretreatment, gasification reaction, synthesis gas purification and cooling, ash treatment; S102: Determine the principle of functional unit division. According to the physical structure and process function of the coal gasification unit, determine the principle of functional unit division, including division according to equipment type and process stage. Based on the independence and integrity of the functional units, each functional unit can only independently complete a specific function, and its input and output are clearly defined. For equipment and process steps that need to work together, they are combined into one functional unit; S103: Preliminary division of functional units: According to the determined division principle, the coal gasification device is preliminarily divided into functional units, starting from coal feeding to ash treatment, and each link is gradually disassembled into functional units, and each functional unit is numbered and named using a hierarchical coding method; S104: Functional definition: for each divided functional unit, clarify its main function, including the role of the functional unit in the entire coal gasification unit, and its processing methods for materials, energy and information; S105: Determine the input parameters of each functional unit, including the properties, flow, temperature, pressure of the material, and the input form of energy and control signals; S106: Functional details and feature description, describing the working characteristics of the functional unit, including but not limited to working cycle, operation mode, start and stop conditions; S107: Integration and recording of digital function lists. Use spreadsheet software to create a digital function list table. The column headings of the table include but are not limited to the function unit number, name, subsystem, main function, input parameters, output parameters, working characteristics, and special requirements. In the digital function list, set the association relationship between the function units.

3. The method for safety analysis and prediction of a coal gasification device based on functional analysis according to claim 1 is characterized in that: The S2 specifically includes the following steps: S201: Based on the system dynamics simulation software, combined with the actual operation data of the coal gasification unit and the process mechanism model, the functional network topology diagram is dynamically simulated to simulate the state changes and mutual influence relationships of each node in the functional network topology diagram under different working conditions, and the functional state dynamic transfer matrix is ​​constructed; S202: In the functional network topology diagram, key nodes and vulnerable links are determined through node degree distribution and clustering coefficient. Key nodes are core function executors, high flow and high energy transmission points, and control hubs in the coal gasification device; vulnerable links are high-risk material transmission links in the coal gasification device, connection components under complex working conditions, and backup links that rely on a single device; S203: using the function state dynamic transfer matrix, observing the state changes of key nodes and vulnerable links under different working conditions; S204: Fault propagation path analysis: sort out the fault propagation path according to the dynamic simulation results of the functional network.

4. The method for safety analysis and prediction of a coal gasification device based on functional analysis according to claim 1 is characterized in that: In S3, the construction of the convolutional neural network-long short-term memory network fusion architecture includes the following steps: S301: Clean and normalize the collected sensor data, and design the CCN module based on it, including the design of the convolution layer and the pooling layer, and input the designed feature data into the LSTM layer in chronological order; S302: A fully connected layer is connected after the LSTM layer as the output layer. The number of neurons in the output layer is the number of failures in the coal gasification unit. The Softmax activation function is used to normalize the results of the output layer so that the output results are expressed as the probability distribution of each failure mode. S303: Train and optimize the convolutional neural network-long short-term memory network fusion architecture.

5. The method for safety analysis and prediction of a coal gasification device based on functional analysis according to claim 1 is characterized in that: The S4 includes the following steps: S401: Based on the functional analysis and historical fault data of the coal gasification unit, a fault mode library is established. The fault modes include equipment failure, process failure, and external interference failure. Each fault mode is described in detail, including the manifestation characteristics of the fault, the cause of the fault, and the location of the fault. S402: Determine the risk quantification index and the risk value. The formula is: Rv k =P k +S k +C k ; Where k is the failure mode, P k is the probability of occurrence of fault k, S k is the influence range coefficient of the failure mode, C k is the consequence severity factor of the failure mode, Rv k is the risk value of fault k, and the impact range of the fault mode is obtained based on the constructed coal gasification unit functional network topology diagram; S403: Calculate the total risk value based on the risk value, the formula is: TRv=∑ k Rv k ; Where TRv is the total risk value; S404: Fault probability calculation, calculating the probability of fault occurrence based on the Bayesian network method; S405: Output of risk assessment results. Based on the calculation results of risk quantification indicators, a risk assessment report of the coal gasification unit is output. The report includes the risk value of each functional unit, the main failure mode and its probability of occurrence, and the total risk value of the entire unit, and the risk assessment results are graded.

6. The method for safety analysis and prediction of a coal gasification device based on functional analysis according to claim 5 is characterized in that: In S405, the classification is based on classification factors, wherein the classification factors include: probability of failure, scope of failure impact, and severity of failure consequences, wherein the probability of failure is recorded as P, and based on the result of failure probability calculation, it is divided into extremely low probability: P < 0.01; low probability: 0.01 ≤ P < 0.1; medium probability: 0.1 ≤ P < 0.5; high probability: 0.5 ≤ P; Based on the constructed coal gasification unit functional network topology, the fault impact range is obtained, which is denoted as S and is divided into local impact, where the fault only affects a single non-critical functional unit: S < 0.2; subsystem impact, where the fault affects multiple functional units within a subsystem: 0.2 ≤ S < 0.6; system-wide impact, where the fault affects multiple subsystems of the entire coal gasification unit: 0.6 ≤ S; The severity of the consequences of the failure is recorded as C, which is divided into minor consequences. The failure has a short-term and quickly recoverable impact on production, and basically no harm to safety and the environment: C=1; medium consequences: the failure causes production interruption, which requires certain repairs and adjustments to recover, and has potential risks to safety and the environment, C=3; serious consequences: the failure causes long-term production stagnation, serious damage to equipment, and great harm to safety and the environment: C=5; catastrophic consequences: the failure causes an irreversible disaster: C=7.

7. The method for safety analysis and prediction of a coal gasification device based on functional analysis according to claim 6 is characterized in that: The definition of risk classification standards based on the risk factors includes low risk, which meets one of the following conditions: the probability of failure is extremely low, the scope of failure impact is local impact, and the severity of failure consequence is minor consequence; The probability of failure is low, the scope of failure is local, and the severity of failure consequence is minor, and Rv = P × S × C < 0.2; Medium risk: one of the following conditions is met: the probability of failure is medium, the scope of failure impact is local, the severity of failure consequence is minor, and 0.2≤Rv<0.6; The probability of failure is low, the scope of failure impact is subsystem impact, the severity of failure consequence is medium, and 0.2≤Rv<0.6; High risk, meeting one of the following conditions: the probability of failure is high, the scope of failure impact is local, the severity of failure consequence is minor, and 0.6≤Rv; The probability of failure is medium, the scope of failure impact is subsystem impact, the severity of failure consequence is medium, and 0.6≤Rv; The probability of failure is low, the scope of failure impact is system-wide impact, the severity of failure consequence is severe, and 0.6≤Rv; Very high risk, meeting one of the following conditions: the probability of failure is high, the scope of failure impact is subsystem impact, the severity of failure consequence is severe, and Rv ≥ 1.0; The probability of failure is medium, the scope of failure impact is system-wide impact, the severity of failure consequence is severe, and Rv≥1.0; In any case, the severity of the consequences of the failure is catastrophic.

8. The method for safety analysis and prediction of a coal gasification device based on functional analysis according to claim 1 is characterized in that: The S5 comprises the following steps: S501: Anomaly detection: develop an anomaly detection model based on immune algorithm, adopt negative selection algorithm, generate a set of detectors representing self-mode, which is used to identify non-self-mode that is different from self-mode, establish abnormal thresholds for each fault mode, and generate detectors corresponding to the mean, variance and change trend of historical normal operation data of the gasification unit. When the matching degree between the actual data of the gasification unit and the detector is lower than the set threshold, it is judged as abnormal, and the working state of the gasification unit is predicted according to the abnormal signal; S502: Use immune network theory to build a dynamic monitoring network, so that different detectors form a network relationship and dynamically adjust the detection sensitivity through mutual stimulation and inhibition. When multiple detectors in a local area send out abnormal signals at the same time, the interaction between them can enhance the overall prediction signal strength, which is used to avoid ignoring potential safety hazards due to minor false alarms of a single detector; S503: Predict signal propagation. After an abnormality is detected, the prediction signal is quickly transmitted to relevant functional units and operators according to the scope of fault impact; S504: Establish a self-repair strategy library, and formulate corresponding repair strategies for different types of faults based on the prediction signals. The strategies include equipment redundancy switching, automatic parameter adjustment, intelligent control strategy adjustment, and starting emergency repair equipment. Comprehensively consider the repair time, repair cost, and factors affecting the overall operation of the device to select the most appropriate repair plan.

9. The method for safety analysis and prediction of a coal gasification device based on functional analysis according to claim 8 is characterized in that: The S502 includes the following steps: S50201: Treat each detector as a node in the network and establish a connection relationship between them; S50202: Setting stimulation and inhibition rules between detectors. When a detector detects a slight abnormal signal, it will send stimulation signals to other detectors connected to it. When at least two detectors detect abnormal signals at the same time and the signal strength has not exceeded the set threshold, inhibition occurs to avoid overreaction. S50203: Use sensors to collect the operating data of the gasification device in real time and input the data into each detector. The detector determines whether the data is abnormal based on its own set threshold and the stimulation and inhibition signals received from other detectors, and generates corresponding monitoring signals. The abnormal monitoring signals are transmitted and integrated in the dynamic monitoring network according to their network relationships.

Images