Risk assessment system and method based on artificial intelligence

Abstract

The invention discloses a risk assessment system and method based on artificial intelligence, relates to the technical field of risk assessment, and solves the technical problems that in the prior art, the requirements for skill and experience of personnel are high, the analysis process is tedious and easy to make mistakes, and combination with other analysis methods is difficult. The method comprises the following steps: acquiring process data and equipment data of a methanol synthesis reactor; analyzing the process data based on a pre-trained HAZOP analysis and identification model to obtain an identification result; calculating an aging evaluation coefficient based on the equipment data; analyzing the identification result to obtain the risk probability of the methanol synthesis reactor and the failure probability of the suggested measures; determining the priority of the suggested measures based on the failure probability of the suggested measures; and calculating a risk assessment coefficient based on the risk probability and the aging assessment coefficient, and judging whether to carry out early warning based on the risk assessment coefficient, thereby solving the technical problem.

Description

Technical Field

[0001] The present invention belongs to the field of risk assessment, and specifically relates to a risk assessment system and method based on artificial intelligence. Background Art

[0002] Methanol is an important chemical raw material, widely used in many fields such as chemical industry, pharmaceutical industry and energy. As the core equipment for producing methanol, the operation safety and production efficiency of methanol synthesis reactor are crucial. Methanol synthesis involves many complex process flows, including raw gas supply and pretreatment, heating, reaction, cooling and separation, purification and storage, and waste gas treatment. The parameter control of each link (such as temperature, pressure, flow, etc.) has a direct impact on the safety and efficiency of the overall process.

[0003] Although the existing technology can systematically identify and evaluate the potential risks in methanol synthesis reactors through the HAZOP (Hazard and Operability Analysis) method, the HAZOP (Hazard and Operability Analysis) analysis method faces many challenges in practical application, such as high requirements on personnel skills and experience, cumbersome and error-prone analysis process, and difficulty in combining with other analysis methods.

[0004] Therefore, the present invention solves the above-mentioned problems by proposing a risk assessment system and method based on artificial intelligence. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a risk assessment system and method based on artificial intelligence, which are used to solve the technical problems that although the prior art can systematically identify and evaluate the potential risks in the methanol synthesis reactor through the HAZOP method, the HAZOP analysis method faces many challenges in practical application, such as high requirements on personnel skills and experience, cumbersome and error-prone analysis process, and difficulty in combining with other analysis methods.

[0006] To achieve the above-mentioned object, a first aspect of the present invention provides a risk assessment system based on artificial intelligence, comprising: a data acquisition module, a data processing module and a risk assessment module;

[0007] The data acquisition module is used to obtain process data and equipment data of the methanol synthesis reactor;

[0008] The data processing module: analyzes the process data based on the pre-trained HAZOP analysis and recognition model to obtain the recognition result; calculates the aging assessment coefficient based on the equipment data; and,

[0009] By analyzing the identification results, the risk probability of the methanol synthesis reactor and the failure probability of the recommended measures are obtained;

[0010] The risk assessment module determines the priority of the recommended measures based on the failure probability of the recommended measures; calculates the risk assessment coefficient based on the risk probability and the aging assessment coefficient, and determines whether to issue an early warning based on the risk assessment coefficient.

[0011] Preferably, the obtaining of process data and equipment data of the methanol synthesis reactor includes:

[0012] The process of the methanol synthesis reactor is divided into several key nodes according to the process flow; wherein the several key nodes include: raw gas supply and pretreatment process, raw gas heating process, methanol synthesis reaction process, product cooling and separation process, methanol purification and storage process, and waste gas treatment process;

[0013] Obtaining process data at each key node; wherein the process data includes: process parameters and equipment parameters;

[0014] The equipment data of the compressor in the methanol synthesis reactor is collected by a data acquisition device; wherein the equipment data includes: vibration frequency and output power.

[0015] It should be noted that the process parameters include temperature, pressure, flow, humidity, sulfur content, heating rate, exhaust gas concentration, etc.; the equipment parameters include compressor speed, compressor power, filter or adsorber working efficiency, desulfurization tower operating temperature and pressure, cooling medium flow and temperature, condenser and cooler refrigerant flow and temperature, etc.

[0016] Preferably, the analysis of process data based on the pre-trained HAZOP analysis and recognition model includes:

[0017] Extract historical process data of methanol synthesis reactor from the database, and the corresponding analysis results obtained through HAZOP analysis; the analysis results include: initial events, deviation causes, accident consequences, existing safety measures and recommended measures corresponding to the initial events;

[0018] Integrate historical process data into standard input data and analysis results into standard output data; train an artificial intelligence model based on the standard input data and standard output data to obtain a pre-trained HAZOP analysis and recognition model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network;

[0019] The process data acquired in real time is input into the HAZOP analysis and recognition model to obtain a recognition result; wherein the content attribute of the recognition result is consistent with the content attribute of the analysis result.

[0020] It should be noted that the database refers to the existing data set of historical HAZOP analysis reports on methanol synthesis reactors;

[0021] Deviation causes refer to the specific reasons that cause process parameters to deviate from the normal range, such as equipment aging, operating errors, changes in the external environment, etc.

[0022] Accident consequences refer to the most serious consequences that may be caused by deviations, such as production interruption, equipment damage, personal injury, etc.

[0023] Existing security measures refer to the existing security protection measures in the current system and their effectiveness evaluation;

[0024] Recommended measures refer to improvement suggestions put forward to reduce risks, such as optimizing operating procedures, upgrading equipment, adding redundant systems, etc.

[0025] Preferably, the aging assessment coefficient is calculated based on the equipment data, including:

[0026] The vibration frequency is marked as Z and the output power is marked as W;

[0027] The aging assessment coefficient is calculated by the formula: LP = A×e^[Z / (Z+1)]+B×e^[(W-ZW)^2 / ZW^2]; where LP is the aging assessment coefficient, A and B are aging influencing factors, and 0<A<1, 0<B<1, A<B, and ZW is the optimal output power.

[0028] It should be noted that the aging impact factor is set by experts in this field based on the historical operating data of the compressor in the methanol synthesis reactor;

[0029] Vibration frequency and output power have a significant impact on the aging of the compressor in the methanol synthesis reactor; high vibration frequency usually means increased friction and wear between mechanical parts, resulting in increased fatigue damage and premature failure of seals, which accelerates the aging process of the compressor; low vibration frequency means smoother operation, reduced mechanical wear and fatigue damage, which helps to extend equipment life and reduce the risk of leakage; similarly, high output power will increase the heat load and mechanical stress of the compressor, leading to increased internal temperature, lubricant deterioration, and increased risk of component deformation or fracture, which in turn accelerates the aging process; too low output power will also cause the compressor to face a series of problems, such as poor lubrication, carbon deposits, seal failure and insufficient cooling, which will also significantly accelerate the aging process of the compressor;

[0030] The optimum output power refers to the power level when the compressor operates efficiently and stably under its design conditions. In this state, the performance indicators of the compressor (such as temperature, vibration, mechanical stress, etc.) are in the best balance, which can not only ensure production efficiency but also maximize the life of the equipment.

[0031] Preferably, the risk probability analysis step of the methanol synthesis reactor comprises:

[0032] Extract the initial events identified by the HAZOP analysis identification model and the existing safety measures corresponding to the initial events;

[0033] Based on the start-up operation data of the methanol synthesis reactor, the number of occurrences of the initial event within the preset time period T, and the number of triggering and failure times of the existing safety measures corresponding to the initial event within the preset time period T are obtained;

[0034] By formula: The probability of occurrence of each initial event is calculated; among them, CS i It refers to the probability of occurrence of the i-th initial event, i is the sequence number of the initial event, i = {1, 2, 3…, n}, n is the total number of initial events, X i It refers to the number of times the i-th initial event occurs within the preset time period T;

[0035] By formula The failure probability of the existing safety measures corresponding to the initial event is calculated; where S i,j It refers to the failure probability of the jth existing safety measure corresponding to the i-th initial event, j is the serial number of the existing safety measure, j = {1, 2, 3…, m}, m is the total number of existing safety measures, Y i,j G refers to the number of failures of the jth existing safety measure corresponding to the i-th initial event within the preset time period T. i,j It refers to the number of times the jth existing safety measure corresponding to the i-th initial event is triggered within the preset time period T;

[0036] The risk probability of the initial event is calculated based on the occurrence probability of the initial event and the failure probability of the existing safety measures corresponding to the initial event.

[0037] It should be noted that the initial event refers to the triggering event that may cause an accident or potential dangerous situation.

[0038] Preferably, the risk probability of the initial event is calculated based on the probability of occurrence of the initial event and the failure probability of the existing safety measures corresponding to the initial event, and includes:

[0039] By formula The total failure probability of existing safety measures corresponding to the initial event is calculated; where SX i It refers to the total failure probability of existing safety measures corresponding to the i-th initial event;

[0040] By calculating CS i and SX i The product of is used to obtain the risk probability of the initial event.

[0041] Preferably, determining the priority of the suggested measures based on the failure probability of the suggested measures includes:

[0042] Extract recommended actions corresponding to the initial event;

[0043] Obtain the failure probability of the recommended measures through the Delphi method;

[0044] The recommended measures are ranked in ascending order of failure probability; the higher the ranked recommended measures are, the higher the priority they are adopted.

[0045] Preferably, the risk assessment coefficient is calculated based on the risk probability and the aging assessment coefficient, including:

[0046] Label the risk probability as Fi;

[0047] The risk assessment coefficient is calculated by the formula: P = ɑ × ∑Fi + β × LP; where P is the risk assessment coefficient, ɑ and β are weight coefficients, and ɑ + β = 1, and Fi refers to the risk probability of the i-th initial event.

[0048] Preferably, the determining whether to issue an early warning based on the risk assessment coefficient includes:

[0049] Determine whether the risk assessment coefficient is greater than the preset risk assessment coefficient threshold; if yes, generate an early warning message and send it to the client; if not, continue to monitor and judge.

[0050] It should be noted that the preset risk assessment coefficient threshold is set by experts in this field based on experience.

[0051] A second aspect of the present invention provides a risk assessment method based on artificial intelligence, comprising:

[0052] Step 1: Obtain process data and equipment data of the methanol synthesis reactor;

[0053] Step 2: Analyze the process data based on the pre-trained HAZOP analysis and recognition model to obtain the recognition results; calculate the aging assessment coefficient based on the equipment data;

[0054] Step 3: By analyzing the identification results, the risk probability of the methanol synthesis reactor and the failure probability of the recommended measures are obtained;

[0055] Step 4: Determine the priority of the recommended measures based on the failure probability of the recommended measures; calculate the risk assessment coefficient based on the risk probability and the aging assessment coefficient, and determine whether to issue an early warning based on the risk assessment coefficient.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] 1. Although the prior art can systematically identify and evaluate the potential risks in a methanol synthesis reactor through the HAZOP method, the HAZOP analysis method faces many challenges in practical application, such as high requirements for personnel skills and experience, cumbersome and error-prone analysis process, and difficulty in combining with other analysis methods. The present invention obtains process data and equipment data of a methanol synthesis reactor; analyzes the process data based on a pre-trained HAZOP analysis recognition model to obtain an identification result; calculates an aging assessment coefficient based on the equipment data; analyzes the identification result to obtain the risk probability of the methanol synthesis reactor and the failure probability of the recommended measures; determines the priority of the recommended measures based on the failure probability of the recommended measures; calculates a risk assessment coefficient based on the risk probability and the aging assessment coefficient, and determines whether to issue an early warning based on the risk assessment coefficient, thereby solving the technical problems of the prior art that the prior art has high requirements for personnel skills and experience, cumbersome and error-prone analysis process, and difficulty in combining with other analysis methods.

[0058] 2. The present invention can accurately quantify the aging degree of the compressor by real-time collection and processing of equipment data such as the vibration frequency and output power of the compressor, and using a formula to calculate the aging assessment coefficient, so as to timely discover potential aging problems, extend the life of the equipment, and reduce production interruptions and safety accidents caused by equipment failures; based on the pre-trained HAZOP analysis and recognition model, the initial event and its consequences are automatically identified, and the probability of occurrence of the initial event and the failure probability of existing safety measures are calculated in combination with historical operation data, thereby ensuring the accuracy and comprehensiveness of risk assessment and reducing dependence on manual experience; by comprehensively analyzing the probability of occurrence of the initial event and the failure probability of existing safety measures, a more accurate risk probability can be obtained to help decision makers formulate more scientific and reasonable risk management strategies; the Delphi method is used to obtain the failure probability of the recommended measures, and the priority is determined according to the failure probability, so that high-risk events can be handled first, thereby improving the efficiency and effectiveness of responding to emergencies; by generating early warning information based on the risk assessment coefficient, it is ensured that high-risk events can be discovered and handled in a timely manner, greatly reducing the risk of accidents and improving the safety and stability of the entire system. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0060] Figure 1 A schematic diagram of a system module according to an embodiment of the present invention;

[0061] Figure 2 A schematic diagram of a specific process of an embodiment of the present invention;

[0062] Figure 3 Schematic diagram of the method steps of an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than 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.

[0064] See also Figure 1-Figure 2 , the first aspect of the present invention provides an artificial intelligence-based risk assessment system, including: a data acquisition module, a data processing module and a risk assessment module;

[0065] The data acquisition module is used to obtain process data and equipment data of the methanol synthesis reactor;

[0066] The data processing module: analyzes the process data based on the pre-trained HAZOP analysis and recognition model to obtain the recognition result; calculates the aging assessment coefficient based on the equipment data; and obtains the risk probability of the methanol synthesis reactor and the failure probability of the recommended measures by analyzing the recognition result;

[0067] The risk assessment module determines the priority of the recommended measures based on the failure probability of the recommended measures; calculates the risk assessment coefficient based on the risk probability and the aging assessment coefficient, and determines whether to issue an early warning based on the risk assessment coefficient.

[0068] Obtain process data and equipment data of methanol synthesis reactor, including:

[0069] The process of the methanol synthesis reactor is divided into several key nodes according to the process flow; wherein the several key nodes include: raw gas supply and pretreatment process, raw gas heating process, methanol synthesis reaction process, product cooling and separation process, methanol purification and storage process, and waste gas treatment process;

[0070] Obtaining process data at each key node; wherein the process data includes: process parameters and equipment parameters;

[0071] The equipment data of the compressor in the methanol synthesis reactor is collected by a data acquisition device; wherein the equipment data includes: vibration frequency and output power.

[0072] Analyze process data based on pre-trained HAZOP analysis and recognition models, including:

[0073] Extract historical process data of methanol synthesis reactor from the database, and the corresponding analysis results obtained through HAZOP analysis; the analysis results include: initial events, deviation causes, accident consequences, existing safety measures and recommended measures corresponding to the initial events;

[0074] Integrate historical process data into standard input data and analysis results into standard output data; train an artificial intelligence model based on the standard input data and standard output data to obtain a pre-trained HAZOP analysis and recognition model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network;

[0075] The process data acquired in real time is input into the HAZOP analysis and recognition model to obtain a recognition result; wherein the content attribute of the recognition result is consistent with the content attribute of the analysis result.

[0076] The aging assessment coefficients are calculated based on the equipment data, including:

[0077] The vibration frequency is marked as Z and the output power is marked as W;

[0078] The aging assessment coefficient is calculated by the formula: LP = A×e^[Z / (Z+1)]+B×e^[(W-ZW)^2 / ZW^2]; where LP is the aging assessment coefficient, A and B are aging influencing factors, and 0<A<1, 0<B<1, A<B, and ZW is the optimal output power.

[0079] The risk probability analysis steps for methanol synthesis reactors include:

[0080] Extract the initial events identified by the HAZOP analysis identification model and the existing safety measures corresponding to the initial events;

[0081] Based on the start-up operation data of the methanol synthesis reactor, the number of occurrences of the initial event within the preset time period T, and the number of triggering and failure times of the existing safety measures corresponding to the initial event within the preset time period T are obtained;

[0082] By formula: The probability of occurrence of each initial event is calculated; among them, CS i It refers to the probability of occurrence of the i-th initial event, i is the sequence number of the initial event, i = {1, 2, 3…, n}, n is the total number of initial events, X i It refers to the number of times the i-th initial event occurs within the preset time period T;

[0083] By formula The failure probability of the existing safety measures corresponding to the initial event is calculated; where S i,jIt refers to the failure probability of the jth existing safety measure corresponding to the i-th initial event, j is the serial number of the existing safety measure, j = {1, 2, 3…, m}, m is the total number of existing safety measures, Y i,j G refers to the number of failures of the jth existing safety measure corresponding to the i-th initial event within the preset time period T. i,j It refers to the number of times the jth existing safety measure corresponding to the i-th initial event is triggered within the preset time period T;

[0084] The risk probability of the initial event is calculated based on the occurrence probability of the initial event and the failure probability of the existing safety measures corresponding to the initial event.

[0085] The risk probability of the initial event is calculated based on the probability of occurrence of the initial event and the failure probability of the existing safety measures corresponding to the initial event, including:

[0086] By formula The total failure probability of existing safety measures corresponding to the initial event is calculated; where SX i It refers to the total failure probability of existing safety measures corresponding to the i-th initial event;

[0087] By calculating CS i and SX i The product of is used to obtain the risk probability of the initial event.

[0088] Prioritize the recommended actions based on their probability of failure, including:

[0089] Extract recommended actions corresponding to the initial event;

[0090] Obtain the failure probability of the recommended measures through the Delphi method;

[0091] The recommended measures are ranked in ascending order of failure probability; the higher the ranked recommended measures are, the higher the priority they are adopted.

[0092] The risk assessment coefficient is calculated based on the risk probability and the aging assessment coefficient, including:

[0093] Label the risk probability as Fi;

[0094] The risk assessment coefficient is calculated by the formula: P = ɑ × ∑Fi + β × LP; where P is the risk assessment coefficient, ɑ and β are weight coefficients, and ɑ + β = 1, and Fi refers to the risk probability of the i-th initial event.

[0095] Determine whether to issue an early warning based on the risk assessment coefficient, including:

[0096] Determine whether the risk assessment coefficient is greater than the preset risk assessment coefficient threshold; if yes, generate an early warning message and send it to the client; if not, continue to monitor and judge.

[0097] For example, suppose there is a chemical plant that uses a methanol synthesis reactor to produce methanol. In order to ensure the safe operation and efficient production of the reactor, the plant decides to adopt the artificial intelligence-based risk assessment system and method proposed in the present invention. The details are as follows:

[0098] 1. Data acquisition module;

[0099] Process data collection:

[0100] The process of methanol synthesis reactor is divided into several key nodes according to the process flow:

[0101] Raw gas supply and pretreatment process;

[0102] Raw gas heating process;

[0103] Methanol synthesis reaction process;

[0104] Product cooling and separation process;

[0105] Methanol purification and storage process;

[0106] Waste gas treatment process;

[0107] The process data in each key node includes:

[0108] Process parameters such as temperature, pressure, flow, humidity, sulfur content, heating rate, exhaust gas concentration, etc.

[0109] Equipment parameters such as compressor speed, compressor power, filter or adsorber working efficiency, desulfurization tower operating temperature and pressure, cooling medium flow and temperature, condenser and cooler refrigerant flow and temperature.

[0110] Equipment data collection:

[0111] The equipment data of the compressor in the methanol synthesis reactor is collected by data acquisition equipment, including vibration frequency (Z) and output power (W).

[0112] 2. Data processing module;

[0113] Analysis based on pre-trained HAZOP analysis recognition model:

[0114] The historical process data of methanol synthesis reactors and the corresponding analysis results obtained through HAZOP analysis were extracted from the database. These analysis results include initial events, deviation causes, accident consequences, existing safety measures and recommended measures.

[0115] The historical process data are integrated as standard input data, and the analysis results are integrated as standard output data. Based on these data, a convolutional neural network (CNN) or a deep belief network (DBN) is trained to obtain a pre-trained HAZOP analysis and recognition model.

[0116] The process data acquired in real time is input into the HAZOP analysis and identification model to obtain the identification results. For example, the following initial events and their related information are identified:

[0117] Initial event: The temperature in the raw gas heating process is too high.

[0118] Cause of deviation: May be due to heater failure or operator error.

[0119] Consequences of the accident: It may cause the reactor to overheat, leading to seal failure and toxic gas leakage.

[0120] Existing safety measures: Temperature alarm system and emergency shutdown system (ESD) are installed.

[0121] Recommended actions: Optimize heater maintenance schedule and add redundant heating control systems.

[0122] Calculate the aging assessment factor based on the equipment data:

[0123] Assume that the vibration frequency (Z) of the compressor is 10 Hz, the output power (W) is 50 kW, and the optimum output power (ZW) is 45 kW. Calculate the aging assessment coefficient according to the formula: LP = A × e^[Z / (Z+1)] + B × e^[(W-ZW)^2 / ZW^2];

[0124] Assume that the aging influence factors A = 0.3 and B = 0.7. Substitute the numerical values ​​for calculation:

[0125] LP = 1.528;

[0126] 3. Risk assessment module;

[0127] The risk probability is calculated based on the probability of the initial event occurring and the probability of failure of existing safety measures:

[0128] Assume that in the past 6 months, there have been 2 initial overtemperature events in the feed gas heating process. The temperature alarm system has been triggered 10 times during this period, with 1 failure; the emergency shutdown system (ESD) has been triggered 5 times during this period, with no failures.

[0129] Calculate the probability of the initial event occurring:

[0130] Assuming the time period T is 6 months and the total time is 180 days, then:

[0131] CS i=0.011;

[0132] Calculate the failure probability of existing safety measures:

[0133] For temperature alarm systems:

[0134] S i,j =0.1;

[0135] For Emergency Shutdown System (ESD):

[0136] S i,j =0;

[0137] Then the risk probability of the initial event is 0.0011;

[0138] Determine priorities based on the probability of failure of the proposed actions;

[0139] Assume that the failure probability of the recommended measures obtained through the Delphi method is as follows:

[0140] Optimize the maintenance plan for the heater: failure probability is 0.05;

[0141] Add redundant heating control system: failure probability is 0.02;

[0142] Sort by failure probability from small to large, with the priority being:

[0143] 1. Add redundant heating control system (failure probability 0.02);

[0144] 2. Optimize the maintenance plan of the heater (failure probability 0.05);

[0145] Calculate the risk assessment factor based on the risk probability and the aging assessment factor:

[0146] Assuming the weight coefficients α = 0.7, β = 0.3, then:

[0147] P = 0.45917;

[0148] Assuming the preset risk assessment coefficient threshold is 0.3, then:

[0149] Since P>0.3, a warning message is generated and sent to the client, indicating that measures need to be taken to reduce the risk.

[0150] Through the above examples, it is demonstrated how to use the artificial intelligence-based risk assessment system and method proposed in the present invention to solve the potential risk problems in the methanol synthesis reactor. The system can not only accurately quantify the aging degree of the compressor and timely discover potential aging problems, but also automatically identify the initial event and its consequences, and calculate the probability of occurrence of the initial event and the failure probability of existing safety measures in combination with historical operation data, ensuring the accuracy and comprehensiveness of the risk assessment. In addition, through the intelligent early warning mechanism, it helps users take preventive measures in a timely manner, greatly reducing the risk of accidents and improving the safety and stability of the entire system.

[0151] See also Figure 3 The second aspect of the present invention provides a risk assessment method based on artificial intelligence, comprising:

[0152] Step 1: Obtain process data and equipment data of the methanol synthesis reactor;

[0153] Step 2: Analyze the process data based on the pre-trained HAZOP analysis and recognition model to obtain the recognition results; calculate the aging assessment coefficient based on the equipment data;

[0154] Step 3: By analyzing the identification results, the risk probability of the methanol synthesis reactor and the failure probability of the recommended measures are obtained;

[0155] Step 4: Determine the priority of the recommended measures based on the failure probability of the recommended measures; calculate the risk assessment coefficient based on the risk probability and the aging assessment coefficient, and determine whether to issue an early warning based on the risk assessment coefficient.

[0156] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0157] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A risk assessment system based on artificial intelligence, characterized in that: include: Data collection module, data processing module and risk assessment module; The data acquisition module is used to obtain process data and equipment data of the methanol synthesis reactor; The data processing module: analyzes the process data based on the pre-trained HAZOP analysis and recognition model to obtain recognition results; Calculate the aging assessment factor based on the equipment data; and, By analyzing the identification results, the risk probability of the methanol synthesis reactor and the failure probability of the recommended measures are obtained; The risk assessment module determines the priority of the suggested measures based on the failure probability of the suggested measures; The risk assessment coefficient is calculated based on the risk probability and the aging assessment coefficient, and whether to issue an early warning is determined based on the risk assessment coefficient.

2. The artificial intelligence-based risk assessment system according to claim 1, characterized in that: The process data and equipment data of the methanol synthesis reactor are obtained, including: The process of the methanol synthesis reactor is divided into several key nodes according to the process flow; wherein the several key nodes include: raw gas supply and pretreatment process, raw gas heating process, methanol synthesis reaction process, product cooling and separation process, methanol purification and storage process, and waste gas treatment process; Obtaining process data at each key node; wherein the process data includes: process parameters and equipment parameters; The equipment data of the compressor in the methanol synthesis reactor is collected by a data acquisition device; wherein the equipment data includes: vibration frequency and output power.

3. The artificial intelligence-based risk assessment system according to claim 1, characterized in that: The process data is analyzed based on the pre-trained HAZOP analysis and recognition model, including: Extract historical process data of methanol synthesis reactor from the database, and the corresponding analysis results obtained through HAZOP analysis; the analysis results include: initial events, deviation causes, accident consequences, existing safety measures and recommended measures corresponding to the initial events; Integrate historical process data into standard input data and analysis results into standard output data; train an artificial intelligence model based on the standard input data and standard output data to obtain a pre-trained HAZOP analysis and recognition model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network; The process data acquired in real time is input into the HAZOP analysis and recognition model to obtain a recognition result; wherein the content attribute of the recognition result is consistent with the content attribute of the analysis result.

4. The artificial intelligence-based risk assessment system according to claim 1, characterized in that: The aging assessment coefficient is calculated based on the equipment data, including: The vibration frequency is marked as Z and the output power is marked as W; The aging assessment coefficient is calculated by the formula: LP = A×e^[Z / (Z+1)]+B×e^[(W-ZW)^2 / ZW^2]; where LP is the aging assessment coefficient, A and B are aging influencing factors, and 0<A<1, 0<B<1, A<B, and ZW is the optimal output power.

5. The artificial intelligence-based risk assessment system according to claim 1, characterized in that: The risk probability analysis step of the methanol synthesis reactor comprises: Extract the initial events identified by the HAZOP analysis identification model and the existing safety measures corresponding to the initial events; Based on the start-up operation data of the methanol synthesis reactor, the number of occurrences of the initial event within the preset time period T, and the number of triggering and failure times of the existing safety measures corresponding to the initial event within the preset time period T are obtained; By formula: The probability of occurrence of each initial event is calculated; among them, CS i It refers to the probability of occurrence of the i-th initial event, i is the sequence number of the initial event, i = {1, 2, 3…, n}, n is the total number of initial events, X i It refers to the number of times the i-th initial event occurs within the preset time period T; By formula The failure probability of the existing safety measures corresponding to the initial event is calculated; where S i,j It refers to the failure probability of the jth existing safety measure corresponding to the i-th initial event, j is the serial number of the existing safety measure, j = {1, 2, 3…, m}, m is the total number of existing safety measures, Y i,j G refers to the number of failures of the jth existing safety measure corresponding to the i-th initial event within the preset time period T. i,j It refers to the number of times the jth existing safety measure corresponding to the i-th initial event is triggered within the preset time period T; The risk probability of the initial event is calculated based on the occurrence probability of the initial event and the failure probability of the existing safety measures corresponding to the initial event.

6. The artificial intelligence-based risk assessment system according to claim 5, characterized in that: The risk probability of the initial event is calculated based on the probability of occurrence of the initial event and the failure probability of the existing safety measures corresponding to the initial event, including: By formula The total failure probability of existing safety measures corresponding to the initial event is calculated; where SX i It refers to the total failure probability of existing safety measures corresponding to the i-th initial event; By calculating CS i and SX i The product of is used to obtain the risk probability of the initial event.

7. The artificial intelligence-based risk assessment system according to claim 1, characterized in that: The determining the priority of the recommended measures based on the failure probability of the recommended measures includes: Extract recommended actions corresponding to the initial event; Obtain the failure probability of the recommended measures through the Delphi method; The recommended measures are ranked in ascending order of failure probability; the higher the ranked recommended measures are, the higher the priority they are adopted.

8. The artificial intelligence-based risk assessment system according to claim 1, characterized in that: The risk assessment coefficient is calculated based on the risk probability and the aging assessment coefficient, including: Label the risk probability as Fi; The risk assessment coefficient is calculated by the formula: P = ɑ × ∑Fi + β × LP; where P is the risk assessment coefficient, ɑ and β are weight coefficients, and ɑ + β = 1, and Fi refers to the risk probability of the i-th initial event.

9. The artificial intelligence-based risk assessment system according to claim 1, characterized in that: The determining whether to issue an early warning based on the risk assessment coefficient includes: Determine whether the risk assessment coefficient is greater than the preset risk assessment coefficient threshold; if yes, generate an early warning message and send it to the client; if not, continue to monitor and judge.

10. An artificial intelligence-based risk assessment method, applied to an artificial intelligence-based risk assessment system according to any one of claims 1 to 9, characterized in that: include: Step 1: Obtain process data and equipment data of the methanol synthesis reactor; Step 2: Analyze the process data based on the pre-trained HAZOP analysis and recognition model to obtain the recognition results; The aging assessment coefficient is calculated based on the equipment data; Step 3: By analyzing the identification results, the risk probability of the methanol synthesis reactor and the failure probability of the recommended measures are obtained; Step 4: Determine the priority of the recommended measures based on their failure probability; The risk assessment coefficient is calculated based on the risk probability and the aging assessment coefficient, and whether to issue an early warning is determined based on the risk assessment coefficient.

Images