A digital intelligent control method and device for multi-modal process safety risks

Abstract

The present invention discloses a digital intelligent control method and device for multi-modal process safety risks, belonging to the technical field of safety production monitoring in the process industry. The method includes: deconstructing process safety risk knowledge in a modal manner to construct a first mapping relationship between safety risk modes and parameters; digitally perceiving and monitoring physical quantity data to construct a second mapping relationship between physical quantities and parameters; after obtaining real-time data and constructing the first and second mapping relationships, performing intelligent extraction of identifiers, multi-modal intelligent fusion analysis, multi-modal intelligent identification and prediction, intelligent decision-making, real-time early warning and control of risk grading, so as to realize digital intelligent control. The implementation device of the method includes a first construction module, a second construction module and an intelligent decision-making and control module. The present invention applies digital and digital intelligent technical methods to the safety control of the process production site, which can effectively avoid on-site operation risks, reduce the burden on front-line workers and managers, and reduce the occurrence of safety accidents.

Description

Technical Field

[0001] The present invention belongs to the technical field of safety production monitoring in the process industry, and more specifically, relates to a digital and intelligent control method and device for multi-modal process safety risks. Background Art

[0002] Currently, the process industry in China has a huge scale and highly concentrated production capacity. The production capacities of industries such as steel, power, cement, and paper-making all rank first in the world. The overall production process, technical equipment, and automation level of the industry have been greatly improved, and the industry scale is constantly expanding. At the same time, the unique production mode of the process industry also has relatively prominent industry problems. In terms of process safety production control, the traditional control process is usually that the central control personnel issue a series of control instructions to the equipment on the production line through the upper computer and the lower computer to ensure that the process production line is in a normal and safe state. However, due to reasons such as the frequent change of raw materials in the process industry, the complex mechanism of the production process, and the continuous and uninterruptible process, it is often impossible to comprehensively monitor some raw material components, production equipment status, and process parameters in real time, resulting in the current situation of difficult perception, difficult decision-making, and difficult control.

[0003] In addition, the control means for safety risks in the process industry often overlap with the control of product production quality. If any process has problems, it will not only cause product quality problems but also affect the safe operation of the entire process production line. However, there is currently no dedicated monitoring and control method or system device for safety risks in the process production process. Summary of the Invention

[0004] In view of the above-mentioned defects or improvement requirements of the prior art, the present invention provides a system method and device for digital and intelligent control of multi-modal process safety risks, aiming to obtain real-time data of relevant physical quantities, and based on the mappings between the constructed process safety risk modes and process safety risk mode parameters, and between process safety risk mode parameters and physical quantities, perform intelligent computational analyses such as data identification, parameter identification, mode identification, and multi-modal fusion to identify the current safety risk mode, and then perform hierarchical early warning and decision-making control according to the current safety risk mode, thereby solving the technical problem of difficult decision-making control for safety risks in the process production process in the prior art.

[0005] To achieve the above object, according to one aspect of the present invention, there is provided a digital and intelligent control method for multi-modal process safety risks, including:

[0006] S1 - Modal Decomposition of Process Safety Risk Knowledge: Establish a mapping between the process safety risk mode and the process safety risk mode parameters, denoted as the first mapping relationship; the process safety risk mode includes: the types of process safety accidents and their corresponding safety risk factors; the process safety risk mode parameters are the characterization parameters corresponding to the safety risk factors.

[0007] S2 - Digital Sensing of Multimodal Parameters of Process Safety Risk: Use front - end sensing devices to identify the sensitive points and frequencies of changes in the process safety risk mode parameters, and set a monitoring period to obtain real - time data of the corresponding physical quantities, and establish a mapping between the process safety risk mode parameters and the monitored physical quantities, denoted as the second mapping relationship.

[0008] S3 - Digital and Intelligent Control of Multimodal Process Safety Risk: According to the real - time data of the currently monitored physical quantities, the established first mapping relationship and the second mapping relationship, perform data identification, parameter identification, mode identification, and intelligent fusion analysis and calculation of safety risk multimodality to identify the current safety risk mode; realize hierarchical early warning and decision - making control of process safety risk.

[0009] In one embodiment, the S1 includes:

[0010] S101: Establish a process safety risk mode set, including: a process safety accident type set A and a safety risk factor set A' corresponding to the accident types; A = {A1, …, A i , … A m}, A' = {A 11 , …, A ij , … A mn}, A i is the i - th accident type, m is the total number of accident types, A ij is the j - th risk factor corresponding to the i - th accident type, and n is the total number of risk factors for the m - th accident type.

[0011] S102: Establish a process safety risk mode parameter set B, B = {B1, B2, B3}, where B1 is a subset of operation environment - related parameters, B2 is a subset of equipment status - related parameters, and B3 is a subset of process logic - related parameters.

[0012] S103: Characterize the first mapping relationship through the change function f of the corresponding parameters. Let x be any parameter in the parameter set B, then

[0013] In one embodiment, the S101 includes:

[0014] All safety accident types that have actually occurred and have a probability of occurring but have not occurred during the production process of the target process are used as the elements A of the process safety accident type set A.i ;

[0015] Regarding the risk factors that may lead to each type of safety accident as elements A in the process safety risk set A'. ij , thereby establishing the process safety risk mode set.

[0016] In one embodiment, the S2 includes:

[0017] S201: Identify the spatial points sensitive to the change of process safety risk mode parameters and the parameter change frequency;

[0018] S202: Determine the positions for deploying the front-end sensing devices and set the sensing period, so as to obtain the real-time data of the monitored physical quantities;

[0019] S203: Construct the second mapping relationship between the process safety risk mode parameters and the monitored physical quantities.

[0020] In one embodiment,

[0021] The S201 includes: Select the most sensitive points as the deployment positions according to the actual operating environment and operating equipment conditions of the target process production;

[0022] The S202 includes: Select 2 to 10 times the change frequency of the process safety risk mode parameters as the monitoring frequency.

[0023] In one embodiment, the S202 includes:

[0024] When the process safety risk mode parameters are operation environment type parameters or equipment status type parameters with continuous and rapid changes, then select 6 - 10 times the change frequency of the process safety risk mode parameters;

[0025] When the process safety risk mode parameters are process logic type parameters, then select 2 - 5 times the change frequency of the process safety risk mode parameters.

[0026] In one embodiment, the S3 includes:

[0027] S301: According to the real-time data of the currently monitored physical quantities, the constructed first mapping relationship and the second mapping relationship, perform data identification, parameter identification, and mode identification;

[0028] S302: Based on the identification extraction results, establish a safety risk multi-modal intelligent fusion model to identify the current safety risk mode;

[0029] S303: If there is or will be an abnormality in the current safety risk mode, perform hierarchical safety real-time decision-making early warning and send risk control instructions.

[0030] According to another aspect of the present invention, there is provided a digital intelligent control device for multi-modal process safety risks, which is used to execute the above method, and includes:

[0031] A first construction module, which is used to construct a mapping between the process safety risk mode and the process safety risk mode parameters, denoted as the first mapping relationship; the process safety risk mode includes: the type of process safety accident and its corresponding safety risk factors; the process safety risk mode parameters are the characterization parameters corresponding to the safety risk factors;

[0032] A second construction module, which is used to sense the change of the process safety risk mode parameters by using a front-end device to obtain real-time data of the monitored physical quantity, and construct a second mapping relationship between the process safety risk mode parameters and the monitored physical quantity;

[0033] An intelligent decision-making and control module, which is used to realize functions such as intelligent extraction of identifiers, multi-modal intelligent fusion analysis, safety mode identification, real-time hierarchical early warning, and intelligent control according to the sensed current monitored physical quantity data, the first mapping relationship, and the second mapping relationship.

[0034] Generally speaking, compared with the prior art through the above technical solutions conceived by the present invention, the following beneficial effects can be achieved:

[0035] (1) Make the safety risk knowledge in the process industry more standardized and systematic, which is conducive to improving the ability of production personnel and managers at all levels in this industry to respond to safety risks.

[0036] (2) Applying digital and intelligent technical methods to the safety control of the process production site can effectively avoid on-site operation risks, reduce the burden on front-line workers and managers, and reduce the occurrence of safety accidents.

[0037] (3) It is conducive to improving the current situation of insufficient digitalization in the safety risk control of the process industry, and combining hot technologies such as machine learning and edge computing to make up for the deficiencies of existing methods, and further improve the digital intelligent control level of production safety risks in the process industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a flowchart of a digital intelligent control method for multi-modal process safety risks in an embodiment of the present invention;

[0039] Figure 2 is a flowchart of S1 in a digital intelligent control method for multi-modal process safety risks in an embodiment of the present invention;

[0040] Figure 3 is a schematic diagram of the relationship between a process safety risk mode set and a parameter set in an embodiment of the present invention;

[0041] Figure 4a It is a schematic flowchart of S2 in the digital control method for multi-modal process safety risks in an embodiment of the present invention;

[0042] Figures 4b - 4f It is a schematic diagram showing the mapping relationship between monitored physical quantities and modal parameters in an embodiment of the present invention;

[0043] Figure 5 It is a schematic flowchart of S3 in the digital control method for multi-modal process safety risks in an embodiment of the present invention;

[0044] Figure 6 It is a schematic diagram showing the composition of the digital control device for multi-modal process safety risks in an embodiment of the present invention. Detailed implementation manners

[0045] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0046] As Figure 1 shown, the present invention provides a digital control method for multi-modal process safety risks, including:

[0047] To achieve the above object, according to one aspect of the present invention, there is provided a digital control method for multi-modal process safety risks, including:

[0048] S1 - Modal decomposition of process safety risk knowledge: Construct a mapping between the process safety risk modality and the process safety risk modal parameters, denoted as the first mapping relationship; the process safety risk modality includes: the types of process safety accidents and their corresponding safety risk factors; the process safety risk modal parameters are the characterization parameters corresponding to the safety risk factors;

[0049] S2 - Digital perception of multi-modal parameters of process safety risks: Use a front-end perception device to identify the sensitive points and frequencies of changes in the process safety risk modal parameters, and set a monitoring period to obtain real-time data of the corresponding physical quantities, and construct a mapping between the process safety risk modal parameters and the monitored physical quantities, denoted as the second mapping relationship;

[0050] Digital control of multi-modal process safety risks: Based on the real-time data of the currently monitored physical quantities, the established first mapping relationship and the second mapping relationship, perform data identification, parameter identification, modal identification, and multi-modal intelligent fusion analysis and calculation of safety risks to identify the current safety risk modality; realize hierarchical early warning and decision-making control of process safety risks. In one embodiment, S1 includes:

[0051] S101: Establish a set of process safety risk modalities, including: a set of process safety accident types A and a set of safety risk factors A' corresponding to the accident types; A = {A1, …, A i , … A m}, A' = {A 11 , …, A ij , … A mn}, A i is the i-th accident type, m is the total number of accident types, A ij is the j-th risk factor corresponding to the i-th accident type, and n is the total number of risk factors for the m-th accident type;

[0052] S102: Establish a set of process safety risk modal parameters B, B = {B1, B2, B3}, where B1 is a subset of operation environment parameters, B2 is a subset of equipment status parameters, and B3 is a subset of process logic parameters;

[0053] S103: Characterize the first mapping relationship through the change function f of the corresponding parameters. Let x be any parameter in the parameter set B, then

[0054] Specifically, based on engineering science and safety science, according to the specific process steps of the target process, on the basis of clarifying the common safety accident types and risk factors in aspects such as equipment, environment, and personnel, establish a set of process safety risk modalities, a set of process safety risk modal parameters, and a relationship model between process safety risk modalities and parameters.

[0055] Further, the set of process safety risk modalities includes a set of process safety accident types A and a set of safety risk factors A' corresponding to the accident types. Among them, the elements in the set of process safety accident types A are a specific accident type A i , that is, A = {A1, A2, …, A i}; the elements in the set of safety risk factors A' are the risk factors A i corresponding to a certain accident type A ij , that is, A' = {A 11 , A 12 , …, A 21 , A 22 , …, A ij}.

[0056] Furthermore, the process safety risk modal parameter set B is a further decomposition of the process safety risk modal set and is a characterization parameter of the risk factors that cause process safety accidents. A lowercase letter is used to represent a certain parameter, that is, B = {a, b, c,...}. According to the parameter type, it can be divided into the operation environment parameter subset B1, the equipment status parameter subset B2, and the process logic parameter subset B3, then B = {B1, B2, B3}. The operation environment parameters include the temperature, humidity, atmospheric pressure, wind speed, etc. of the operation area. The equipment status parameters include the stress, strain, deformation, inclination angle, length, width, etc. of the equipment structural components. The process logic parameters include the vibration frequency, vibration amplitude, speed, pressure, flow rate, etc. of the process control.

[0057] Furthermore, the relationship model between the process safety risk modal and parameters includes the relationship model between the process safety risk modal and the corresponding parameters. The evolution law of the occurrence of safety accidents can be characterized by the changes of specific risk modal parameters. That is to say, the influence relationship between the safety risk factor set A' and the safety accident type set A can be characterized by the change function f of the corresponding parameters. Let x be any parameter in the parameter set B, then

[0058] In one of the embodiments, S101 includes:

[0059] All the safety accident types that actually occurred and had the probability of occurring but did not occur during the production process of the target process are taken as the elements A of the process safety accident type set A i ;

[0060] The risk factors that cause each safety accident are taken as the elements A in the process safety risk set A' ij , thereby establishing the process safety risk modal set.

[0061] Furthermore, the specific parameters included in the process safety risk modal parameter set B depend on the process safety risk modal set of the target process, that is, the process safety accident type set A and the safety risk factor set A'. All the safety accident types that actually occurred or had the probability of occurring but did not occur during the production process of the target process are taken as the elements A of the process safety accident type set A i The risk factors that cause a certain safety accident are taken as the elements A in the process safety risk set ij, establish a set of process safety risk modes. Modal parameters are the means to characterize safety risk modes. According to the formation causes of each safety accident and each risk factor, sort out the modal parameters involved therein, and establish a set of process safety risk modal parameters B. Then, on the premise of establishing the set of process safety risk modes and the set of process safety risk modal parameters as described above, further explore the internal relationship between the process safety risk modes and the corresponding parameters, and establish a relationship model between the process safety risk modes and the parameters. It should be noted that since different processes have different sets of safety accident types A and sets of safety risk factors A′, the corresponding sets of process safety risk modal parameters B also have specificities.

[0062] In one embodiment, S2 includes:

[0063] S201: Identify the spatial points sensitive to changes in process safety risk modal parameters and the parameter change frequencies;

[0064] S202: Determine the positions for arranging the front-end sensing devices and set the sensing periods, so as to obtain real-time data of the monitored physical quantities;

[0065] S203: Construct the second mapping relationship between the process safety risk modal parameters and the monitored physical quantities.

[0066] Specifically, use the front-end sensing devices to sense the changes in process safety risk modal parameters, collect real-time data of the corresponding physical quantities, and then establish a mapping relationship model between the two based on the association path between the monitored physical quantities and the process safety risk modal parameters. Specifically, it involves the identification of spatial points sensitive to changes in process safety risk modal parameters and the identification of parameter change frequencies of process safety risk modal parameters.

[0067] Furthermore, there is a one-to-one or many-to-one mapping relationship between the monitored physical quantities and the process safety risk modal parameters, that is, there is one or more physical quantities corresponding to a certain type of attribute modal parameter. According to the mapping relationships between various types of attribute modal parameters involved in the target process and their monitored physical quantities, establish a mapping model between the process safety risk modal parameters and the monitored physical quantities, including the monitoring positions and monitoring frequencies.

[0068] In one embodiment, S201 includes:

[0069] Select the most sensitive points as the layout positions according to the actual operation environment and operating equipment conditions of the target process production. Further, when identifying the sensitive points and selecting the layout positions of the parameter sensing devices, according to the actual operation environment and operating equipment conditions of the target process production, the most sensitive points should be preferentially selected as the layout positions. If it is too difficult to arrange at the most sensitive points, the second most sensitive points can be selected as the layout positions, and so on.

[0070] In one embodiment, S202 includes:

[0071] Select 2 to 10 times the change frequency of the process safety risk modal parameters as the monitoring frequency. Specifically, based on the identification of sensitive points, determine the position P of the parameter perception device, and based on the identification of the change frequency, set the monitoring frequency T of the perception device. The number of monitoring times should be 2 to 10 times the parameter change frequency.

[0072] In one embodiment, S202 includes:

[0073] When the process safety risk modal parameter is a continuously and rapidly changing operation environment parameter or equipment status parameter, then select 6 - 10 times the change frequency of the process safety risk modal parameter;

[0074] When the process safety risk modal parameter is a process logic parameter, then select 2 - 5 times the change frequency of the process safety risk modal parameter.

[0075] Specifically, when identifying the parameter change frequency and setting the monitoring frequency of the perception device, it should also be based on the actual operation environment and operating equipment conditions of the target process production. Priority should be given to selecting 2 to 10 times the parameter change frequency as the monitoring frequency. For continuously and rapidly changing operation environment parameters and equipment status parameters, take the larger value (6 - 10 times), and for process logic parameters, take the smaller value (2 - 5 times).

[0076] For example, in combination with the tundish, which is a key process equipment in the continuous casting process of steel, introduce the specific implementation of S1 - Modalization and Parameter Decomposition of Process Safety Risk Knowledge, such as Figure 2 shown, including S101 - Obtain the process safety risk modal set, S102 - Obtain the process safety risk modal parameter set, and S103 - Establish the relationship model between the process safety risk modal and parameters.

[0077] S101 - Construct the process safety risk modal set: The continuous casting process has many process steps, strong front - back correlation, high operation risk, and the production equipment mostly operates under high temperature and high pressure. As one of the core node equipment in the continuous casting process, the tundish's safety accident types mainly involve gas explosion, molten steel splashing and scalding, breakout explosion, etc. Then the safety accident type set A of the tundish can be expressed as:

[0078] {A1, A2, A3}

[0079] Such as Figure 3 shown, where,

[0080] A1 is a gas explosion accident;

[0081] A2 is a molten steel scalding accident;

[0082] A3 is an accident of breakout explosion.

[0083] Gas leakage in the tundish baking area is a risk factor leading to gas explosion accidents. Insufficient safe height of the tundish molten steel, malfunction of the liquid level automatic measurement system, insufficient molten steel protective slag, defective submerged nozzle structure, and unstable molten steel flow rate at the submerged nozzle are risk factors leading to molten steel scalding accidents. Unstable lifting speed of the tundish car, misalignment of the stopper rod, jamming of the stopper rod, shaking of the stopper rod, breakage of the stopper rod, and too fast control speed of the stopper rod are factors leading to breakout explosion accidents. Then, the set of safety risk factors A' of the tundish can be expressed as:

[0084] {A 11 ,A 21 ,A 22 ,A 23 ,A 24 ,A 31 ,A 32 ,A 33 ,A 34 ,A 35 ,A 36}

[0085] As Figure 3 shown, where

[0086] A 11 is the risk factor of gas leakage in the tundish baking area;

[0087] A 21 is the risk factor of insufficient safe height of the tundish molten steel and malfunction of the liquid level automatic measurement system;

[0088] A 22 is the risk factor of insufficient molten steel protective slag;

[0089] A 23 is the risk factor of defective submerged nozzle structure;

[0090] A 24 is the risk factor of unstable molten steel flow rate at the submerged nozzle;

[0091] A 31 is the risk factor of unstable lifting speed of the tundish car;

[0092] A 32 is the risk factor of misalignment of the stopper rod;

[0093] A 33 is the risk factor of jamming of the stopper rod;

[0094] A 34 is the risk factor of shaking of the stopper rod;

[0095] A 35 is the risk factor of breakage of the stopper rod;

[0096] A 36 is a risk factor for the too-fast stopper control speed.

[0097] S102 - Construct the modal parameter set of process safety risks: The direct cause of the explosion is the excessive gas concentration caused by gas leakage in the tundish baking area. Therefore, the gas concentration in the tundish baking area can be used as the modal parameter of A1 to characterize whether the target process is in a safe state. When the safe height of the molten steel in the tundish is insufficient and the liquid level automatic measurement system fails, the molten steel may splash out from the upper mouth of the tundish, causing personal burns. At the same time, when the molten steel surface in the tundish leaks red due to insufficient flux, it may also cause the molten steel to splash and cause personal burns. In addition, when the tundish bottom nozzle has structural defects such as cracks and breakages, or the molten steel flow rate control is unstable, it often causes the molten steel to splash and cause personal burns. Here, the safe height of the molten steel in the tundish, the surface state of the molten steel, the structural state of the bottom nozzle, and the molten steel flow rate at the bottom nozzle can be used as the four modal parameters of A2 to characterize whether the target process is in a safe state. During the preparation for starting casting, once the lifting speed of the tundish car is unstable or too fast, it may cause the molten steel inside the tundish to slosh out due to inertia, resulting in an explosion accident. In addition, when pouring molten steel from the tundish into the mold, the failure of the stopper control caused by the misalignment, jamming, shaking, or fracture of the stopper, or the too-fast control speed, may cause the molten steel in the mold to overflow, and then lead to a breakout explosion accident. Here, the lifting speed of the tundish car, the position of the tundish stopper, the stress state of the tundish stopper, the structural state of the tundish stopper, and the control speed of the tundish stopper are used as the five modal parameters of A3 to characterize whether the target process is in a safe state. All the parameter elements in the modal parameter set B of process safety risks are:

[0098] {a, b, c, d, e, f, g, h, i, j};

[0099] As Figure 3 shown,

[0100] a represents the gas concentration in the tundish baking area, which belongs to the operation environment type parameter;

[0101] b represents the safe height of the molten steel in the tundish, which belongs to the process logic type parameter;

[0102] c represents the surface state of the molten steel, which belongs to the process logic type parameter;

[0103] d represents the structural state of the bottom nozzle, which belongs to the equipment state type parameter;

[0104] e represents the molten steel flow rate at the bottom nozzle, which belongs to the process logic type parameter;

[0105] f represents the lifting speed of the tundish car, which belongs to the process logic type parameter;

[0106] g represents the position of the tundish stopper rod, which belongs to the equipment status type of parameters;

[0107] h represents the stress state of the tundish stopper rod, which belongs to the equipment status type of parameters;

[0108] i represents the structural state of the tundish stopper rod, which belongs to the equipment status type of parameters;

[0109] j represents the control speed of the tundish stopper rod, which belongs to the process logic type of parameters.

[0110] The process safety risk modal parameter set B can be further expressed as: {B1, B2, B3}.

[0111] Such as Figure 3 shown, where

[0112] B1 = {a};

[0113] B2 = {d, g, h, i};

[0114] B3 = {b, c, e, f, j};

[0115] S103 - Construct the relationship model between the process safety risk modal and parameters, that is, the first mapping relationship: Such as Figure 3 shown, for most process safety risk modals, the relationship with their parameters is one-to-one, but not all process safety risk modals and their parameters have a strict one-to-one correspondence. Some safety risk modals can be characterized by the same parameter. For example, for A 33 - the jamming of the tundish stopper rod and A 34 - the swaying of the tundish stopper rod, these two safety risk modals can both be characterized by the parameter h - the stress state of the tundish stopper rod to represent their changes. In addition, in general processes, there are also cases where the same parameter is used to characterize the changes of three or more safety risk modals.

[0116] It should be noted that although the process safety risk modal set, the process safety risk modal parameter set, and the first mapping relationship have been formally described here, it does not limit their expression forms and expression scopes.

[0117] Such as Figure 4a shown, in S2, digital perception is to use digital means to sense the changes of process safety risk modal parameters through measuring equipment, collect the real-time data of the corresponding physical quantities, and then establish the mapping relationship model between the monitored physical quantities and the process safety risk modal parameters, that is, the second mapping relationship.

[0118] S201 refers to the analysis and determination of the most sensitive and significant position of parameter changes and the frequency characteristics of parameter changes over time based on the modal deconstruction of S1-process safety risk knowledge and the corresponding equipment working principle when the modal parameters of tundish safety risk change. For example, when the tundish is pouring molten steel, the speed of the stopper rod is changed by manual control or automatic control of the system to change the operating rod outside the tundish. Then the most sensitive point of parameter j should be the connection between the outside operating rod and the tundish stopper rod. At the same time, the frequency characteristics of the change of the tundish stopper rod speed over time are related to the actual pouring situation. When the liquid level in the crystallizer is not appropriate, the stopper rod mechanism should be controlled to control the speed of the molten steel flowing into the crystallizer from the tundish outlet by changing the stopper rod speed, so that the liquid level in the crystallizer remains stable, that is, the change frequency of parameter j can be identified according to the actual operation situation. Parameter j-tundish stopper rod control speed belongs to the process logic type parameter. The change frequency of process logic type parameters generally needs to be identified based on the specific operation process, while the operating environment type parameters and equipment status type parameters are in a state to be monitored during the entire process of process production operation.

[0119] S202 is to deploy parameter sensing devices and set the monitoring frequency according to the parameter sensitive points and change frequencies identified in S201. It should be noted that when selecting the deployment points of parameter sensing devices, the most sensitive points should be selected first according to the actual situation of the target process production. If the deployment difficulty at the most sensitive point is too high, the secondary sensitive point can be selected as the deployment position, and the deployment position is not unique. When setting the monitoring frequency of the sensing device, it should also be based on the actual situation of the target process production. In order to perceive the data as comprehensively and in real time as possible, the monitoring frequency should be selected 2 to 10 times the parameter change frequency. The process logic parameters take small values, and the continuous and rapid change parameters take large values.

[0120] S203 is to mine and construct a second mapping relationship between the monitored physical quantity and the corresponding parameter based on the layout positions and monitoring frequencies of all parameter sensing devices determined in S202.

[0121] When monitoring the gas concentration in the tundish baking zone with parameter a, if there is only one actual monitoring position P a and monitoring frequency T a , then the mapping relationship model between the parameter a and the monitored physical quantity can be expressed as Figure 4b ; If there are two or more actual monitoring locations for parameter a, the mapping relationship model between parameter a and the monitored physical quantity can be expressed as Figure 4c and Figure 4d Parameter h-intermediate plug rod displacement state is a characterization of A 33 -Stick stick and A 34 - Equipment status parameters for the safe state of stopper rod shaking. The monitored equipment is the stopper rod. The displacement monitoring frequency is T his the same, but the specific displacement monitoring position P h1 and P h2 are different. The monitoring position for the stopper sticking should be the connecting part, and the monitoring position for the stopper swaying should be the rod body. Then, the mapping relationship model between the parameter h and the monitored physical quantity can be expressed as Figure 4e ; if in the actual process, for the monitoring position P k of a certain parameter k is the same, while the monitoring frequency T k1 and T k2 are different, then the mapping relationship model between the parameter k and the monitored physical quantity can be expressed as Figure 4f .

[0122] In the actual process, there is a situation where one attribute modal parameter corresponds to multiple monitored physical quantities. At this time, the mapping relationship model between the monitored physical quantity and the modal parameter is a combination of the above Figures 4b - 4f multiple mapping relationship models.

[0123] It should be noted that although multiple mapping relationship models between the parameters and the monitored physical quantities have been described here, it does not limit their types and scopes because of this.

[0124] For example Figure 5 as shown, in one embodiment, S3 includes:

[0125] S301: According to the real-time data of the current monitored physical quantity, the constructed first mapping relationship and the second mapping relationship, layer by layer, intelligently extract the data identifier, parameter identifier, and modal identifier;

[0126] S302: Based on the identification extraction result, establish a multi-modal intelligent fusion model for safety risks to identify the current safety risk modality;

[0127] S303: If there is or will be an abnormality in the current safety risk modality, according to the classification of the safety level, conduct real-time intelligent decision-making early warning and send risk control instructions.

[0128] Furthermore, S3 involves a digital and intelligent application platform integrating real-time data reception, automatic identification extraction, multi-modal intelligent fusion analysis, safety risk level classification, and intelligent decision-making control. This platform includes functional units for data reception, digital-to-analog conversion, identification extraction, multi-modal data intelligent fusion, safety status hierarchical early warning, and intelligent decision-making control, and can realize functions such as data reception, identification extraction, multi-modal intelligent fusion analysis, real-time early warning, and intelligent decision-making control for multi-modal and multi-parameter data from multiple levels. In another aspect of the present invention, a digital and intelligent control device for process safety risk modalities may include a first construction module, a second construction module, and a decision control module, as Figure 6 shown.

[0129] The first construction module is used to construct the mapping between the process safety risk mode and the process safety risk mode parameters, denoted as the first mapping relationship; the process safety risk mode includes: the types of process safety accidents and their corresponding safety risk factors; the process safety risk mode parameters are the characterization parameters corresponding to the safety risk factors.

[0130] The second construction module is used to monitor the change of the process safety risk mode parameters by using the front-end sensing device to obtain the monitored physical quantity, and construct the mapping between the process safety risk mode parameters and the monitored physical quantity, denoted as the second mapping relationship.

[0131] The intelligent decision-making and control module is used to realize functions such as intelligent extraction of identification, multi-modal intelligent fusion analysis, safety mode identification, real-time hierarchical early warning, and intelligent control according to the currently sensed monitored physical quantity data, the first mapping relationship, and the second mapping relationship.

[0132] Those skilled in the art can easily understand that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A digital intelligent control method for multi-modal process safety risks, characterized in that, Including: S1 - Modal Decomposition of Process Safety Risk Knowledge: Establish a mapping between the process safety risk mode and the process safety risk mode parameters, denoted as the first mapping relationship; The process safety risk mode includes: the types of process safety accidents and their corresponding safety risk factors; the process safety risk mode parameters are the characterization parameters corresponding to the safety risk factors; S2 - Digital Sensing of Multimodal Parameters of Process Safety Risk: Use front - end sensing devices to identify the sensitive points and frequencies of changes in the process safety risk mode parameters, and set a monitoring period to obtain real - time data of the monitored physical quantities, and establish a mapping between the process safety risk mode parameters and the monitored physical quantities, denoted as the second mapping relationship; S3 - Digital and Intelligent Control of Multimodal Process Safety Risk: According to the real - time data of the currently monitored physical quantities, the established first mapping relationship and the second mapping relationship, perform data identification, parameter identification, mode identification, and intelligent fusion analysis and calculation of safety risk multimodality to identify the current safety risk mode, and achieve hierarchical early warning and decision - making control of process safety risk.

2. The digital intelligent control method for multi-modal process safety risks according to claim 1, wherein The S1 includes: S101: Establish a process safety risk mode set, including: a process safety accident type set A and a safety risk factor set A' corresponding to the accident type; A = {A1, …, A i , … A m}, A' = {A 11 , …, A ij , … A mn}, A i is the i-th accident type, m is the total number of accident types, A ij is the j-th risk factor corresponding to the i-th accident type, and n is the total number of risk factors for the m-th accident type; S102: Establish a set B of process safety risk mode parameters, B = {B1, B2, B3}, where B1 is a subset of operation environment - type parameters, B2 is a subset of equipment status - type parameters, and B3 is a subset of process logic - type parameters; S103: Characterize the first mapping relationship through the variation function f of the corresponding parameters. Let x be any parameter in the parameter set B, then 3. The digital intelligent control method for multi-modal process safety risks according to claim 2, wherein, The S101 includes: All types of safety accidents that have actually occurred and are likely to occur but have not occurred during the production process of the target process are regarded as elements A of the set A of process safety accident types i ; The risk factors that lead to each type of safety accident are taken as the elements A in the process safety risk set A'. ij , thereby establishing the process safety risk mode set.

4. The digital intelligent control method for multi-modal process safety risks according to claim 1, characterized in that The S2 includes: S201: Identify the spatial points and parameter change frequencies that are sensitive to changes in process safety risk mode parameters; S202: Determine the positions for deploying front - end sensing devices and set the sensing period to obtain real - time data of the monitored physical quantities; S203: Establish the second mapping relationship between the process safety risk mode parameters and the monitored physical quantities.

5. The digital and intelligent control method and control system method for multimodal process safety risk according to claim 4, characterized in that The S201 includes: Select the most sensitive points as the deployment positions according to the actual operation environment and operating equipment conditions of the target process production; The S202 includes: Select 2 to 10 times the change frequency of the process safety risk mode parameters as the monitoring frequency.

6. The digital intelligent control method and control system for multi-modal process safety risks according to claim 5, characterized in that, The S202 includes: When the process safety risk mode parameters are operation environment - type parameters or equipment status - type parameters with continuous and rapid changes, then select 6 - 10 times the change frequency of the process safety risk mode parameters; When the process safety risk mode parameters are process logic - type parameters, then select 2 - 5 times the change frequency of the process safety risk mode parameters.

7. The digital intelligent control method for multi-modal process safety risks according to claim 1, wherein, The S3 includes: S301: According to the real - time data of the currently monitored physical quantities, the established first mapping relationship and the second mapping relationship, perform data identification, parameter identification, and mode identification; S302: Based on the identification extraction results, establish an intelligent fusion model for safety risk multimodality to identify the current safety risk mode; S303: If there is or will be an abnormality in the current safety risk mode, conduct hierarchical real - time safety decision - making early warning and send risk control instructions.

8. A digital intelligent control device for multi-modal process safety risks, characterized in that, For executing the method according to any one of claims 1 - 7, including: The first construction module is used to construct the first mapping relationship between the process safety risk mode and the process safety risk mode parameters; the process safety risk mode includes: the types of process safety accidents and their corresponding safety risk factors; the process safety risk mode parameters are the characterization parameters corresponding to the safety risk factors; The second construction module is used to utilize the front-end device to sense the change of the process safety risk mode parameters to obtain the real-time data of the monitored physical quantity, and construct the second mapping relationship between the process safety risk mode parameters and the monitored physical quantity; The intelligent decision-making and control module is used to realize the functions of intelligent extraction of identification, multi-modal intelligent fusion analysis, safety mode identification, real-time hierarchical early warning and intelligent control according to the sensed current monitored physical quantity data, the first mapping relationship and the second mapping relationship.

Images