DCS fault diagnosis method and system for whole flow of fluorine chemical production line

Abstract

The invention relates to the technical field of fault diagnosis, provides a DCS fault diagnosis method and system for the whole process of a fluorine chemical production line, and solves the problems that it is difficult to represent the multi-stage nonlinear dynamic logic of the fluorine chemical production process and the fault source cannot be accurately positioned. According to the invention, the multi-source data stream output by the distributed control system is monitored in real time; identifying the current process state according to the change characteristics of the temperature data and the pressure data of the fluorination reactor; in the second process stage, coupling the concentration of the anhydrous hydrogen fluoride with the temperature of the fluorination reactor to generate a combined process characteristic; constructing an operation logic network based on the process state sequences of the first to third process stages in combination with the combined process characteristics; and traversing and analyzing the multi-source data stream, and positioning the fault source of the fluorine chemical production line. According to the technical scheme provided by the invention, the digital modeling of the full-flow process state of the fluorine chemical production line and the accurate positioning of the fault source are realized, and the accuracy and processing efficiency of fault diagnosis are also improved.

Description

Technical Field

[0001] This application relates to the field of fault diagnosis technology, and in particular to a DCS fault diagnosis method and system for the entire process of a fluorochemical production line. Background Technology

[0002] In the field of fluorochemical production, especially in the synthesis, reaction, and purification processes involving highly reactive and corrosive fluorides such as anhydrous hydrogen fluoride and aluminum fluoride, production lines are typically equipped with a distributed control system (DCS) to achieve automated operation. The DCS system monitors and collects multi-source heterogeneous data streams in real time, including reactor temperature / pressure, reactant concentration, tail gas scrubbing tower pH value, and flow rate. However, fluorochemical processes are characterized by strong nonlinearity, multi-stage dynamic transitions such as normal fluorination, aluminum fluoride crystallization induction, and emergency shutdown transitions, as well as strong coupling of multiple parameters. Faults often occur in a process-oriented, concealed, and multi-source manner, making it difficult to provide early warning and accurate location through instantaneous threshold alarms of a single parameter.

[0003] Currently, one existing technical solution employs a fault detection and diagnosis method based on multivariate statistical analysis. Specifically, this method first standardizes the multi-source data collected by the DCS, and then uses algorithms such as Principal Component Analysis (PCA) or Partial Least Squares (PLS) to establish a statistical model describing the normal operation of production. During the online monitoring phase, this method calculates statistical indicators such as Hotelling's T² and SPE in the statistical model for new sample data in real time and compares them with preset control limits. When the statistical indicators exceed the control limits, an anomaly is determined to have occurred in the process. Subsequently, methods such as contribution graph analysis are used to trace the process variables that contribute the most to the anomaly indicators, thereby assisting in determining the source of the fault.

[0004] However, this existing solution still has shortcomings when applied to fault diagnosis of the entire process of fluorochemical production lines. First, its core is a linear statistical model based on steady-state or quasi-steady-state assumptions, which is difficult to effectively characterize the complex process logic of nonlinear and multi-stage dynamic transformations in fluorochemical processes, resulting in insufficient sensitivity to detect anomalies in key nonlinear processes such as crystallization induction and state transition. Second, this method focuses on global detection of statistical deviations in multivariate data, lacking modeling of the inherent temporal correlations and causal relationships in the process flow, making it difficult to understand the propagation path of faults from the perspective of state sequences. Finally, its fault tracing relies on the calculation of the contribution of original process variables, which cannot effectively integrate and explain the coupling characteristics across process stages, such as the linkage between reactant concentration and equipment temperature, thus making it difficult to accurately locate the root cause of the fault. For example, it cannot clearly distinguish whether the fault is caused by reactor scaling or a chain reaction triggered by the failure of the downstream scrubbing system. Summary of the Invention

[0005] This application provides a DCS fault diagnosis method and system for the entire process of a fluorochemical production line, which solves the problems in the prior art that make it difficult to characterize the multi-stage nonlinear dynamic logic of the fluorochemical production process, and that it is impossible to effectively model and analyze the correlation between process stages based on multi-source data to achieve accurate location of the root cause of the fault.

[0006] Firstly, this application provides a DCS fault diagnosis method for the entire process of a fluorochemical production line, including:

[0007] The distributed control system output of the fluorochemical production line is monitored in real time. The multi-source data stream includes temperature and pressure data of the fluorination reactor, anhydrous hydrogen fluoride concentration data, and pH and flow data of the tail gas scrubbing tower.

[0008] Based on the variation characteristics of temperature and pressure data from the fluorination reactor, the current process state of the fluorination production line is identified, including the first process stage, the second process stage, and the third process stage.

[0009] When the current process state is identified as the second process stage, the anhydrous hydrogen fluoride concentration data is coupled with the temperature data of the fluorination reactor to generate a joint process feature.

[0010] Based on the process state sequence from the first process stage to the third process stage, and combined with the joint process characteristics, an operational logic network is constructed.

[0011] The multi-source data stream is traversed and analyzed using the aforementioned operational logic network to pinpoint the root cause of the fault in the fluorochemical production line.

[0012] Optionally, based on the variation characteristics of temperature and pressure data from the fluorination reactor, the current process state of the fluorochemical production line is identified. The current process state includes a first process stage, a second process stage, and a third process stage, including:

[0013] The temperature and pressure data of the fluorination reactor during a first preset time period are obtained as the first combined data.

[0014] Calculate the rate of change of temperature data in the first combination of data, and use it as the first rate of change;

[0015] Calculate the rate of change of pressure data in the first combination of data, and use it as the second rate of change.

[0016] When the duration of the first rate of change being within a first preset rate range and the second rate of change being within a second preset rate range reaches a first preset time length, the current process state is identified as the first process stage.

[0017] After being identified as the first process stage, if the first rate of change moves from the first preset rate range to the third preset rate range, and the second rate of change moves from the second preset rate range to the fourth preset rate range, then the current process state is identified and the process enters the second process stage.

[0018] If, during the duration of the second process stage, the first rate of change is greater than the fifth preset rate range, the current process state is identified and the process enters the third process stage.

[0019] Optionally, when the current process state is identified as the second process stage, the anhydrous hydrogen fluoride concentration data is coupled with the temperature data of the fluorination reactor to generate joint process features, including:

[0020] When the current process state is identified as the second process stage, the anhydrous hydrogen fluoride concentration sequence and the fluorination reactor tube wall temperature sequence corresponding to the second process stage in time are obtained.

[0021] The anhydrous hydrogen fluoride concentration sequence is processed to extract the fluctuation period of the anhydrous hydrogen fluoride concentration sequence, which is used as the concentration fluctuation period.

[0022] The temperature sequence of the tube wall of the fluorination reactor is processed, and the decreasing slope of the temperature sequence is extracted as the decreasing slope of the tube wall temperature.

[0023] Calculate the reciprocal of the concentration fluctuation period to obtain the concentration fluctuation frequency;

[0024] Calculate the absolute value of the slope of the temperature drop in the pipe wall;

[0025] The concentration fluctuation frequency is multiplied by the absolute value of the slope of the pipe wall temperature drop, and the product is used as the joint process feature.

[0026] Optionally, based on the process state sequence from the first process stage to the third process stage, and in conjunction with the joint process characteristics, an operational logic network is constructed, including:

[0027] Map the first process stage to a first process stage node, map the second process stage to a second process stage node, and map the third process stage to a third process stage node.

[0028] Establish a first connection relationship from the first process stage node to the second process stage node, wherein the first connection relationship has a first direction, and the first direction indicates the transition from the first process stage to the second process stage;

[0029] Establish a second connection relationship from the second process stage node to the third process stage node, wherein the second connection relationship has a second direction, the second direction indicating the transition from the second process stage to the third process stage;

[0030] Obtain the combined process features;

[0031] The numerical value of the joint process feature is assigned to the first connection relationship as the attribute value of the first connection relationship;

[0032] A runtime logic network is constructed based on the first process stage node, the second process stage node, the third process stage node, and the first and second connection relationships with the attribute values ​​attached.

[0033] Optionally, based on the first process stage node, the second process stage node, the third process stage node, and the first connection relationship and the second connection relationship with the attribute values ​​attached, an operational logic network is constructed, including:

[0034] The first process stage node, the second process stage node, and the third process stage node are considered as a set of nodes in the running logic network.

[0035] The first connection relationship and the second connection relationship are used as the set of connection relationships in the running logical network;

[0036] The attribute value of the first connection relationship is used as the weight value of the first connection relationship;

[0037] Set a preset weight value for the second connection relationship;

[0038] Based on the set of nodes, the set of connection relationships, the weight value of the first connection relationship, and the preset weight value of the second connection relationship, the topology of the running logic network is generated;

[0039] Mark the first process stage node as the starting node of the running logic network;

[0040] The third process stage node is marked as the termination node of the running logic network;

[0041] The starting node, the ending node, and the topology are combined to construct a running logical network.

[0042] Optionally, the multi-source data stream is traversed and parsed using the operational logic network to locate the root cause of the fault in the fluorochemical production line, including:

[0043] Based on the topology of the running logic network, the traversal path from the starting node to the ending node is determined.

[0044] Based on the changing trends of each data in the multi-source data stream, the current process stage node of the fluorochemical production line is identified.

[0045] Starting from the current process stage node, traverse along the traversal path towards the termination node;

[0046] During the traversal, obtain the weight value of the current connection relationship;

[0047] Based on the matching relationship between the weight value and the corresponding data in the multi-source data stream, it is determined whether the current process status is abnormal.

[0048] If the current process status is determined to be abnormal, the root cause of the fault in the fluorochemical production line is determined based on the process stage node corresponding to the current connection relationship.

[0049] Optionally, based on the matching relationship between the weight value and the corresponding data in the multi-source data stream, determining whether the current process state is abnormal includes:

[0050] Obtain the weight value of the current connection relationship;

[0051] Extract process parameter data corresponding to the current connection relationship from the multi-source data stream;

[0052] Based on the correlation between the feature values ​​and weight values ​​of the process parameter data, a matching degree index is calculated.

[0053] The matching degree index is compared with a preset threshold.

[0054] Determine whether the current process status is abnormal based on the comparison results.

[0055] Secondly, this application provides a DCS fault diagnosis system for the entire process of a fluorochemical production line, including:

[0056] The monitoring module is used to monitor the multi-source data stream output by the distributed control system of the fluorochemical production line in real time. The multi-source data stream includes temperature and pressure data of the fluorination reactor, anhydrous hydrogen fluoride concentration data, and pH and flow data of the tail gas scrubbing tower.

[0057] The identification module is used to identify the current process state of the fluorochemical production line based on the change characteristics of the temperature and pressure data of the fluorination reactor. The current process state includes a first process stage, a second process stage, and a third process stage.

[0058] The coupling module is used to couple the anhydrous hydrogen fluoride concentration data with the temperature data of the fluorination reactor when the current process state is identified as the second process stage, to generate a joint process feature.

[0059] A construction module is used to construct an operational logic network based on the process state sequence from the first process stage to the third process stage, combined with the joint process features;

[0060] The positioning module is used to traverse and parse the multi-source data stream using the operating logic network to locate the root cause of the fault in the fluorochemical production line.

[0061] Thirdly, this application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement a DCS fault diagnosis method for the entire process of a fluorochemical production line as described in the first aspect above.

[0062] Fourthly, this application provides a computer storage medium storing a computer program, which, when executed by a computer, implements a DCS fault diagnosis method for the entire process of a fluorochemical production line as described in the first aspect.

[0063] The beneficial effects of this application are:

[0064] This application utilizes an operational logic network constructed based on the multi-stage process state sequence of fluorochemicals to effectively describe the dynamic evolution logic of the entire process from the first process stage, the second process stage to the third process stage. Specifically, the process state sequence network constructed based on multi-source data streams can structurally characterize the nonlinear state transition relationships of the fluorochemical production line at different process stages, overcoming the limitations of existing multivariate statistical analysis methods in effectively modeling the inherent nonlinear temporal correlations of the process flow. This results in higher sensitivity for state identification and anomaly detection for key nonlinear process processes such as aluminum fluoride crystallization induction and emergency shutdown transition.

[0065] Furthermore, in the process of constructing the operational logic network, the process state sequence is combined with the joint process features generated in the second process stage, realizing the fusion of process logic across stages and data sources. This enables the matching degree analysis based on the feature values ​​of process parameter data and the weight values ​​of network connection relationships when using the operational logic network to traverse and parse multi-source data streams. Fault diagnosis is performed from the dual dimensions of global state evolution and local feature association, thereby accurately distinguishing whether the fault originates from scaling in the fluorination reactor or from a chain reaction caused by an abnormality in the downstream tail gas scrubbing system.

[0066] These or other aspects of this application will become more apparent from the description of the following embodiments. Attached Figure Description

[0067] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0068] Figure 1 A flowchart of a DCS fault diagnosis method for the entire process of a fluorochemical production line provided in this application is shown.

[0069] Figure 2 This application provides a schematic diagram of the structure of a DCS fault diagnosis system for the entire process of a fluorochemical production line.

[0070] Figure 3 A schematic diagram of the structure of a computing device provided in this application is shown. Detailed Implementation

[0071] To enable those skilled in the art to better understand the present application, the technical solution of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0072] In some of the processes described in the specification, claims, and accompanying drawings of this application, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 101, 102, etc., are merely used to distinguish different operations and do not themselves represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a chronological order, nor do they limit "first" and "second" to different types.

[0073] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0074] Figure 1 This application provides a flowchart of a DCS fault diagnosis method for the entire process of a fluorochemical production line, as shown in the following figure. Figure 1 As shown, the method includes:

[0075] Step 101: Monitor the multi-source data stream output by the distributed control system of the fluorochemical production line in real time. The multi-source data stream includes temperature and pressure data of the fluorination reactor, anhydrous hydrogen fluoride concentration data, and pH and flow data of the tail gas scrubbing tower.

[0076] In this step, the distributed control system refers to an automated system used for industrial process control, which is used to achieve centralized monitoring and management of the production line. It is achieved by multiple control units being distributed and working collaboratively.

[0077] Multi-source data streams refer to continuous data sequences composed of various types of data from different monitoring devices and sensors, used to provide comprehensive information on the operating status of the production line, obtained by collecting parameters such as temperature, pressure, concentration, pH value, and flow rate.

[0078] Temperature data refers to the measured values ​​reflecting the reaction temperature inside the fluorination reactor. It is used to monitor the thermodynamic state of the reaction and is acquired in real time by temperature sensors.

[0079] Pressure data refers to the measured values ​​reflecting the internal pressure of the fluorination reactor, used to monitor the reaction kinetics, and is acquired in real time by pressure sensors.

[0080] The anhydrous hydrogen fluoride concentration data refers to the measured mass fraction of anhydrous hydrogen fluoride in the reaction products, which is used to evaluate the reaction conversion efficiency and is obtained in real time through an online analyzer.

[0081] The pH value data of the exhaust gas scrubbing tower refers to the measured value of acidity and alkalinity in the scrubbing liquid, which is used to monitor the exhaust gas treatment effect and is obtained in real time through pH meter.

[0082] Flow data refers to the measured flow rate of the washing liquid in the washing tower, which is used to evaluate the operating status of the washing system and is obtained through real-time monitoring by a flow meter.

[0083] In this step, temperature data of the fluorination reactor is first collected in real time by temperature sensors deployed on the fluorochemical production line, and pressure data of the fluorination reactor is collected in real time by pressure sensors. Next, the concentration of anhydrous hydrogen fluoride in the reaction products is measured in real time by an online analyzer. Simultaneously, the pH value of the scrubbing liquid is monitored in real time by a pH meter installed in the tail gas scrubbing tower, and the flow rate of the scrubbing liquid is monitored in real time by a flow meter. Then, the collected temperature, pressure, anhydrous hydrogen fluoride concentration, pH, and flow rate data are centrally received and stored by the data acquisition module of the distributed control system, forming a multi-source data stream. Finally, the multi-source data stream is transmitted to the fault diagnosis system for further processing.

[0084] For example, in a fluorochemical company in location A, temperature and pressure data of the fluorination reactor are first collected once per second by temperature and pressure sensors installed on the reactor. At the same time, an online analyzer measures the concentration of anhydrous hydrogen fluoride in the reaction product every 30 seconds. Then, pH and flow meters in the tail gas scrubbing tower collect pH and flow data of the scrubbing liquid every minute. Finally, all the collected data are aggregated by the data acquisition module of the distributed control system to form a multi-source data stream containing five dimensions: temperature, pressure, concentration, pH, and flow, and transmitted in real time to the fault diagnosis server in the central control room.

[0085] Step 102: Based on the variation characteristics of the temperature and pressure data of the fluorination reactor, identify the current process state of the fluorination production line. The current process state includes the first process stage, the second process stage, and the third process stage.

[0086] Optionally, step 102 may specifically include:

[0087] Step 1021: Obtain the temperature and pressure data of the fluorination reactor within a first preset time period, as the first combined data.

[0088] Step 1022: Calculate the rate of change of temperature data in the first combination of data, and use it as the first rate of change.

[0089] Step 1023: Calculate the rate of change of pressure data in the first combination of data, as the second rate of change.

[0090] Step 1024: When the duration of the first change rate being within the first preset rate range and the second change rate being within the second preset rate range reaches the first preset time length, the current process state is identified as the first process stage.

[0091] Step 1025: After identifying the first process stage, if the first rate of change moves from the first preset rate range to the third preset rate range, and the second rate of change moves from the second preset rate range to the fourth preset rate range, then the current process state is identified and the process enters the second process stage.

[0092] Step 1026: If, during the duration of the second process stage, the first rate of change is greater than the fifth preset rate range, the current process state is identified and the third process stage is entered.

[0093] In this step, the current process status refers to the specific process stage that the fluorochemical production line is in during operation, and is used to characterize the dynamic changes in the production process.

[0094] The first process stage refers to the normal fluorination reaction stage of the fluorochemical production line, which is used to characterize the process state of stable changes in reactor temperature and pressure.

[0095] The second process stage refers to the aluminum fluoride crystallization induction stage, which is used to characterize the process state in which the reactor temperature and pressure begin to deviate from the normal range. It is identified by the rate of temperature change and the rate of pressure change as they move from the first preset range to the second preset range.

[0096] The third process stage refers to the emergency shutdown transition stage, which is used to characterize the process state of rapid changes in reactor temperature and pressure. It is identified by the rate of temperature change exceeding a preset range.

[0097] The first preset time period refers to a continuous time period used to calculate the rate of temperature change and the rate of pressure change. It is used to obtain enough data for rate calculation and is set through configuration.

[0098] The first set of data refers to the collection of temperature and pressure data collected within a first preset time period, used to calculate the rate of temperature change and the rate of pressure change.

[0099] The first rate of change refers to how quickly the temperature data changes over time, and is used to reflect the drastic degree of temperature change in the reactor.

[0100] The second rate of change refers to how quickly the pressure data changes over time, and is used to reflect the drastic degree of pressure change in the reactor.

[0101] The first preset rate range refers to the reasonable range of temperature change rate during the normal fluorination stage, used to determine whether it is in the first process stage. The preset rate range and preset time length below are all obtained through historical data and process parameter settings.

[0102] The second preset rate range refers to the reasonable range of pressure change rate during the normal fluorination stage, used to determine whether it is in the first process stage.

[0103] The first preset time length refers to the shortest duration for which the rate of temperature and pressure change needs to be kept within a preset range, used to confirm that the process status is stable.

[0104] The third preset rate range refers to the range of temperature change rates during the aluminum fluoride crystallization induction stage, used to determine whether to enter the second process stage.

[0105] The fourth preset rate range refers to the range of pressure change rate during the aluminum fluoride crystallization induction stage, used to determine whether to enter the second process stage.

[0106] The fifth preset rate range refers to the range of temperature change rate during the emergency shutdown transition phase, used to determine whether to enter the third process stage.

[0107] In this step, firstly, temperature and pressure data of the fluorination reactor within a first preset time period are extracted from a multi-source data stream to form a first combined data set. Secondly, the temperature data in the first combined data set is processed using a differential algorithm to calculate the difference between temperature data at adjacent time points, and then divided by the time interval to obtain a first rate of change. Simultaneously, the pressure data in the first combined data set is processed using a differential algorithm to calculate the difference between pressure data at adjacent time points, and then divided by the time interval to obtain a second rate of change. Next, a comparator compares the first rate of change with a first preset rate range, and compares the second rate of change with a second preset rate range. When the first rate of change is within the first preset rate range and the second rate of change is within the second preset rate range, the first rate of change is determined. When the duration of the rate of change within the second preset rate range reaches the first preset time length, the current process state is determined to be the first process stage. Then, after the first process stage is identified, the changing trends of the first and second rates of change are continuously monitored by a monitor. When the first rate of change moves from the first preset rate range to the third preset rate range, and the second rate of change moves from the second preset rate range to the fourth preset rate range, the current process state is determined to have entered the second process stage. Finally, during the second process stage, the first rate of change is compared with the fifth preset rate range by a comparator. When the first rate of change is greater than the fifth preset rate range, the current process state is determined to have entered the third process stage.

[0108] For example, following the specific implementation of the previous step, in a fluorochemical enterprise in location A, the temperature and pressure data within the last 5 minutes are first extracted from a multi-source data stream as the first combined data. Next, a differential algorithm is used to calculate the rate of change of the temperature and pressure data, resulting in a first rate of change of 0.5 degrees Celsius per minute and a second rate of change of 0.1 MPa per minute. Since the first rate of change falls within a first preset rate range of 0.3 to 0.8 degrees Celsius per minute, and the second rate of change falls within a second preset rate range of 0.05 to 0.15 MPa per minute, and the duration has reached a first preset time length of 3 minutes, the current process state is determined to be the first process stage. Then, the temperature rate of change gradually increases to 1.2 degrees Celsius per minute, entering the third preset rate range, while the pressure rate of change becomes 0.2 MPa per minute, entering the fourth preset rate range, and the current process state is determined to have entered the second process stage. Finally, after 2 minutes in the second process stage, the temperature rate of change suddenly increases to 3.5 degrees Celsius per minute, exceeding the fifth preset rate range, and the current process state is determined to have entered the third process stage.

[0109] This step, by analyzing the changing characteristics of temperature and pressure data from the fluorination reactor, can accurately identify the process state of the fluorochemical production line, including the normal fluorination stage, the aluminum fluoride crystallization induction stage, and the emergency shutdown transition stage. This enables real-time monitoring and status identification of dynamic changes in the production process, providing accurate process status information for subsequent joint process feature generation and fault diagnosis. At the same time, by setting multiple preset rate ranges and duration thresholds, different process stages can be effectively distinguished, avoiding misjudgments and omissions, and improving the reliability of status identification.

[0110] Step 103: When the current process state is identified as the second process stage, the anhydrous hydrogen fluoride concentration data is coupled with the temperature data of the fluorination reactor to generate a joint process feature.

[0111] Optionally, step 103 may specifically include:

[0112] Step 1031: When the current process state is identified as the second process stage, obtain the anhydrous hydrogen fluoride concentration sequence and the fluorination reactor tube wall temperature sequence corresponding to the second process stage in time.

[0113] Step 1032: Process the anhydrous hydrogen fluoride concentration sequence and extract the fluctuation period of the anhydrous hydrogen fluoride concentration sequence as the concentration fluctuation period.

[0114] Step 1033: Process the temperature sequence of the fluorination reactor tube wall and extract the descending slope of the temperature sequence as the tube wall temperature descending slope.

[0115] Step 1034: Calculate the reciprocal of the concentration fluctuation period to obtain the concentration fluctuation frequency.

[0116] Step 1035: Calculate the absolute value of the slope of the tube wall temperature decrease.

[0117] Step 1036: Multiply the concentration fluctuation frequency with the absolute value of the slope of the tube wall temperature drop to obtain the product as a joint process feature.

[0118] In this step, the combined process characteristics refer to the characteristic values ​​that reflect the correlation between changes in anhydrous hydrogen fluoride concentration and changes in the temperature of the fluorination reactor, and are used to characterize the coupling relationship between reactant concentration and equipment temperature in the second process stage.

[0119] The anhydrous hydrogen fluoride concentration sequence refers to a sequence of anhydrous hydrogen fluoride concentration data collected continuously during the second process stage, used to analyze the concentration change over time.

[0120] The fluorination reactor tube wall temperature sequence refers to a sequence of tube wall temperature data continuously collected during the second process stage, used to analyze the variation of tube wall temperature over time.

[0121] The concentration fluctuation period refers to the time interval between adjacent peaks or troughs in the anhydrous hydrogen fluoride concentration sequence, used to reflect the periodic characteristics of concentration changes.

[0122] The tube wall temperature decrease slope refers to the rate at which the temperature decreases over time in the tube wall temperature sequence of the fluorination reactor. It is used to reflect the decreasing trend of the tube wall temperature and is obtained by linear fitting of the tube wall temperature sequence of the fluorination reactor.

[0123] Concentration fluctuation frequency refers to the number of times the concentration of anhydrous hydrogen fluoride fluctuates per unit time, and is used to reflect the speed of concentration change.

[0124] The absolute value refers to the magnitude of the value regardless of the sign. It is used to eliminate the influence of the negative sign on the slope of the pipe wall temperature drop and is obtained through mathematical calculations.

[0125] In this step, firstly, when the current process state is identified as the second process stage, anhydrous hydrogen fluoride concentration data completely corresponding to the time period of this second process stage is extracted from the multi-source data stream through time alignment processing to form an anhydrous hydrogen fluoride concentration sequence. Simultaneously, the fluorination reactor tube wall temperature data for the same time period is extracted to form a fluorination reactor tube wall temperature sequence. Secondly, the anhydrous hydrogen fluoride concentration sequence is transformed from the time domain to the frequency domain using a fast Fourier transform algorithm. Then, the dominant frequency component is identified in the frequency domain, and the concentration fluctuation period is calculated using the conversion relationship between frequency and period. Next, the fluorination reactor tube wall temperature... The temperature sequence of the tube wall in the fluorination reactor was analyzed by linear regression using the least squares method to fit a straight line that best represents the temperature change trend. The slope parameter of the tube wall temperature drop was extracted from the slope parameter of this line. Then, the concentration fluctuation period was calculated by dividing the value by the concentration fluctuation period using the reciprocal formula. At the same time, the absolute value of the tube wall temperature drop slope was obtained by removing its negative sign using the absolute value function. Finally, the calculated concentration fluctuation frequency and the absolute value of the tube wall temperature drop slope were multiplied together to obtain the product, which was used as the joint process characteristic.

[0126] For example, following the specific implementation of the previous step, in a fluorochemical enterprise in location A, after identifying the current process state as the second process stage, the anhydrous hydrogen fluoride concentration data for the most recent 10 minutes is extracted from a multi-source data stream to form an anhydrous hydrogen fluoride concentration sequence. Simultaneously, the fluorination reactor tube wall temperature data for the same time period is extracted to form a fluorination reactor tube wall temperature sequence. Next, a fast Fourier transform algorithm is used to perform spectral analysis on the anhydrous hydrogen fluoride concentration sequence. By identifying the main peak position in the spectrum, the concentration fluctuation period is calculated to be 2 minutes. Then, a minimum... The quadratic method was used to linearly fit the temperature sequence of the fluorination reactor tube wall. By calculating the slope of the fitted line, the slope of the tube wall temperature decrease was found to be 0.8 degrees per minute. Then, by performing a reciprocal operation, 1 was divided by the concentration fluctuation period of 2 minutes to obtain a concentration fluctuation frequency of 0.5 times per minute. The negative value of the tube wall temperature decrease slope was processed by absolute value operation to obtain an absolute value of 0.8. Finally, the concentration fluctuation frequency of 0.5 was multiplied by the absolute value of the tube wall temperature decrease slope of 0.8 to obtain a product result of 0.4. This product result of 0.4 was determined as the joint process characteristic.

[0127] This step couples anhydrous hydrogen fluoride concentration data with fluorination reactor temperature data in the second process stage to generate joint process features. This effectively characterizes the correlation between reactant concentration changes and equipment temperature changes. At the same time, the extracted concentration fluctuation period and pipe wall temperature drop slope reflect key dynamic features in the fluorochemical production process. Through mathematical operations, these features are integrated into joint process features, providing important feature information for the subsequent construction of the operational logic network.

[0128] Step 104: Based on the process state sequence from the first process stage to the third process stage, and in conjunction with the joint process characteristics, construct the operation logic network.

[0129] Optionally, step 104 may specifically include:

[0130] Step 1041: Map the first process stage to a first process stage node, map the second process stage to a second process stage node, and map the third process stage to a third process stage node.

[0131] Step 1042: Establish a first connection relationship from the first process stage node to the second process stage node, wherein the first connection relationship has a first direction, and the first direction indicates the transition from the first process stage to the second process stage.

[0132] Step 1043: Establish a second connection relationship from the second process stage node to the third process stage node, wherein the second connection relationship has a second direction, and the second direction indicates the transition from the second process stage to the third process stage.

[0133] Step 1044: Obtain the combined process features.

[0134] Step 1045: Assign the value of the joint process feature to the first connection relationship as the attribute value of the first connection relationship.

[0135] Step 1046: Construct an operational logic network based on the first process stage node, the second process stage node, the third process stage node, and the first connection relationship and the second connection relationship with the attribute values ​​attached.

[0136] Optionally, step 1046 may include the following steps: using the first process stage node, the second process stage node, and the third process stage node as a set of nodes in the running logic network; using the first connection relationship and the second connection relationship as a set of connection relationships in the running logic network; using the attribute value of the first connection relationship as the weight value of the first connection relationship; setting a preset weight value for the second connection relationship; generating a topology of the running logic network based on the node set, the set of connection relationships, the weight value of the first connection relationship, and the preset weight value of the second connection relationship; marking the first process stage node as the starting node of the running logic network; marking the third process stage node as the ending node of the running logic network; and combining the starting node, the ending node, and the topology to construct the running logic network.

[0137] In this step, the process state sequence refers to the sequential arrangement from the first process stage to the second process stage and then to the third process stage. It is used to describe the evolution of the process state of the fluorochemical production line and is obtained through real-time monitoring and state identification.

[0138] The operational logic network refers to a network structure composed of process stage nodes and connections, used to characterize the process logic and state transition relationships of a fluorochemical production line.

[0139] The first process stage node refers to the network node representing the first process stage, used to indicate the process status of the normal fluorination stage.

[0140] The second process stage node refers to the network node representing the second process stage, used to indicate the process status of the aluminum fluoride crystallization induction stage.

[0141] The third process stage node refers to the network node representing the third process stage, used to indicate the process status during the emergency shutdown transition phase.

[0142] The first connection relationship refers to the directed connection from the first process stage node to the second process stage node, which is used to represent the state transition relationship from the normal fluorination stage to the aluminum fluoride crystallization induction stage. It is obtained by establishing a directed connection between the two nodes.

[0143] The first direction refers to the direction of the first connection relationship, which is used to indicate the direction of the process state transition.

[0144] The second connection relationship refers to the directed connection from the second process stage node to the third process stage node. It is used to represent the state transition relationship from the aluminum fluoride crystallization induction stage to the emergency shutdown transition stage, and is obtained by establishing a directed connection between two nodes.

[0145] The second direction refers to the direction in which the second connection relationship points, and is used to indicate the direction of process state transition.

[0146] Attribute values ​​refer to the numerical values ​​of the characteristics assigned to the connection relationship, which are used to characterize the strength or importance of the connection relationship. They are obtained by assigning the numerical values ​​of the joint process characteristics to the first connection relationship.

[0147] A node set refers to the collection of all nodes in a running logical network, used to store and manage node information in the network.

[0148] The connection set refers to the collection of all connection relationships in the running logical network, used to store and manage connection relationship information in the network.

[0149] The weight value refers to the quantitative value of the importance of the connection relationship, which is used to reflect the probability or risk of process state transformation. It is obtained by using the attribute value of the first connection relationship as the weight value.

[0150] The preset weight value refers to the pre-set weight value of the connection relationship, which is used to indicate the importance of the second connection relationship. It is obtained through configuration settings.

[0151] Topology refers to the connection method of nodes and connections in a logical network. It is used to describe the overall architecture of the network and is obtained by combining the set of nodes and the set of connections.

[0152] The starting node refers to the initial node of the running logic network, used to represent the initial stage of the process state.

[0153] Termination nodes refer to the end nodes of the running logic network, used to represent the final stage of the process state.

[0154] In this step, firstly, the first, second, and third process stages in the process state sequence are processed separately using a node mapping algorithm, mapping each process stage to a corresponding network node, resulting in first, second, and third process stage nodes. Secondly, a directed graph construction algorithm is used to establish a directed connection between the first and second process stage nodes, setting the connection direction to be from the first process stage node to the second process stage node, resulting in a first connection relationship with a first direction. Simultaneously, a directed graph construction algorithm is used to establish a directed connection between the second and third process stage nodes, setting the connection direction to be from the second process stage node to the third process stage node, resulting in a second connection relationship with a second direction. The values ​​of the joint process features are then obtained from the joint process feature storage location through a data acquisition interface. Finally, attribute assignment processing is performed... The obtained joint process feature values ​​are assigned to the first connection relationship as its attribute value. Then, a network construction algorithm is used to aggregate the nodes of the first, second, and third process stages to form a node set. The first and second connection relationships are then aggregated to form a connection relationship set. The attribute values ​​of the first connection relationships are processed and converted into weight values. The second connection relationships are processed and assigned preset weight values. A topology is generated using graph theory based on the node set, connection relationship set, the weight values ​​of the first and second connection relationships, and the preset weight values ​​of the second connection relationships. Finally, node labeling is used to mark the first process stage nodes as start nodes and the third process stage nodes as end nodes. The start nodes, end nodes, and topology are combined through structural combination processing to ultimately construct the operational logic network.

[0155] For example, following the specific implementation of the previous step, in a fluorochemical enterprise in location A, firstly, a node mapping algorithm is used to map the first process stage to node N1, the second process stage to node N2, and the third process stage to node N3; secondly, a directed graph construction algorithm is used to establish a directed connection E1 between node N1 and node N2, with the direction set from N1 to N2, and a directed connection E2 is established between node N2 and node N3, with the direction set from N2 to N3; then, the joint process feature value of 0.4 is obtained from the memory through a data acquisition interface, and the attribute assignment process is performed to... The joint process feature value of 0.4 is assigned as an attribute value to connection E1. Then, the network construction algorithm is used to aggregate nodes N1, N2, and N3 into a node set, and connections E1 and E2 into a connection relationship set. The attribute value of 0.4 for connection E1 is converted into its weight value, and a preset weight value of 0.6 is set for connection E2. Finally, the graph theory method is used to generate a topology based on the node set and the connection relationship set. Node labeling is used to mark node N1 as the starting node and node N3 as the ending node. Then, the starting node, ending node, and topology are combined through structural combination processing to complete the construction of the running logic network.

[0156] This step, by constructing an operational logic network, achieves digital modeling of the entire process state of the fluorochemical production line. It clearly characterizes the complete process logic and state transition relationship from the normal fluorination stage to the emergency shutdown transition stage. At the same time, by using joint process feature values ​​as connection weights, it quantifies the possibility of process state transitions, providing a reliable structured framework for subsequent fault diagnosis and improving the accuracy of fault location.

[0157] Step 105: Use the running logic network to traverse and parse the multi-source data stream to locate the root cause of the fault in the fluorochemical production line.

[0158] Optionally, step 105 may specifically include:

[0159] Step 1051: Based on the topology of the running logic network, determine the traversal path from the starting node to the ending node.

[0160] Step 1052: Identify the current process stage node of the fluorochemical production line based on the changing trends of each data in the multi-source data stream.

[0161] Step 1053: Starting from the current process stage node, traverse along the traversal path towards the termination node.

[0162] Step 1054: During the traversal, obtain the weight value of the current connection relationship.

[0163] Step 1055: Based on the matching relationship between the weight value and the corresponding data in the multi-source data stream, determine whether the current process status is abnormal.

[0164] Optionally, step 1055 may include the following steps: obtaining the weight value of the current connection relationship; extracting process parameter data corresponding to the current connection relationship from the multi-source data stream; calculating a matching degree index based on the correlation between the feature value of the process parameter data and the weight value; comparing the matching degree index with a preset threshold; and determining whether the current process state is abnormal based on the comparison result.

[0165] Step 1056: If the current process status is determined to be abnormal, the root cause of the fluorochemical production line fault is determined based on the process stage node corresponding to the current connection relationship.

[0166] In this step, the root cause of the fluorochemical production line failure refers to the location or link that is the fundamental cause of the abnormal operation of the production line. It is used to guide troubleshooting and maintenance decisions and is obtained through traversal and analysis of the operating logic network.

[0167] A traversal path refers to an ordered sequence of nodes and connections from the starting node to the ending node, used to guide the sequential process of fault diagnosis.

[0168] The trend of change refers to the direction and magnitude of the change in process parameter data over time, which is used to determine the development direction of the process status. It is obtained by performing trend analysis on multi-source data streams.

[0169] Process parameter data refers to process variable data related to a specific process stage, used to evaluate the operating status of that process stage, and is obtained by extracting corresponding data from multi-source data streams.

[0170] Eigenvalues ​​are numerical values ​​extracted from process parameter data that reflect its characteristics and are used to characterize the degree of abnormality in the process state.

[0171] The matching degree index refers to the quantitative value of the degree of matching between the weight value and the feature value of the process parameter data, which is used to determine whether the process status is abnormal.

[0172] The preset threshold is the critical value for judging whether the matching degree index is abnormal. It is used to distinguish between normal and abnormal states and is set through historical data and process requirements.

[0173] In this step, firstly, the topology of the running logical network is analyzed using a depth-first search algorithm. Starting from the initial node, all possible paths are recursively explored along the connection directions to ultimately determine a traversal path from the initial node to the final node. Secondly, a time series analysis algorithm is used to analyze the trends of temperature, pressure, anhydrous hydrogen fluoride concentration, pH, and flow rate data from the multi-source data stream. The changing trends of each data point are calculated using a polynomial fitting method. Then, the calculated trends are compared and matched with the characteristics of the process stage nodes to identify the current process stage node of the fluorochemical production line. Finally, the identified current process stage node is used as the basis for further analysis. Starting from the current node, a path tracing algorithm is used to traverse the path towards the termination node. During the traversal, the weight values ​​of the current connection relationship are obtained through a weight extraction interface. Then, data filtering is used to extract process parameter data corresponding to the current connection relationship from multi-source data streams. A feature extraction algorithm is used to perform statistical analysis on the process parameter data to extract feature values. The Pearson correlation coefficient calculation method is then used to calculate the matching degree index based on the correlation between feature values ​​and weight values. A threshold comparison algorithm is used to compare the matching degree index with a preset threshold for judgment. Finally, a fault location algorithm is used to determine the root cause of the fault in the fluorochemical production line based on the process stage node corresponding to the current connection relationship.

[0174] For example, following the specific implementation of the previous step, in a fluorochemical enterprise in location A, the topology of the operating logic network is first analyzed using a depth-first search algorithm to determine the traversal path as node N1 → connection E1 → node N2 → connection E2 → node N3. Next, a time series analysis algorithm is used to perform trend analysis on the multi-source data stream, calculating that the temperature data shows an upward trend and the pressure data shows a downward trend. Node matching is then used to identify the current process stage node as N2. Then, starting from node N2, a path tracing algorithm is used to traverse along the traversal path towards node N3. When traversing connection E2, a weighted algorithm is used to determine the connection path. The interface is re-extracted to obtain the preset weight value of 0.6 for connection E2; then, anhydrous hydrogen fluoride concentration data and pH value data are extracted from the multi-source data stream through data filtering processing. According to the feature extraction algorithm, the concentration fluctuation frequency is calculated to be 0.7 times per minute, and the pH value decrease slope is 0.3 times per minute. Subsequently, the matching degree index is calculated to be 0.42 using the Pearson correlation coefficient calculation method. Then, the threshold comparison algorithm is used to compare with the preset threshold of 0.5 to determine that the current process status is abnormal; finally, the fault location algorithm is used to determine that the root cause of the fault is the abnormal anhydrous hydrogen fluoride concentration caused by scaling in the fluorination reactor based on the process stage nodes N2 and N3 corresponding to connection E2.

[0175] This step utilizes a running logic network to traverse and parse multi-source data streams, enabling a comprehensive diagnosis of the operating status of the fluorochemical production line according to the process logic sequence. By analyzing the matching relationship between weight values ​​and process parameter data, it accurately identifies abnormalities in the process status. Simultaneously, based on the process stage nodes corresponding to the traversal path and connection relationships, it precisely locates the root cause of the fault, providing a reliable basis for quickly eliminating faults. This not only achieves intelligent fault diagnosis for the entire fluorochemical production line process but also improves the efficiency and accuracy of fault handling.

[0176] Figure 2 This application provides a schematic diagram of the structure of a DCS fault diagnosis system for the entire process of a fluorochemical production line, as shown in the following figure. Figure 2 As shown, the system includes:

[0177] Monitoring module 21 is used to monitor in real time the multi-source data stream output by the distributed control system of the fluorochemical production line. The multi-source data stream includes temperature and pressure data of the fluorination reactor, anhydrous hydrogen fluoride concentration data, and pH and flow data of the tail gas scrubbing tower.

[0178] The identification module 22 is used to identify the current process state of the fluorochemical production line based on the change characteristics of the temperature and pressure data of the fluorination reactor. The current process state includes a first process stage, a second process stage, and a third process stage.

[0179] The coupling module 23 is used to couple the anhydrous hydrogen fluoride concentration data with the temperature data of the fluorination reactor when the current process state is identified as the second process stage, to generate a joint process feature.

[0180] Construction module 24 is used to construct an operational logic network based on the process state sequence from the first process stage to the third process stage, combined with the joint process features;

[0181] The positioning module 25 is used to traverse and parse the multi-source data stream using the operating logic network in order to locate the root cause of the fault in the fluorochemical production line.

[0182] Figure 2 The aforementioned DCS fault diagnosis system for the entire process of a fluorochemical production line can perform... Figure 1 The implementation principle and technical effects of the DCS fault diagnosis method for the entire process of a fluorochemical production line described in the illustrated embodiment will not be repeated here. The specific operation methods of each module and unit in the DCS fault diagnosis system for the entire process of a fluorochemical production line in the above embodiments have been described in detail in the embodiments related to this method, and will not be elaborated upon here.

[0183] In one possible design, Figure 2The DCS fault diagnosis system for the entire process of a fluorochemical production line shown in the embodiment can be implemented as a computing device, such as... Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0184] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by the processing component 32.

[0185] The processing component 32 is used for the above Figure 1 The embodiment describes a DCS fault diagnosis method for the entire process of a fluorochemical production line.

[0186] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above-described method. Alternatively, the processing component may be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method.

[0187] Storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0188] Of course, computing devices may also include other components, such as input / output interfaces, display components, communication components, etc.

[0189] Input / output interfaces provide interfaces between processing components and peripheral interface modules, which can be output devices, input devices, etc.

[0190] The communication components are configured to facilitate wired or wireless communication between computing devices and other devices.

[0191] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server, and the aforementioned processing components, storage components, etc., can be basic server resources rented or purchased from the cloud computing platform.

[0192] This application also provides a computer storage medium storing a computer program, which, when executed by a computer, can perform the above-described functions. Figure 1 The embodiment shown illustrates a DCS fault diagnosis method for the entire process of a fluorochemical production line.

[0193] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0194] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0195] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0196] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A DCS fault diagnosis method for the entire process of a fluorochemical production line, characterized in that, include: The distributed control system output of the fluorochemical production line is monitored in real time. The multi-source data stream includes temperature and pressure data of the fluorination reactor, anhydrous hydrogen fluoride concentration data, and pH and flow data of the tail gas scrubbing tower. Based on the variation characteristics of temperature and pressure data from the fluorination reactor, the current process state of the fluorination production line is identified, including the first process stage, the second process stage, and the third process stage. When the current process state is identified as the second process stage, the anhydrous hydrogen fluoride concentration data is coupled with the temperature data of the fluorination reactor to generate a joint process feature. Based on the process state sequence from the first process stage to the third process stage, and combined with the joint process characteristics, an operational logic network is constructed. The multi-source data stream is traversed and analyzed using the aforementioned operational logic network to pinpoint the root cause of the fault in the fluorochemical production line.

2. The DCS fault diagnosis method for the entire process of a fluorochemical production line according to claim 1, characterized in that, Based on the changing characteristics of temperature and pressure data from the fluorination reactor, the current process state of the fluorination production line is identified. This current process state includes a first process stage, a second process stage, and a third process stage, comprising: The temperature and pressure data of the fluorination reactor during a first preset time period are obtained as the first combined data. Calculate the rate of change of temperature data in the first combination of data, and use it as the first rate of change; Calculate the rate of change of pressure data in the first combination of data, and use it as the second rate of change. When the duration of the first rate of change being within a first preset rate range and the second rate of change being within a second preset rate range reaches a first preset time length, the current process state is identified as the first process stage. After being identified as the first process stage, if the first rate of change moves from the first preset rate range to the third preset rate range, and the second rate of change moves from the second preset rate range to the fourth preset rate range, then the current process state is identified and the process enters the second process stage. If, during the duration of the second process stage, the first rate of change is greater than the fifth preset rate range, the current process state is identified and the process enters the third process stage.

3. The DCS fault diagnosis method for the entire process of a fluorochemical production line according to claim 1, characterized in that, When the current process state is identified as the second process stage, the anhydrous hydrogen fluoride concentration data is coupled with the temperature data of the fluorination reactor to generate joint process characteristics, including: When the current process state is identified as the second process stage, the anhydrous hydrogen fluoride concentration sequence and the fluorination reactor tube wall temperature sequence corresponding to the second process stage in time are obtained. The anhydrous hydrogen fluoride concentration sequence is processed to extract the fluctuation period of the anhydrous hydrogen fluoride concentration sequence, which is used as the concentration fluctuation period. The temperature sequence of the tube wall of the fluorination reactor is processed, and the decreasing slope of the temperature sequence is extracted as the decreasing slope of the tube wall temperature. Calculate the reciprocal of the concentration fluctuation period to obtain the concentration fluctuation frequency; Calculate the absolute value of the slope of the temperature drop in the pipe wall; The concentration fluctuation frequency is multiplied by the absolute value of the slope of the pipe wall temperature drop, and the product is used as the joint process feature.

4. The DCS fault diagnosis method for the entire process of a fluorochemical production line according to claim 1, characterized in that, Based on the process state sequence from the first process stage to the third process stage, and combined with the aforementioned joint process characteristics, an operational logic network is constructed, including: Map the first process stage to a first process stage node, map the second process stage to a second process stage node, and map the third process stage to a third process stage node. Establish a first connection relationship from the first process stage node to the second process stage node, wherein the first connection relationship has a first direction, and the first direction indicates the transition from the first process stage to the second process stage; Establish a second connection relationship from the second process stage node to the third process stage node, wherein the second connection relationship has a second direction, the second direction indicating the transition from the second process stage to the third process stage; Obtain the combined process features; The numerical value of the joint process feature is assigned to the first connection relationship as the attribute value of the first connection relationship; A runtime logic network is constructed based on the first process stage node, the second process stage node, the third process stage node, and the first and second connection relationships with the attribute values ​​attached.

5. The DCS fault diagnosis method for the entire process of a fluorochemical production line according to claim 4, characterized in that, Based on the first process stage node, the second process stage node, the third process stage node, and the first connection relationship and the second connection relationship with the attached attribute values, an operational logic network is constructed, including: The first process stage node, the second process stage node, and the third process stage node are considered as a set of nodes in the running logic network. The first connection relationship and the second connection relationship are used as the set of connection relationships in the running logical network; The attribute value of the first connection relationship is used as the weight value of the first connection relationship; Set a preset weight value for the second connection relationship; Based on the set of nodes, the set of connection relationships, the weight value of the first connection relationship, and the preset weight value of the second connection relationship, the topology of the running logic network is generated; Mark the first process stage node as the starting node of the running logic network; The third process stage node is marked as the termination node of the running logic network; The starting node, the ending node, and the topology are combined to construct a running logical network.

6. The DCS fault diagnosis method for the entire process of a fluorochemical production line according to claim 1, characterized in that, The multi-source data stream is traversed and parsed using the aforementioned operational logic network to locate the root cause of faults in the fluorochemical production line, including: Based on the topology of the running logic network, the traversal path from the starting node to the ending node is determined. Based on the changing trends of each data in the multi-source data stream, the current process stage node of the fluorochemical production line is identified. Starting from the current process stage node, traverse along the traversal path towards the termination node; During the traversal, obtain the weight value of the current connection relationship; Based on the matching relationship between the weight value and the corresponding data in the multi-source data stream, it is determined whether the current process status is abnormal. If the current process status is determined to be abnormal, the root cause of the fault in the fluorochemical production line is determined based on the process stage node corresponding to the current connection relationship.

7. The DCS fault diagnosis method for the entire process of a fluorochemical production line according to claim 6, characterized in that, Based on the matching relationship between the weight value and the corresponding data in the multi-source data stream, determine whether the current process state is abnormal, including: Obtain the weight value of the current connection relationship; Extract process parameter data corresponding to the current connection relationship from the multi-source data stream; Based on the correlation between the feature values ​​and the weight values ​​of the process parameter data, a matching degree index is calculated; The matching degree index is compared with a preset threshold. Determine whether the current process status is abnormal based on the comparison results.

8. A DCS fault diagnosis system for the entire process of a fluorochemical production line, characterized in that, include: The monitoring module is used to monitor the multi-source data stream output by the distributed control system of the fluorochemical production line in real time. The multi-source data stream includes temperature and pressure data of the fluorination reactor, anhydrous hydrogen fluoride concentration data, and pH and flow data of the tail gas scrubbing tower. The identification module is used to identify the current process state of the fluorochemical production line based on the change characteristics of the temperature and pressure data of the fluorination reactor. The current process state includes a first process stage, a second process stage, and a third process stage. The coupling module is used to couple the anhydrous hydrogen fluoride concentration data with the temperature data of the fluorination reactor when the current process state is identified as the second process stage, to generate joint process characteristics. A construction module is used to construct an operational logic network based on the process state sequence from the first process stage to the third process stage, combined with the joint process features; The positioning module is used to traverse and parse the multi-source data stream using the operating logic network to locate the root cause of the fault in the fluorochemical production line.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement a DCS fault diagnosis method for the entire process of a fluorochemical production line as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The device contains a computer program that, when executed by a computer, implements a DCS fault diagnosis method for the entire process of a fluorochemical production line as described in any one of claims 1 to 7.

Images