A multi-signal fusion-based zb45 unit intelligent early warning method and system

Abstract

This invention relates to the field of industrial automation monitoring, and particularly to an intelligent early warning method and system for ZB45 generating units based on multi-signal fusion. To address the problem of improving early warning accuracy, an intelligent early warning method for ZB45 generating units based on multi-signal fusion is provided, comprising: S1, acquiring the X2 detection signal and the glue level detection signal, fusing the multi-signal data using a dynamic weighting mechanism to generate an optimized multi-signal fusion result; S2, generating a graded response command; S3, recording the parameters and status information during the early warning and shutdown process; S4, dynamically updating the weight allocation parameters and the alarm threshold for severe anomalies, returning to steps S1-S4, updating the weights and the alarm threshold for severe anomalies, and completing the closed-loop optimization process. Using the method described in this invention improves the early warning accuracy.

Description

Technical Field

[0001] This invention relates to the field of industrial automation monitoring, and in particular to an intelligent early warning method and system for ZB45 generator units based on multi-signal fusion. Background Technology

[0002] With the development of industrial automation, complex equipment such as rigid box packaging units, like the ZB45 unit, are increasingly widely used in industrial production. However, during operation, the ZB45 unit often suffers from low accuracy in identifying potential equipment faults due to sensor signal acquisition errors, limitations of single signal judgment logic, and insufficient ability to analyze correlations between multiple signals. This results in false alarms and missed alarms, seriously affecting the stability of equipment operation and production efficiency.

[0003] Traditional equipment early warning methods typically rely on threshold judgments based on a single signal, making it difficult to handle multi-signal fusion issues in complex signal environments. In actual operation, when equipment is in an abnormal state (such as detection signal bypass, glue level detection failure, or sensor signal interference), the logic of a single signal is prone to deviation, making it impossible to accurately determine the fault type and severity. Furthermore, existing methods lack closed-loop optimization mechanisms based on historical data, making it difficult to dynamically adjust signal weight allocation and alarm thresholds according to real-time changes, thus failing to meet the high-precision, intelligent early warning requirements of complex industrial environments.

[0004] Therefore, how to accurately identify potential equipment faults through the fusion and correlation analysis of multiple signals, and trigger early warnings or shutdown control by combining intelligent judgment logic, has become an important technical challenge to ensure the safety of equipment operation and improve production efficiency. Summary of the Invention

[0005] This invention provides a method and system for intelligent early warning of ZB45 units based on multi-signal fusion, in order to solve the problem of how to accurately identify potential faults of ZB45 units and intelligently trigger early warning or shutdown measures based on multiple real-time signals through dynamic signal fusion and anomaly judgment logic, so as to ensure the safe and stable operation of the equipment.

[0006] This invention is achieved using the following technical solution:

[0007] A smart early warning method for ZB45 generator units based on multi-signal fusion includes:

[0008] S1. Acquire the X2 detection signal and the adhesive position detection signal, and fuse the multi-signal data using a dynamic weighting mechanism to generate an optimized multi-signal fusion result;

[0009] S2. Perform logical judgment on the optimized multi-signal fusion results generated in step S1, extract key abnormal parameters and classify them according to the severity of the abnormality, and generate graded response instructions.

[0010] S3. Receive graded response instructions, trigger warning signals for minor anomalies, trigger shutdown signals for serious anomalies, and record parameters and status information during the warning and shutdown process.

[0011] S4. Based on the warning and shutdown record sequence obtained in step S3, analyze the anomaly handling results, dynamically update the weight allocation parameters and the alarm threshold for severe anomalies, return to steps S1-S4, update the weights and the alarm threshold for severe anomalies, and complete the closed-loop optimization process.

[0012] A ZB45 unit intelligent early warning system based on multi-signal fusion includes:

[0013] The dynamic signal acquisition module is used to acquire the X2 detection signal and the adhesive position detection signal, and to fuse the multi-signal data using a dynamic weighting mechanism to generate an optimized multi-signal fusion result.

[0014] The anomaly detection module is used to make logical judgments on the optimized multi-signal fusion results, extract key anomaly parameters, classify them according to the severity of the anomalies, and generate graded response instructions.

[0015] The intelligent early warning module is used to trigger early warning signals for minor anomalies and shutdown signals for serious anomalies, while recording parameters and status information during the early warning and shutdown process;

[0016] The closed-loop feedback module analyzes the abnormal handling results based on the acquired warning and shutdown record sequence, dynamically updates the weight allocation parameters and the alarm threshold for severe abnormalities, and returns the results to the dynamic signal acquisition module.

[0017] The beneficial effects of this invention are as follows: (1) Multi-signal fusion improves detection accuracy: This invention uses a dynamic signal acquisition module to acquire X2 detection signal, empty mold box detection pulse signal and glue position detection signal, and uses a dynamic weighting mechanism to weight and fuse multi-signal data, avoiding false alarms and missed alarms caused by a single signal processing method, and significantly improving the accuracy of anomaly detection;

[0018] (2) Graded response improves early warning efficiency: By performing logical analysis on the fused signal results, graded response instructions are generated according to the severity of the anomalies. At the same time, based on the graded response instructions, early warning signals for minor anomalies or shutdown signals for severe anomalies are quickly triggered, ensuring that different fault conditions can be responded to accurately and quickly, thus improving the operating efficiency of the equipment.

[0019] (3) Closed-loop feedback enables adaptive optimization: Based on the analysis of abnormal handling results from early warning and shutdown record data, dynamic adjustments are made; weight allocation parameters and alarm thresholds are used to form optimized signal control commands, which further improves the system's adaptive capability and enables it to maintain efficient operation in complex industrial environments. Attached Figure Description

[0020] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart of the intelligent early warning method described in this invention;

[0023] Figure 2 This is a structural block diagram of the intelligent early warning system described in this invention. Detailed Implementation

[0024] To better understand the above-mentioned objectives, features, and advantages of the present invention, the solutions of the present invention will be further described below. It should be noted that, unless otherwise specified, the embodiments of the present invention and the features thereof can be combined with each other.

[0025] Many specific details are set forth in the following description in order to provide a full understanding of the invention, but the invention may also be practiced in other ways different from those described herein; obviously, the embodiments in the specification are only some embodiments of the invention, and not all embodiments.

[0026] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0027] like Figure 1 As shown, a smart early warning method for ZB45 generator units based on multi-signal fusion includes:

[0028] S1. Acquire the X2 detection signal and the adhesive position detection signal, and fuse the multi-signal data using a dynamic weighting mechanism to generate an optimized multi-signal fusion result;

[0029] The specific steps of step S1 include:

[0030] S11. Acquire the X2 detection signal and the adhesive position detection signal, and perform preliminary processing on the acquired signals to form a multi-signal initial data sequence;

[0031] Specifically, the steps include the following:

[0032] S111, Obtain the normally closed contact signal of the X2 detection button. Empty mold box detection pulse signal and adhesive position detection signal ;

[0033] S112. Perform preliminary processing on the signal obtained in step S11, namely...

[0034] right Perform logical judgments and extract the switch state sequence. ;

[0035] right Perform pulse counting and extract the time series of the running status. , ,…, };

[0036] right Voltage amplitude sampling is performed to generate the corresponding adhesive level data sequence. , ,…, };

[0037] S113. Unify the timeline of the signals after the preliminary processing in step S12 to form a multi-signal initial data sequence:

[0038]

[0039] in, For the initial data matrix of multiple signals, Signals , , Time series data on the same time axis;

[0040] S12. Based on the dynamic weighting mechanism, the initial multi-signal data sequence formed in step S11 is weighted and fused to filter out abnormal interference signals and generate a dynamic weighted fused signal sequence.

[0041] Specifically, the steps include the following:

[0042] S121. Extract signal feature values ​​from the multi-signal initial data sequence formed in step S113. That is to Extracting frequency features This indicates how frequently the X2 detection button is triggered;

[0043] right Extracting pulse amplitude This indicates the operating strength of the empty mold box;

[0044] right Extracting the average voltage This indicates the stability of the adhesive position;

[0045] S122. Calculate the dynamic weight of each signal. :

[0046]

[0047] Where i∈{X2,B222,G}, Represents the normalized weights of the signal;

[0048] S123. The initial data matrix of the multi-signal formed in step S113. Perform weighted fusion to generate dynamic weighted fusion signal data. :

[0049]

[0050] in, This represents the original data of the j-th signal;

[0051] S124. The dynamic weighted fusion signal data in step S123. Abnormal signals are identified, and noise data exceeding a set threshold is removed to obtain a filtered dynamic weighted fusion signal sequence. ;

[0052] S13. Perform logical processing on the dynamic weighted fusion signal sequence obtained in step S12, extract the correlation features between signals, and generate the optimized multi-signal fusion result.

[0053] Specifically, the steps include the following:

[0054] S131. Using Pearson correlation coefficient Measurement signal and The linear correlation is shown in the following formula:

[0055]

[0056] Where Cov represents covariance. and These represent the x-th and y-th fused signal sequences, respectively. and They represent and Standard deviation;

[0057] S132, Based on correlation results By extracting feature parameters between signals through logical rules, a feature vector F={ , ,…, };

[0058] S133. Integrate the feature vector F into the optimized multi-signal fusion result. This results in an optimized multi-signal fusion outcome. ,in This is the fusion feature vector between signals, representing the signal strength or correlation characteristics after dynamic weight fusion;

[0059] S2. Perform logical judgment on the optimized multi-signal fusion results generated in step S1, extract key abnormal parameters and classify them according to the severity of the abnormality, and generate graded response instructions.

[0060] Specifically, the steps include the following:

[0061] S21. Combine the system operating status signals, perform logical judgments and extract abnormal signal parameters to generate an abnormal parameter sequence;

[0062] The specific steps are as follows:

[0063] S211, The optimized multi-signal fusion result formed in step S133 Obtain key feature data sequences from the data;

[0064] S212. Extract real-time operating status parameters R={ from equipment operating status signals. , ,…, },in Time-series characteristics representing the current operating status of the device, such as operating load and trigger frequency;

[0065] S213, based on Based on the logical rules of R, the following abnormal signal determination formula is constructed:

[0066]

[0067] in, The parameter is the i-th abnormal signal. , These are the thresholds for signal strength and operating status, respectively. This is a weighting factor used to balance the influence of characteristic signals and operating states;

[0068] Finally, the abnormal parameter sequence A={ is generated. , ,…, };

[0069] S22. From the abnormal parameter sequence generated in step S21, based on the comparison between abnormal features and signal thresholds, the abnormal state is determined, the abnormal situation is classified, and an abnormal classification label is generated.

[0070] The specific steps include the following:

[0071] S221, The abnormal parameter sequence A = { generated in step S21} , ,…, The system retrieves all abnormal signal values ​​and compares them with preset abnormal feature thresholds to determine the abnormal state. The abnormal determination rules are as follows:

[0072]

[0073] in, This is the label for the i-th abnormal state; and These are the alarm thresholds for minor and severe abnormalities, respectively.

[0074] This generates an anomaly classification label set C={ , ,…, };

[0075] S222. Based on the category label set C, calculate the response priority according to the following logic:

[0076]

[0077] in, Indicates the priority of graded responses; The weighting factor for the classification labels is as follows: "minor anomalies" have a lower weight, while "serious anomalies" have a higher weight.

[0078] according to Generate hierarchical response instructions Hierarchical response instructions Includes a set of minor abnormal instructions and severe exception instruction set ,Right now ;

[0079] S3. Receive graded response instructions, trigger warning signals for minor anomalies, trigger shutdown signals for serious anomalies, and record parameters and status information during the warning and shutdown process.

[0080] Specifically, the steps include the following:

[0081] S31. Based on the anomaly level, trigger different warning signals to generate warning instructions for minor anomalies or shutdown instructions for severe anomalies:

[0082] Based on the tiered response instructions, the following logical rules are used to determine whether to issue a warning or shut down the system:

[0083]

[0084] Where P represents the trigger signal type, 1 is a stop signal, and 0 is a warning signal;

[0085] When P=0, a warning instruction for minor anomalies is generated. Used to remind operators;

[0086] When P=1, a stop instruction for a serious anomaly is generated. Used to immediately stop equipment operation;

[0087] S32. Lock the shutdown command and warning command to ensure that the operator cannot forcibly reset if the abnormality is not handled, and record the lock status and parameter information:

[0088] For shutdown commands triggered by severe anomalies The signal is locked to the servo stop endpoint. The locking logic is defined as follows:

[0089]

[0090] Where L represents the locked state, 1 represents the locked state, and 0 represents the unlocked state.

[0091] Simultaneously, after the shutdown command is triggered, the locked state L, the abnormal parameter sequence A, and the current equipment operating state R are recorded as a set of state information. Its expression is:

[0092]

[0093] S33. Perform time-series processing on all status information sets to generate a complete sequence of warning and shutdown records. :

[0094] ;

[0095] S4. Based on the warning and shutdown record sequence obtained in step S3, analyze the anomaly handling results, dynamically update the weight allocation parameters and the alarm threshold for severe anomalies, and return to steps S1-S4 to update the weights. and alarm thresholds for serious anomalies Complete the closed-loop optimization process.

[0096] S41. The sequence of warning and shutdown records generated in step S33 Extract key data items from the middle. Including abnormal signal parameters Locked state L i and exception handling time :

[0097]

[0098] For the extracted The abnormal signals are categorized based on their source and type (e.g., X2 detection signals, adhesive position signals) to form classification data. :

[0099]

[0100] in, ={ All records originating from the X2 signal} This is a collection of abnormal records of adhesive position signals;

[0101] S42. Calculate the number of exception triggers based on the classification data. and average processing time :

[0102]

[0103]

[0104] in, Indicates abnormal parameters Has an exception been triggered? (1 if triggered, 0 otherwise)

[0105] S43. Results of anomaly handling Each signal category in Calculation logic accuracy :

[0106]

[0107] Among them, the accurate trigger count represents the number of times that the signal triggering logic matches the actual exception;

[0108] S43, according to As a result, the dynamic weights for signal categories Adjustments will be made:

[0109]

[0110] in, The adjusted dynamic weights; k is the adjustment coefficient; Accuracy rate preset by humans for the target;

[0111] S44. Use the adjusted dynamic weights and number of exceptions triggered Recalculate the alarm threshold for severe anomalies:

[0112]

[0113] in, The updated alarm threshold for severe anomalies; c is the adjustment coefficient; This represents the desired number of triggers.

[0114] S45. Repeat steps S1-S4 to update the weights. and alarm thresholds for serious anomalies Complete the closed-loop optimization process.

[0115] like Figure 2 As shown, a ZB45 unit intelligent early warning system based on multi-signal fusion includes:

[0116] The dynamic signal acquisition module is used to acquire X2 detection signal, CH detection signal and colloid position detection signal, and use a dynamic weighting mechanism to fuse the multi-signal data to generate an optimized multi-signal fusion result;

[0117] The anomaly detection module is used to make logical judgments on the optimized multi-signal fusion results, extract key anomaly parameters, classify them according to the severity of the anomalies, and generate graded response instructions.

[0118] The intelligent early warning module is used to trigger early warning signals for minor anomalies and shutdown signals for serious anomalies, while recording parameters and status information during the early warning and shutdown process;

[0119] The closed-loop feedback module analyzes the anomaly handling results based on the acquired early warning and shutdown record sequences, dynamically updates the weight allocation parameters and alarm thresholds for severe anomalies, and forms a closed-loop feedback.

[0120] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the present invention. Although detailed descriptions have been provided 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 or all of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments, and they should all be covered within the protection scope of the claims.

Claims

1. A smart early warning method for ZB45 generator units based on multi-signal fusion, characterized in that, include: S1. Acquire the X2 detection signal and the adhesive position detection signal, and fuse the multi-signal data using a dynamic weighting mechanism to generate an optimized multi-signal fusion result; S2. Perform logical judgment on the optimized multi-signal fusion results generated in step S1, extract key abnormal parameters and classify them according to the severity of the abnormality, and generate graded response instructions. S3. Receive graded response instructions, trigger warning signals for minor anomalies, trigger shutdown signals for serious anomalies, and record parameters and status information during the warning and shutdown process. S4. Based on the warning and shutdown record sequence obtained in step S3, analyze the anomaly handling results, dynamically update the weight allocation parameters and the alarm threshold for severe anomalies, and return to steps S1-S4 to update the weights. and alarm thresholds for severe anomalies Complete the closed-loop optimization process.

2. The intelligent early warning method for ZB45 units based on multi-signal fusion according to claim 1, characterized in that, The specific steps of step S1 include: S11. Acquire the X2 detection signal and the adhesive position detection signal, and perform preliminary processing on the acquired signals to form a multi-signal initial data sequence; S12. Based on the dynamic weighting mechanism, the initial multi-signal data sequence formed in step S11 is weighted and fused to filter out abnormal interference signals and generate a dynamic weighted fused signal sequence. S13. Perform logical processing on the dynamic weighted fusion signal sequence obtained in step S12, extract the correlation features between signals, and generate the optimized multi-signal fusion result.

3. The intelligent early warning method for ZB45 units based on multi-signal fusion according to claim 2, characterized in that, The specific steps of step S11 include: S111, Obtain the normally closed contact signal of the X2 detection button. Empty mold box detection pulse signal and adhesive position detection signal ; S112. Perform preliminary processing on the signal obtained in step S11, namely... right Perform logical judgments and extract the switch state sequence. ; right Perform pulse counting and extract the time series of the running status. , ,…, }; right Voltage amplitude sampling is performed to generate the corresponding adhesive level data sequence. , ,…, }; S113. Unify the timeline of the signals after the preliminary processing in step S12 to form a multi-signal initial data sequence: ； in, For the initial data matrix of multiple signals, Signals , , Time series data on the same time axis.

4. The intelligent early warning method for ZB45 units based on multi-signal fusion according to claim 3, characterized in that, The specific steps of step S12 include: S121. Extract signal feature values ​​from the multi-signal initial data sequence formed in step S113. ,Right now right Extracting frequency features This indicates how frequently the X2 detection button is triggered; right Extracting pulse amplitude This indicates the operating strength of the empty mold box; right Extracting the average voltage This indicates the stability of the adhesive position; S122. Calculate the dynamic weight of each signal. : ； Where i∈{X2,B222,G}, Represents the normalized weights of the signal; S123. Perform weighted fusion on the initial multi-signal data sequence formed in step S113 to generate dynamic weighted fused signal data. : ； in, This represents the original data of the j-th signal; S124. The dynamic weighted fusion signal data in step S123. Abnormal signals are identified, and noise data exceeding a set threshold is removed to obtain a filtered dynamic weighted fusion signal sequence. .

5. The intelligent early warning method for ZB45 units based on multi-signal fusion according to claim 4, characterized in that, The specific steps of step S13 include: S131. Using Pearson correlation coefficient Measurement signal and The linear correlation is shown in the following formula: ； Where Cov represents covariance. and These represent the x-th and y-th fused signal sequences, respectively. and They represent and Standard deviation; S132, Based on correlation results By extracting feature parameters between signals through logical rules, a feature vector F={ , ,…, }; S133. Integrate the feature vector F into the optimized multi-signal fusion result. This results in an optimized multi-signal fusion outcome. ,in This is the fusion feature vector between signals, representing the signal strength or correlation characteristics after dynamic weight fusion.

6. The intelligent early warning method for ZB45 units based on multi-signal fusion according to claim 5, characterized in that, The specific steps of step S2 include: S21. Combine the system operating status signals, perform logical judgments and extract abnormal signal parameters to generate an abnormal parameter sequence; The specific steps are as follows: S211, The optimized multi-signal fusion result formed in step S133 Obtain key feature data sequences from the data; S212. Extract real-time operating status parameters R={ from equipment operating status signals. , ,…, },in Time-series characteristics representing the current operating status of the device, such as operating load and trigger frequency; S213, based on Based on the logical rules of R, the following abnormal signal determination formula is constructed: ； in, The parameter is the i-th abnormal signal. , These are the thresholds for signal strength and operating status, respectively. This is a weighting factor used to balance the influence of characteristic signals and operating states; Finally, the abnormal parameter sequence A={ is generated. , ,…, }; S22. From the abnormal parameter sequence generated in step S21, based on the comparison between abnormal features and signal thresholds, the abnormal state is determined, the abnormal situation is classified, and an abnormal classification label is generated. The specific steps include the following: S221, The abnormal parameter sequence A = { generated in step S21} , ,…, The system retrieves all abnormal signal values ​​and compares them with preset abnormal feature thresholds to determine the abnormal state. The abnormal determination rules are as follows: ； in, This is the label for the i-th abnormal state; and These are the alarm thresholds for minor and severe abnormalities, respectively. This generates an anomaly classification label set C={ , ,…, }; S222. Based on the category label set C, calculate the response priority according to the following logic: ； in, Indicates the priority of graded responses; As a weighting factor for the classification labels, "minor anomalies" have a lower weight and "serious anomalies" have a higher weight. according to Generate hierarchical response instructions Hierarchical response instructions Includes a set of minor abnormal instructions and severe exception instruction set ,Right now .

7. The intelligent early warning method for ZB45 units based on multi-signal fusion according to claim 6, characterized in that, The specific steps of step S3 include: S31. Based on the anomaly level, trigger different warning signals to generate warning instructions for minor anomalies or shutdown instructions for severe anomalies: Based on the tiered response instructions, the following logical rules are used to determine whether to issue a warning or shut down the system: ； Where P represents the trigger signal type, 1 is a stop signal, and 0 is a warning signal; When P=0, a warning instruction for minor anomalies is generated. Used to remind operators; When P=1, a stop instruction for a serious anomaly is generated. Used to immediately stop equipment operation; S32. Lock the shutdown command and warning command to ensure that the operator cannot forcibly reset if the abnormality is not handled, and record the lock status and parameter information: For shutdown commands triggered by severe anomalies The signal is locked to the servo stop endpoint. The locking logic is defined as follows: ； Where L represents the locked state, 1 represents the locked state, and 0 represents the unlocked state; Simultaneously, after the shutdown command is triggered, the locked state L, the abnormal parameter sequence A, and the current equipment operating state R are recorded as a set of state information. Its expression is: ； S33. Perform time-series processing on all status information sets to generate a complete sequence of warning and shutdown records. : 。 8. The intelligent early warning method for ZB45 units based on multi-signal fusion according to claim 7, characterized in that, Step S4 specifically includes the following steps: S41. The sequence of warning and shutdown records generated in step S33 Extract key data items from the middle. Including abnormal signal parameters Locked state L i and exception handling time : ； For the extracted The abnormal signals are categorized and classified according to their source and type to form classification data. : ； in, ={ All records originating from the X2 signal} This is a collection of abnormal records of adhesive position signals; S42. Calculate the number of exception triggers based on the classification data. and average processing time : ； ； in, Indicates abnormal parameters Has any exception been triggered? S43. Results of exception handling Each signal category in Calculation logic accuracy : ； Among them, the accurate trigger count represents the number of times that the signal triggering logic matches the actual exception; S43, according to As a result, the dynamic weights for signal categories Adjustments will be made: ； in, The adjusted dynamic weights; k is the adjustment coefficient; Accuracy rate preset by humans for the target; S44. Use the adjusted dynamic weights and number of exceptions triggered Recalculate the alarm threshold for severe anomalies: ； in, The updated alarm threshold for severe anomalies; c is the adjustment coefficient; The desired number of triggers; S45. Repeat steps S1-S4 to update the weights. and alarm thresholds for severe anomalies Complete the closed-loop optimization process.

9. A ZB45 unit intelligent early warning system based on multi-signal fusion, characterized in that, include: The dynamic signal acquisition module is used to acquire X2 detection signal, CH detection signal and colloid position detection signal, and use a dynamic weighting mechanism to fuse the multi-signal data to generate an optimized multi-signal fusion result; The anomaly detection module is used to make logical judgments on the optimized multi-signal fusion results, extract key anomaly parameters, classify them according to the severity of the anomalies, and generate graded response instructions. The intelligent early warning module is used to trigger early warning signals for minor anomalies and shutdown signals for serious anomalies, while recording parameters and status information during the early warning and shutdown process; The closed-loop feedback module analyzes the abnormal handling results based on the acquired warning and shutdown record sequence, dynamically updates the weight allocation parameters and the alarm threshold for severe abnormalities, and returns the results to the dynamic signal acquisition module.

Images