A pneumatic ash conveying abnormal state diagnosis and fault early warning method

Abstract

The application provides a pneumatic ash conveying abnormal state diagnosis and fault early warning method, comprising the following steps: calling preset monitoring tools corresponding to each link of a pneumatic ash conveying system to perform real-time monitoring on the running state of the matching link; based on the real-time monitoring result, determining abnormal information corresponding to the matching link, pre-analyzing the abnormal information, and locking the abnormal position of the matching link; inputting the abnormal information corresponding to the matching link into a matched self-diagnosis alarm model for fault early warning, determining the fault index and fault type corresponding to each early warning information in the fault early warning based on a warning-fault database; obtaining a separate solution corresponding to the fault index and fault type of each early warning information in the matching link, obtaining a comprehensive solution for the matching link, associating the comprehensive solution with the abnormal position, and issuing the comprehensive solution to a manual end for intervention.

Description

Technical Field

[0001] This invention relates to the field of abnormal state diagnosis and fault early warning technology, and in particular to a method for abnormal state diagnosis and fault early warning of pneumatic ash conveying. Background Technology

[0002] Pneumatic conveying systems are widely used pneumatic conveying engineering systems that utilize the energy of airflow to transport granular materials along the airflow direction within a closed pipeline. They are a specific application of fluidization technology and are commonly used conveying systems in power plants and the metallurgical industry.

[0003] Currently, the control of pneumatic ash conveying systems is mainly based on automatic control of ash conveying through preset parameters. When an abnormal state occurs in the ash conveying system, it is impossible to detect the fault type or determine the fault location in a timely manner. This will cause a significant decrease in the operating efficiency of the ash conveying system and an increase in energy consumption. Furthermore, the untimely ash conveying may cause the ash level in the electrostatic precipitator ash hopper to rise, affecting the safe operation of the electrostatic precipitator system, leading to the shutdown of the ash conveying system and equipment damage.

[0004] Therefore, this invention proposes a method for diagnosing abnormal conditions and providing early warning of faults in pneumatic ash conveying. Summary of the Invention

[0005] This invention provides a method for diagnosing abnormal states and providing early warning of faults in pneumatic ash conveying systems. The method uses a preset monitoring tool to monitor the operating status of the pneumatic ash conveying system in real time. After obtaining abnormal information based on the monitoring results, the abnormal information is pre-analyzed to locate the abnormal position. Then, a self-diagnostic alarm model is used to provide early warning of faults based on the abnormal information. Finally, a matching solution is obtained based on the fault index and fault type and sent to the manual end for intervention.

[0006] This invention provides a method for diagnosing abnormal conditions and providing early warning of faults in pneumatic ash conveying systems, including:

[0007] Step 1: Retrieve the preset monitoring tools corresponding to each link of the pneumatic ash conveying system to monitor the operating status of the matching links in real time;

[0008] Step 2: Based on the real-time monitoring results, determine the abnormal information of the corresponding matching link, and perform pre-analysis on the abnormal information to pinpoint the abnormal location of the corresponding matching link;

[0009] Step 3: Input the abnormal information of the corresponding matching link into the matching self-diagnosis alarm model for fault warning. Based on the warning-fault database, determine the fault index and fault type corresponding to each warning information in the fault warning.

[0010] Step 4: Obtain the fault index and individual solutions corresponding to each early warning information in the same matching process, obtain a comprehensive solution for the same matching process, associate the comprehensive solution with the abnormal location, and send it to the manual intervention terminal.

[0011] In one possible implementation, preset monitoring tools corresponding to each stage of the pneumatic ash conveying system are invoked to monitor the operating status of the matching stages in real time, including:

[0012] The type of each link in the pneumatic ash conveying system is determined, and a preset monitoring tool for each link type is determined based on the type-tool mapping table;

[0013] When the pneumatic ash conveying system starts working, the preset monitoring tool controls the operation status of the corresponding matching links in real time.

[0014] In one possible implementation, based on real-time monitoring results, abnormal information in the corresponding matching process is determined, and the abnormal information is pre-analyzed to pinpoint the abnormal location in the corresponding matching process, including:

[0015] Collect the first monitoring data during normal operation of the pneumatic ash conveying system, and extract the normal operation period of each link in the pneumatic ash conveying system and the parameter fluctuation range of normal operation parameters under different normal operation periods based on the first monitoring data;

[0016] Collect real-time monitoring data of the pneumatic ash conveying system during actual operation, and extract the actual operating segment of each link in the pneumatic ash conveying system and the actual working parameters under different actual operating segments based on the real-time monitoring data;

[0017] Compare the fluctuation range of normal operating parameters with the actual operating parameters under the same process;

[0018] Calculate outliers in the comparison results;

[0019]

[0020]

[0021] Among them, y1 i Y1 represents the i-th actual working parameter in the same stage, where the parameter fluctuation range of the normal working parameter matching the i-th actual working parameter is [a1, a2]; i This represents the abnormal factor of the i-th actual working parameter in the same stage; This represents the abnormal adjustment coefficient corresponding to the i-th actual working parameter in the same stage being less than a1; Y0 represents the abnormal adjustment coefficient corresponding to the i-th actual working parameter in the same stage being greater than a1; Y0 represents the second abnormal value of the comparison result in the same stage; n1 represents the number of actual working parameters in the same stage. This represents the parameter weight of the i-th actual working parameter in the same stage, and

[0022] Determine whether the normal operating period and the actual operating period are consistent in the same process. If they are consistent, determine whether the abnormal value is greater than the preset value.

[0023] If it is greater than, then it is determined that there is an anomaly in the same link, and from all y1 i The middle screening meets the requirements as well as The abnormal parameters are used as abnormal information in the same process, where a3 represents the first abnormal threshold and a4 represents the second abnormal threshold.

[0024] Based on the abnormal combination-location mapping table, the location that matches the abnormal information is determined as the abnormal location of the corresponding matching step.

[0025] In one possible implementation, abnormal information from the corresponding matching step is input into the matching self-diagnostic alarm model for fault warning, including:

[0026] Based on the self-diagnosis alarm model, the abnormal cause set of each abnormal parameter involved in the abnormal information and the parameter association relationship between each abnormal parameter and the remaining abnormal parameters are determined, and the first abnormal cause is extracted from the abnormal cause set.

[0027] Based on the self-diagnostic alarm model, a warning message consistent with the first cause of the anomaly is output.

[0028] In one possible implementation, based on an early warning-fault database, the fault index and fault type corresponding to each early warning message are determined, including:

[0029] Each warning message in the fault warning is compared with the warning-fault database.

[0030] Based on the comparison results, the fault index and fault type corresponding to the corresponding early warning information are obtained.

[0031] In one possible implementation, the fault index and fault type corresponding to each early warning information in the matching process are obtained, and a comprehensive solution for the matching process is obtained, including:

[0032] Based on the index-type-solution database, a unique solution is obtained for each early warning information;

[0033] Each individual solution is analyzed, and the corresponding fault-solving characteristics of the individual solution are obtained. The fault-solving characteristics involved in the same individual solution are of different types.

[0034] Cluster analysis is performed on the fault-solving features involved in all individual solutions. The cluster centers and the first number of the first feature with a similarity greater than a preset degree to the cluster centers, as well as the second number of the remaining second features, are obtained in each cluster classification result. A feature construction map of the corresponding classification result is then constructed.

[0035] Based on the feature graph, the third feature with the largest edge weight among the first features is retained;

[0036] Based on the cluster centers, third features, and second features in each cluster analysis result, feature fusion is performed on all individual solutions to obtain a comprehensive solution for the same matching step.

[0037] In one possible implementation, the process of retaining the third feature with the largest edge weight among the first features in the process of constructing the graph based on the features further includes:

[0038] Calculate the edge weight B1 between each first feature and the cluster center;

[0039] B1 = L1 sim(D1,D0)

[0040] Where L1 represents the distance between the first feature and the cluster center; sim(D1,D0) represents the similarity function between the first feature D1 and the cluster center D0.

[0041] In one possible implementation, the comprehensive solution is associated with the anomaly location and sent to a human intervention point, including:

[0042] Establish a relation array between the comprehensive solution and the anomaly location, where the relation array = [comprehensive solution, anomaly location];

[0043] The relation array is then sent to a human operator for intervention.

[0044] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0045] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0046] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0047] Figure 1 This is a flowchart of a method for diagnosing abnormal states and providing early warning of faults in pneumatic ash conveying according to an embodiment of the present invention;

[0048] Figure 2 This is a flowchart illustrating the operation of the pneumatic ash conveying system in an embodiment of the present invention. Detailed Implementation

[0049] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0050] This invention provides a method for diagnosing abnormal states and providing early warning of faults in pneumatic ash conveying systems, such as... Figure 1 As shown, it includes:

[0051] Step 1: Retrieve the preset monitoring tools corresponding to each link of the pneumatic ash conveying system to monitor the operating status of the matching links in real time;

[0052] Step 2: Based on the real-time monitoring results, determine the abnormal information of the corresponding matching link, and perform pre-analysis on the abnormal information to pinpoint the abnormal location of the corresponding matching link;

[0053] Step 3: Input the abnormal information of the corresponding matching link into the matching self-diagnosis alarm model for fault warning. Based on the warning-fault database, determine the fault index and fault type corresponding to each warning information in the fault warning.

[0054] Step 4: Obtain the fault index and individual solutions corresponding to each early warning information in the same matching process, obtain a comprehensive solution for the same matching process, associate the comprehensive solution with the abnormal location, and send it to the manual intervention terminal.

[0055] In this embodiment, the pneumatic ash conveying system is a widely used pneumatic conveying engineering system, commonly used in power plants and metallurgical industries. The pneumatic ash conveying system includes feeding, fluidization and pressurization, and conveying stages, as detailed below. Figure 2 As shown.

[0056] In this embodiment, the preset monitoring tools include a level gauge, a position feedback detector, a pressure transmitter, and a temperature measuring resistance thermometer. A level gauge, a pressure transmitter 1, and a temperature measuring resistance thermometer are installed inside the silo pump. A position feedback detector 1 is installed at the feed valve, a position feedback detector 2 is installed at the discharge valve, a position feedback detector 3 is installed at the primary air inlet valve, a position feedback detector 4 is installed at the secondary air inlet valve, and a pressure transmitter 2 is installed at the ash conveying pipeline as a monitoring tool.

[0057] In this embodiment, the abnormal information includes abnormal time periods, abnormal material levels in level gauges, abnormal detection results from position feedback detectors, abnormal pressure values ​​detected by pressure transmitters, and abnormal resistance values ​​in temperature measurement thermal resistors.

[0058] In this embodiment, in order to obtain real-time monitoring data of the pneumatic ash conveying system during operation, actual working parameters are extracted from the monitoring data and compared with normal working parameters to obtain comparison results. The outlier of the comparison results is calculated, and it is determined whether the outlier is greater than a preset value, thereby determining whether there is an abnormality in the operating status. If there is an abnormality, the location of the abnormal information matching is determined through the corresponding matching table. The above process is a pre-analysis process, and the matching table is set in advance by experts.

[0059] In this embodiment, abnormal locations include locations where malfunctions may occur, such as the silo pump, various valves, and ash conveying pipes.

[0060] In this embodiment, the process of forming the early warning-fault database is as follows: the types of faults are counted according to the categories of monitoring tools, and each type of fault is numbered. The early warning signal corresponding to each fault category is obtained according to the fault category, and the early warning signal is associated with the fault category to form the early warning-fault database. The association of different fault categories with which early warning signal is set in advance to facilitate timely understanding of the fault.

[0061] In this embodiment, the fault index and fault type refer to the different fault types and fault indices corresponding to each type of abnormal information determined after the abnormal information is input into the fault-early warning database.

[0062] In this embodiment, the association between the comprehensive solution and the abnormal location can be achieved by establishing a relation array.

[0063] In this embodiment, each individual solution is analyzed to obtain the corresponding fault-solving features. Then, the fault-solving features involved in all individual solutions are clustered. Based on the results of the clustering analysis, the features of all individual solutions are fused to obtain a comprehensive solution for the matching process.

[0064] The beneficial effects of the above technical solution are as follows: by using preset monitoring tools to monitor the operational status of each link in real time, it is possible to respond to abnormal states in a timely manner, match abnormal information with links, further pinpoint the abnormal location, issue different fault warnings for the abnormal location, match the solution with different fault warnings, and send the solution to the human end, thereby improving the timeliness of abnormal state detection and the accuracy of abnormal location pinpointing. It can reflect the exact location of the abnormality and obtain fault solutions in a timely manner to accurately guide the human end to resolve the fault.

[0065] This invention provides a method for diagnosing abnormal states and providing early warning of faults in pneumatic ash conveying systems. The method involves retrieving preset monitoring tools corresponding to each stage of the pneumatic ash conveying system to monitor the operating status of the matched stages in real time, including:

[0066] The type of each link in the pneumatic ash conveying system is determined, and a preset monitoring tool for each link type is determined based on the type-tool mapping table;

[0067] When the pneumatic ash conveying system starts working, the preset monitoring tool controls the operation status of the corresponding matching links in real time.

[0068] In this embodiment, the link type corresponds one-to-one with each link in the pneumatic ash conveying system.

[0069] In this embodiment, the corresponding monitoring tools for different stages can be obtained from the type-tool mapping table. The level gauge and position feedback detector 1 are used as monitoring tools for the feeding stage; the pressure transmitter 1, position feedback detector 1, position feedback detector 2, and position feedback detector 3 are used as monitoring tools for the fluidization pressurization stage; and the temperature measuring resistance thermometer and pressure transmitter 2 are used as monitoring tools for the conveying stage.

[0070] The beneficial effects of the above technical solution are: by using preset monitoring tools to monitor each link in the pneumatic ash conveying system simultaneously, abnormal situations can be captured in a timely manner when the pneumatic ash conveying system is in an abnormal operating state, and can be mapped to different link types, thus speeding up the detection of faults.

[0071] This invention provides a method for diagnosing abnormal states and providing early warning of faults in pneumatic ash conveying. Based on real-time monitoring results, it determines abnormal information in the corresponding matching links, performs pre-analysis on the abnormal information, and pinpoints the abnormal location in the corresponding matching links, including:

[0072] Collect the first monitoring data during normal operation of the pneumatic ash conveying system, and extract the normal operation period of each link in the pneumatic ash conveying system and the parameter fluctuation range of normal operation parameters under different normal operation periods based on the first monitoring data;

[0073] Collect real-time monitoring data of the pneumatic ash conveying system during actual operation, and extract the actual operating segment of each link in the pneumatic ash conveying system and the actual working parameters under different actual operating segments based on the real-time monitoring data;

[0074] Compare the fluctuation range of normal operating parameters with the actual operating parameters under the same process;

[0075] Calculate outliers in the comparison results;

[0076]

[0077]

[0078] Among them, y1 i Y1 represents the i-th actual working parameter in the same stage, where the parameter fluctuation range of the normal working parameter matching the i-th actual working parameter is [a1, a2]; i This represents the abnormal factor of the i-th actual working parameter in the same stage; This represents the abnormal adjustment coefficient corresponding to the i-th actual working parameter in the same stage being less than a1; Y0 represents the abnormal adjustment coefficient corresponding to the i-th actual working parameter in the same stage being greater than a1; Y0 represents the second abnormal value of the comparison result in the same stage; n1 represents the number of actual working parameters in the same stage. This represents the parameter weight of the i-th actual working parameter in the same stage, and

[0079] Determine whether the normal operating period and the actual operating period are consistent in the same process. If they are consistent, determine whether the abnormal value is greater than the preset value.

[0080] If it is greater than, then it is determined that there is an anomaly in the same link, and from all y1 i The middle screening meets the requirements as well as The abnormal parameters are used as abnormal information in the same process, where a3 represents the first abnormal threshold and a4 represents the second abnormal threshold.

[0081] Based on the abnormal combination-location mapping table, the location that matches the abnormal information is determined as the abnormal location of the corresponding matching step.

[0082] In this embodiment, the first monitoring data is the normal monitoring data obtained by real-time monitoring of the system's operating status through a preset monitoring tool when the pneumatic ash conveying system is operating normally.

[0083] In this embodiment, the parameter fluctuation range refers to the fact that under normal circumstances, the parameters of the system have a certain numerical fluctuation. The parameter values ​​within this fluctuation range can be regarded as normal parameters, and the parameter fluctuation range is set in advance.

[0084] In this embodiment, the pneumatic ash conveying system will generate many sets of actual working parameters during actual operation. By comparing the parameter fluctuation range of each set of actual working parameters with the matching normal working parameters, the specific location of the abnormal information can be determined.

[0085] In this embodiment, the first abnormal threshold and the second abnormal threshold are determined based on the normal operating parameters collected during the normal operation of the pneumatic ash conveying system.

[0086] In this embodiment, the anomaly combination-location mapping table is used to combine different anomaly information when they coexist to determine the specific anomaly situation. Examples include: if the level gauge measures a level that has reached the specified level but the feeding stage has not ended, the system fault is determined to be excessive feeding; if the pressure value detected by pressure transmitter 1 is lower than the preset pressure value and the data fed back by position feedback detectors 1 (position feedback device 2, position feedback device 3) is also abnormal, the two anomalies can be combined to determine the system fault as internal leakage of the system valve in the fluidization pressurization stage; if the pressure value detected by pressure transmitter 1 has reached the preset pressure value but the fluidization pressurization time has not reached the preset time, the system fault is determined to be system overpressure; if the ash conveying system has reached its maximum operating time and the pressure value of pressure transmitter 2 is still greater than the preset pressure value, and the resistance value of the temperature measuring resistor is normal, the system fault is determined to be ash blockage in the conveying main pipe; if the resistance value of the temperature measuring resistor is also abnormal, the system fault is determined to be ash blockage in the silo pump.

[0087] The beneficial effects of the above technical solution are: by comparing the real-time monitoring results with the monitoring data under normal system operation, the consistency of time periods can accurately locate the specific link in the system operation process where the abnormality occurs. Considering the actual situation of parameter fluctuations, the link and location of the abnormal information can be accurately determined, thus achieving precise location of the abnormal state.

[0088] This invention provides a method for diagnosing abnormal states and providing early warning of faults in pneumatic ash conveying, which inputs abnormal information of the corresponding matching link into a matching self-diagnostic alarm model for fault warning, including:

[0089] Based on the self-diagnosis alarm model, the abnormal cause set of each abnormal parameter involved in the abnormal information and the parameter association relationship between each abnormal parameter and the remaining abnormal parameters are determined, and the first abnormal cause is extracted from the abnormal cause set.

[0090] Based on the self-diagnostic alarm model, a warning message consistent with the first cause of the anomaly is output.

[0091] In this embodiment, the first abnormal cause refers to the abnormal cause with the highest correlation obtained by matching each abnormal parameter with each abnormal cause in the abnormal cause set through the parameter correlation relationship between each abnormal parameter and the remaining abnormal parameters. Since the occurrence of each abnormal position is not caused by a single parameter abnormality, but is usually caused by the combined effect of multiple abnormal parameters, it is necessary to establish the correlation relationship between parameters.

[0092] The beneficial effects of the above technical solution are: by matching the correlation between abnormal parameters with the abnormal causes in the abnormal cause set to obtain the abnormal cause with the highest matching degree, and then corresponding it with the early warning information in the self-diagnostic alarm model, the accuracy of the early warning information can be effectively improved.

[0093] This invention provides a method for diagnosing abnormal states and providing early warnings of faults in pneumatic ash conveying systems. Based on an early warning-fault database, the method determines the fault index and fault type corresponding to each early warning message, including:

[0094] Each warning message in the fault warning is compared with the warning-fault database.

[0095] Based on the comparison results, the fault index and fault type corresponding to the corresponding early warning information are obtained.

[0096] In this embodiment, specific fault types include: excessive ash intake, insufficient ash intake, internal leakage of system valves, system overpressure, pipe blockage, etc. By assigning different fault indices to each different fault type, the purpose of characterizing different fault types can be achieved through fault indices.

[0097] The beneficial effects of the above technical solution are: by comparing the early warning information with the data in the database to obtain the fault index and the corresponding fault type, the accuracy of obtaining the specific fault type can be improved, which facilitates the matching of subsequent fault solutions.

[0098] This invention provides a method for diagnosing abnormal states and providing early warning of faults in pneumatic ash conveying. It obtains the fault index and individual solutions corresponding to the fault type for each early warning information in the matching process, and then obtains a comprehensive solution for the matching process, including:

[0099] Based on the index-type-solution database, a unique solution is obtained for each early warning information;

[0100] Each individual solution is analyzed, and the corresponding fault-solving characteristics of the individual solution are obtained. The fault-solving characteristics involved in the same individual solution are of different types.

[0101] Cluster analysis is performed on the fault-solving features involved in all individual solutions. The cluster centers and the first number of the first feature with a similarity greater than a preset degree to the cluster centers, as well as the second number of the remaining second features, are obtained in each cluster classification result. A feature construction map of the corresponding classification result is then constructed.

[0102] Based on the feature graph, the third feature with the largest edge weight among the first features is retained;

[0103] Based on the cluster centers, third features, and second features in each cluster analysis result, feature fusion is performed on all individual solutions to obtain a comprehensive solution for the same matching step.

[0104] In this embodiment, the index-type-solution database is a database formed by matching different fault indices to different fault types and then matching them one by one with the individual solutions for each fault.

[0105] In this embodiment, cluster analysis refers to the analysis process of grouping a set of data objects into multiple classes composed of similar objects. It can be applied in the data preprocessing process. For multidimensional data with complex structures, cluster analysis can be used to aggregate the data, standardize the complex structure data, discover the dependencies between data items, thereby removing or merging data items with close dependencies, and also provide preprocessing functions for some data mining methods.

[0106] In this embodiment, the cluster center refers to the center point of the cluster category. The initial cluster center is a cluster point randomly selected by the clustering algorithm. If you need to view the cluster center situation, you need to focus on the final cluster center.

[0107] In this embodiment, the first feature is the fault feature with the highest similarity to the cluster center, and the second feature is the remaining feature after removing the first feature. The first number is the number of the first feature, and the second number is the number of the second feature.

[0108] In this embodiment, the feature construction map refers to the construction map obtained by combining the first feature of the first number and the second feature of the second number with the corresponding classification results.

[0109] In this embodiment, the edge weight is the weight of the edge, which means the value of the edge connecting two nodes.

[0110] In this embodiment, the third feature refers to the feature with the largest edge weight among the first features in the feature construction graph.

[0111] In this embodiment, feature fusion refers to the specific process of merging different individual solutions according to the cluster centers, second features, and third features in the cluster analysis results to obtain a comprehensive solution for the same matching link.

[0112] The beneficial effects of the above technical solution are: by obtaining clustering results through clustering analysis and then using the clustering analysis results to fuse features of individual solutions, a comprehensive solution for the corresponding matching link can be obtained, which helps to accurately analyze the corresponding solution for the fault category and provides convenience for accurately solving the faults of the pneumatic ash conveying system.

[0113] This invention provides a method for diagnosing abnormal states and providing early warning of faults in pneumatic ash conveying. Based on the feature construction graph, the process of retaining the third feature with the largest edge weight among the first features includes:

[0114] Calculate the edge weight B1 between each first feature and the cluster center;

[0115] B1 = L1 sim(D1,D0)

[0116] Where L1 represents the distance between the first feature and the cluster center; sim(D1,D0) represents the similarity function between the first feature D1 and the cluster center D0.

[0117] In this embodiment, the similarity function is used to measure the similarity between the first feature and the cluster center.

[0118] The beneficial effects of the above technical solution are: by obtaining the similarity between the first feature and the cluster center through the similarity function, the edge weight between the first feature and the cluster center can be obtained. The third feature in the first feature can be obtained by calculating the edge weight, thereby ensuring the determinism of the comprehensive solution and the fault feature.

[0119] This invention provides a method for diagnosing and warning of abnormal states in pneumatic ash conveying systems. The method associates the comprehensive solution with the location of the abnormality and sends it to a manual intervention point, including:

[0120] Establish a relation array between the comprehensive solution and the anomaly location, where the relation array = [comprehensive solution, anomaly location];

[0121] The relation array is then sent to a human operator for intervention.

[0122] In this embodiment, the comprehensive solution and the abnormal location have a one-to-one correspondence.

[0123] The beneficial effects of the above technical solution are: determining the comprehensive solution for abnormal locations and associating the abnormal locations with the comprehensive solution to the manual intervention end can effectively guide the manual labor to accurately find the fault location and promptly resolve the faults that occur during the operation of the pneumatic ash conveying system according to the solution.

[0124] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for diagnosing abnormal state and early warning of fault of pneumatic ash conveying, characterized in that, The method comprises the following steps: Step 1: Real-time monitoring of the running state of the matching link by calling the corresponding preset monitoring tool of each link of the pneumatic ash conveying system; Step 2: Based on the real-time monitoring result, determine the abnormal information of the corresponding matching link, and pre-analyze the abnormal information to lock the abnormal position of the corresponding matching link; Step 3: Input the abnormal information of the corresponding matching link into the matched self-diagnosis alarm model for fault early warning, and determine the fault index and fault type corresponding to each early warning information in the fault early warning based on the early warning-fault database; Step 4: Obtain the individual solution corresponding to the fault index and fault type of each early warning information in the matching link, and obtain the comprehensive solution for the matching link, and associate the comprehensive solution with the abnormal position, and issue it to the artificial end for intervention; The method for obtaining the individual solution corresponding to the fault index and fault type of each early warning information in the matching link and obtaining the comprehensive solution for the matching link comprises: Based on the index-type-solution database, obtain the individual solution of each early warning information; Analyze each individual solution and obtain the fault solving features of the corresponding individual solution, wherein the fault solving features involved in the same individual solution are different types; Cluster analysis is performed on the fault solving features involved in all individual solutions to obtain the cluster center and the first number of the first feature with a similarity greater than a preset degree and the second number of the remaining second feature in each cluster classification result, and a feature construction graph corresponding to the classification result is constructed; Based on the feature construction graph, retain the third feature with the largest edge weight in the first feature; Based on the cluster center, the third feature and the second feature in each cluster analysis result, perform feature fusion on all individual solutions to obtain the comprehensive solution for the matching link; In the process of retaining the third feature with the largest edge weight in the first feature based on the feature construction graph, it further comprises: Calculate the edge weight B1 of each first feature and the cluster center; wherein L1 represents a distance of the first feature to the cluster center; represents a similarity function of the first feature D1 to the cluster center D0.

2. The method according to claim 1, characterized in that, In step 1, the preset monitoring tool corresponding to each link of the pneumatic ash conveying system is called to real-time monitor the running state of the matching link, which comprises: Determine the link type of each link in the pneumatic ash conveying system, and determine the preset monitoring tool of each link type based on the type-tool mapping table; When the pneumatic ash conveying system starts to work, control the preset monitoring tool to real-time monitor the running state of the corresponding matching link.

3. The method according to claim 1, characterized in that, In step 2, based on the real-time monitoring result, determine the abnormal information of the corresponding matching link, and pre-analyze the abnormal information to lock the abnormal position of the corresponding matching link, which comprises: Collect the first monitoring data of the pneumatic ash conveying system during normal operation, and extract the normal operation period of each link in the pneumatic ash conveying system and the parameter fluctuation range of the normal working parameters under different normal operation periods based on the first monitoring data; Collect real-time monitoring data of the pneumatic ash conveying system during actual operation, extract actual operation time periods of each link in the pneumatic ash conveying system and actual working parameters under different actual operation time periods based on the real-time monitoring data; Compare the parameter fluctuation range of the normal working parameters and the actual working parameters under the same link; Calculate the abnormal value of the comparison result; wherein, represents the ith actual working parameter under the same link, wherein the parameter fluctuation range of the normal working parameter matched with the ith actual working parameter is [a1, a2]; represents the abnormal factor of the ith actual working parameter under the same link; represents the abnormal adjustment coefficient corresponding to the ith actual working parameter under the same link when the ith actual working parameter is less than ; represents the abnormal adjustment coefficient corresponding to the ith actual working parameter under the same link when the ith actual working parameter is greater than 1; represents the second abnormal value of the comparison result under the same link; represents the number of actual working parameters under the same link; represents the parameter weight of the ith actual working parameter under the same link, and ; Determine whether the normal operation time period and the actual operation time period under the same link are consistent, if consistent, determine whether the abnormal value is greater than a preset value; If greater, it is determined that the same link exists an abnormality, and abnormal parameters satisfying and are selected from all as abnormal information of the same link, wherein a3 represents a first abnormal threshold; a4 represents a second abnormal threshold; Determine the position matched with the abnormal information as the abnormal position of the corresponding matching link according to the abnormal combination-position mapping table.

4. The method according to claim 1, characterized in that, Input the abnormal information of the corresponding matching link into the matched self-diagnosis alarm model for fault early warning, including: Determine the abnormal reason set of each abnormal parameter involved in the abnormal information and the parameter association relationship between each abnormal parameter and the remaining abnormal parameters based on the self-diagnosis alarm model, and extract the first abnormal reason from the abnormal reason set; Output the early warning information consistent with the first abnormal reason based on the self-diagnosis alarm model.

5. The method according to claim 1, characterized in that, Determine the fault index and fault type corresponding to each early warning information in the fault early warning based on the early warning-fault database, including: Compare each early warning information in the fault early warning with the early warning-fault database respectively; Get the fault index and fault type corresponding to the corresponding early warning information based on the comparison result.

6. The method according to claim 1, characterized in that, Associate the comprehensive solution with the abnormal position and issue it to the artificial end for intervention, including: Establish a relationship array of the comprehensive solution and the abnormal position, where the relationship array=[comprehensive solution, abnormal position]; Issue the relationship array to the artificial end for intervention.

Images