A Fault Diagnosis System for Drum Granulator

Abstract

The invention discloses a fault diagnosis system for a drum granulator, relates to the technical field of fault diagnosis, and solves the technical problems that it is difficult to quickly locate the faulty parts of the drum granulator, it is also difficult to detect the internal data of the equipment, and it is not convenient to quickly locate the fault nodes inside the drum granulator; the parts in the component area are divided into a plurality of similar fault data part groups through a first fault diagnosis module, the fault data of the similar fault data part groups are statistically analyzed, and similar fault analysis models of each component area are established; the parts are grouped according to similar faults, so as to more accurately determine the location of the fault, which is helpful for quickly locating and repairing the fault; at the same time, by analyzing the similar fault data, the potential faults of the parts can be discovered in time; by collecting the operation data of the component area and performing group analysis, it is helpful to deeply understand the operation status of the equipment and determine the scope of the fault.

Description

Technical Field

[0001] The invention belongs to the field of fault diagnosis, in particular to a fault diagnosis system for a drum granulator. Background Art

[0002] The drum granulator is a commonly used granulation equipment, which is widely used in the granulation process of the chemical, food, pharmaceutical and other industries. Due to the complex operating environment, numerous equipment components, and frequent failures, once a failure occurs, it will lead to production stagnation, affecting production plans and production efficiency. Therefore, the development of a set of fault diagnosis system for the drum granulator can effectively improve the operating reliability of the equipment, reduce the cost of fault repair, reduce equipment downtime, and improve production efficiency and product quality. The fault diagnosis system of the drum granulator combines the intelligent system of advanced technologies such as the Internet of Things, big data analysis, and artificial intelligence. It can realize the functions of remote monitoring, fault warning, and rapid fault location, providing production enterprises with more intelligent and efficient management and maintenance methods.

[0003] Most of the existing fault diagnosis systems for drum granulators judge the faulty parts by the similarity of the overall fault data. There is a lack of group detection of different modules of the drum granulator. There is also a lack of grouping of parts with similar fault data in component areas. It is difficult to quickly locate the faulty parts of the drum granulator, and it is also difficult to detect the internal data of the equipment, making it inconvenient to quickly locate the faulty nodes inside the drum granulator. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a fault diagnosis system for a drum granulator, which is used to solve the technical problems that it is difficult to quickly locate the faulty parts of the drum granulator, it is also difficult to detect the internal data of the equipment, and it is not convenient to quickly locate the faulty nodes inside the drum granulator.

[0005] In order to solve the above problems, the first aspect of the present invention provides a fault diagnosis system for a drum granulator, comprising:

[0006] Regional data acquisition module: divides the components of the drum granulator into drum component area, motor transmission component area, hydraulic control component area, power supply component area and cooling component area, and monitors the operating data of each component area;

[0007] Node setting module: obtain the historical data of failures in different component areas of the drum granulator, and use the component with the highest probability of failure in the component area as the fault detection node;

[0008] The first fault diagnosis module: divides the parts in the component area into several similar fault data part groups, collects statistics on the fault data of the similar fault data part groups, establishes similar fault analysis models for each component area, analyzes the collected operation data of the component area through the model, and groups the collected operation data into corresponding similar fault data part groups;

[0009] The second fault diagnosis module: establishes a fault location model for each component area, and determines the faulty parts of the corresponding component area according to the fault detection data of the fault detection node of the component area;

[0010] Fault location module: compares the similar fault data parts grouping corresponding to the operation data of the first fault diagnosis module with the faulty parts of the corresponding component area determined by the second fault diagnosis module, and screens out the overlapping parts as high-risk fault parts.

[0011] As a further solution of the present invention: the regional data acquisition module monitors the operating data of each component area including the following steps:

[0012] For the drum component area and motor transmission component area, monitor vibration, temperature and speed data;

[0013] For the hydraulic control component area, monitor the oil pressure and oil temperature data;

[0014] For the power supply component area, monitor the power and voltage data of each component area;

[0015] For cooling component areas, monitor cooling water flow, pressure and temperature.

[0016] As a further solution of the present invention: the first fault diagnosis module includes:

[0017] Similarity analysis unit: obtaining the fault monitoring data of the corresponding component area when each part in different component areas fails in the historical data, and evaluating the similarity of the fault monitoring data in the component area according to the fault monitoring data;

[0018] Parts grouping unit: according to the similarity of the fault monitoring data in the component area, the parts with similarity evaluation values ​​less than a preset threshold are grouped into the same similar fault data parts group, and the parts in the component area are divided into several similar fault data parts groups;

[0019] Similar fault analysis model building unit: statistics the fault data of similar fault data parts grouping in the component area, and establish similar fault analysis models for each component area;

[0020] Fault detection unit: Through similar fault analysis model, similar fault analysis is performed on the operating data collected in real time from each component area, and based on the similar fault analysis results, it is determined whether there is a fault in each component area of ​​the drum granulator;

[0021] Fault location unit: If a fault occurs, the fault data of the faulty component area is grouped into corresponding similar fault data parts groups according to the similar fault analysis results of the faulty component area.

[0022] As a further solution of the present invention: evaluating the similarity of the fault monitoring data in the component area according to the fault monitoring data comprises the following steps:

[0023] According to the fault monitoring data of the corresponding component area when each part fails in the component area, the similarity of the fault monitoring data in the component area is evaluated by evaluating the similarity of the fault monitoring data of any two parts in the component area;

[0024] For the drum component area and the motor transmission component area, the similarity of the fault monitoring data between the parts in the component area is evaluated by the following formula:

[0025]

[0026] Among them, K 1 For the drum component area or the motor transmission component area, the similarity of the fault monitoring data between the parts, a 1 and a 2 is the vibration data of any two parts fault monitoring data in the drum component area or the motor transmission component area, b 1 and b 2 is the temperature data of any two parts fault monitoring data in the drum component area or the motor transmission component area, c 1 and c 2 The rotation speed data of any two parts fault monitoring data in the drum component area or the motor transmission component area;

[0027] For the hydraulic control component area, the similarity of the fault monitoring data in the component area is evaluated by the following formula:

[0028]

[0029] Among them, K 2 is the hydraulic control component area, the similarity of fault monitoring data between parts, d 1 and d 2 is the oil pressure data of any two parts fault monitoring data in the hydraulic control component area, f 1 and f 2 The oil temperature data of any two parts fault monitoring data in the hydraulic control component area;

[0030] For the power supply component area, the similarity of the fault monitoring data in the component area is evaluated by the following formula:

[0031]

[0032] Among them, K 3 is the power supply component area, the similarity of fault monitoring data between parts, p 1 and p 2 is the power data of the fault monitoring data of any two parts in the hydraulic control component area, u 1 and u 2 The voltage data of the fault monitoring data of any two parts in the hydraulic control component area;

[0033] For the cooling component area, the similarity of the fault monitoring data in the component area is evaluated by the following formula:

[0034]

[0035] Among them, K 4 is the similarity of fault monitoring data between parts in the cooling component area, g 1 and g 2 is the cooling water flow data of the fault monitoring data of any two parts in the cooling component area, h 1 and h 2 is the cooling water pressure data of the fault monitoring data of any two parts in the cooling component area, t 1 and t 2 It is the cooling water temperature data of the fault monitoring data of any two parts in the cooling component area.

[0036] As a further solution of the present invention: according to the similarity of the fault monitoring data in the component area, the parts with similarity evaluation values ​​less than a preset threshold are grouped into the same similar fault data part group, and the parts in the component area are divided into several similar fault data part groups, including the following steps:

[0037] According to the similarity of the fault monitoring data in the component area, for the drum component area and the motor transmission component area, when the similarity of the fault monitoring data is greater than 0.7, the fault monitoring data is divided into the same similar fault data part group;

[0038] For the hydraulic control component area, when the similarity of fault monitoring data is greater than 0.6, the corresponding parts are divided into the same similar fault data parts group;

[0039] For the power supply component area, when the similarity of the fault monitoring data is greater than 0.6, the corresponding parts are divided into the same similar fault data part group;

[0040] For the cooling component area, when the similarity of the fault monitoring data is greater than 0.7, the corresponding parts are divided into the same similar fault data part group.

[0041] As a further solution of the present invention, statistics are collected on the failure data of the parts grouped by similar failure data in the component area, and a similar failure analysis model for each component area is established, including the following steps:

[0042] Performing statistics on the failure data of the parts grouped with similar failure data in the component area, and calculating the average number of the failure data of the parts grouped with similar failure data in each component area;

[0043] According to the evaluation formula of the similarity of fault monitoring data in the component area, the similarity of fault monitoring data between the real-time collected operation data of each component area and the average number of fault data of the similar fault data part grouping in the corresponding component area is calculated, and a similar fault analysis model for each component area is established.

[0044] As a further solution of the present invention: by using a similar fault analysis model, similar fault analysis is performed on the operating data collected in real time from each component area, and according to the similar fault analysis results, it is determined whether each component area of ​​the drum granulator has a fault, including the following steps:

[0045] By establishing similar fault analysis models for each component area, the similarity of fault monitoring data between the real-time collected operation data and the average number of fault data of each similar fault data part group in the corresponding component area is calculated, and similar fault analysis is performed on the real-time collected operation data;

[0046] Obtain similar fault analysis results. If the fault monitoring data similarity between the real-time collected operation data and the average number of fault data of the similar fault data part group in the corresponding component area is greater than 0.5, the corresponding similar fault data part group is screened out, the fault is located in the screened similar fault data part group, and the real-time collected operation data is used as the fault monitoring data. Otherwise, no division is performed.

[0047] If the fault monitoring data similarity between the real-time collected operating data and the average number of fault data of each similar fault data part group in the corresponding component area is less than or equal to 0.5, the similar fault analysis model of the component area determines that there is no fault in each component area of ​​the drum granulator.

[0048] As a further solution of the present invention: the second fault diagnosis module establishes a fault location model for each component area, and determines the faulty parts of the corresponding component area according to the fault detection data of the fault detection node of the component area, including the following steps:

[0049] Obtain historical data of failures in different component areas of the drum granulator, use the component with the highest probability of failure in the component area as a fault detection node, and detect vibration, temperature and pressure data at the fault detection node;

[0050] Numbering other components in different component areas in sequence according to their distances from the fault detection node;

[0051] For the drum component area, motor transmission component area, hydraulic control component area and cooling component area, according to the number of the parts in each component area, the vibration, temperature and pressure data of the fault detection node when each component in the component area fails are detected in turn;

[0052] For the power supply component area, when each component in the component area fails, the current and voltage data of the fault detection node are detected in turn;

[0053] Correlate the detection data of each component detected by the fault detection node with the corresponding part number, and establish a fault location model for each detection component area;

[0054] According to the real-time fault detection data of the fault detection node in the component area, the part number is screened through the established fault location model, and the screened part number is determined as the faulty part number of the corresponding component area.

[0055] As a further solution of the present invention, the detection data of each component detected by the fault detection node when the fault occurs is correlated with the corresponding part number, and a fault location model for each detection component area is established, including the following steps:

[0056] When each component fails, the detection data of the fault detection node is correlated with the corresponding part number;

[0057] For the drum component area, motor transmission component area, hydraulic control component area and cooling component area, the fault location model is established by the following formula. According to the number of the parts in the component area, the deviation between the real-time fault detection data of the fault detection node in the component area and the fault detection data when each part in the component area fails is calculated in sequence:

[0058]

[0059] Among them, E 1 is the deviation between the real-time fault detection data in the drum component area, motor transmission component area, hydraulic control component area and cooling component area and the fault detection data when the i-th part in the component area fails. F is the vibration data of the real-time fault detection data of the fault detection node in the component area. F i is the vibration data in the fault detection data when the i-th part fails, T is the temperature data of the real-time fault detection data of the fault detection node in the component area, and T i is the temperature data in the fault detection data when the i-th part fails, P is the pressure data of the real-time fault detection data of the fault detection node in the component area, and P iis the pressure data in the fault detection data when the i-th part fails;

[0060] When the deviation between the real-time fault detection data and the fault detection data when each part in the component area fails is less than 0.3, the corresponding part number is selected in the drum component area, the motor transmission component area, the hydraulic control component area and the cooling component area;

[0061] For the power supply component area, the fault location model is established by the following formula. According to the number of the parts in the component area, the deviation between the real-time fault detection data of the fault detection node in the component area and the fault detection data when each part in the component area fails is calculated in sequence:

[0062]

[0063] Among them, E 2 is the deviation between the real-time fault detection data in the power supply component area and the fault detection data when the i-th part in the component area fails, I is the current data of the real-time fault detection data of the fault detection node in the component area, and I i is the current data in the fault detection data when the i-th part fails, U is the voltage data of the real-time fault detection data of the fault detection node in the component area, and U i is the voltage data in the fault detection data when the i-th component fails;

[0064] When the deviation between the real-time fault detection data screened out in the drum component area, motor transmission component area, hydraulic control component area and cooling component area and the fault detection data when each part in the component area fails is less than 0.2, the corresponding part number.

[0065] As a further solution of the present invention: the fault location module compares the similar fault data part grouping corresponding to the operation data of the first fault diagnosis module with the faulty parts of the corresponding component area determined by the second fault diagnosis module, and screens out the overlapping parts as high-risk faulty parts, including the following steps:

[0066] The parts with similar fault data corresponding to the operation data of the first fault diagnosis module are grouped as a set, and the parts corresponding to the part numbers of the corresponding component areas determined by the second fault diagnosis module are grouped as another set. The overlapping parts of the two sets are compared, and the parts corresponding to the overlapping parts of the two sets are grouped as high-risk fault parts, and the parts in the non-overlapping parts of the two sets are grouped as medium-risk fault parts.

[0067] Compared with the prior art, the present invention has the following beneficial effects:

[0068] The present invention divides the parts in the component area into several similar fault data part groups through the first fault diagnosis module, collects the fault data of the similar fault data part groups, establishes similar fault analysis models for each component area, analyzes the collected operation data of the component area through the model, and groups them into corresponding similar fault data part groups; grouping the parts according to similar faults facilitates more accurate determination of the fault location, and helps to quickly locate and repair the fault; at the same time, by analyzing the similar fault data, the potential failure mode of the parts can be discovered in time, and then preventive maintenance can be implemented to avoid the occurrence of faults and improve equipment reliability; by collecting the operation data of the component area and performing group analysis, it is helpful to have an in-depth understanding of the operation status of the equipment and determine the scope of the fault; at the same time, through model analysis and grouping, the speed of fault diagnosis is accelerated.

[0069] The present invention obtains historical data of faults occurring in different component areas of a drum granulator through a node setting module, and uses the component with the highest probability of failure in the component area as a fault detection node; a second fault diagnosis module establishes a fault location model for each component area, and determines the faulty parts of the corresponding component area according to the fault detection data of the fault detection node in the component area; by using the component with the highest probability of failure in the component area as a fault detection node, data collection is performed, and faults inside the equipment are usually detected faster than the data of the entire component area; at the same time, by establishing a fault location model for each component area and analyzing the fault detection data of the fault detection node in the component area, it is convenient to timely discover potential fault points. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0071] Figure 1 It is a schematic diagram of the system framework of the present invention;

[0072] Figure 2 It is a schematic diagram of the framework of the first fault diagnosis module of the present invention. DETAILED DESCRIPTION

[0073] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0074] See also Figure 1-Figure 2 The first embodiment of the present invention provides a fault diagnosis system for a drum granulator, comprising:

[0075] Regional data acquisition module: divides the components of the drum granulator into drum component area, motor transmission component area, hydraulic control component area, power supply component area and cooling component area, and monitors the operating data of each component area;

[0076] Node setting module: obtain the historical data of failures in different component areas of the drum granulator, and use the component with the highest probability of failure in the component area as the fault detection node;

[0077] The first fault diagnosis module: divides the parts in the component area into several similar fault data part groups, collects statistics on the fault data of the similar fault data part groups, establishes similar fault analysis models for each component area, analyzes the collected operation data of the component area through the model, and groups the collected operation data into corresponding similar fault data part groups;

[0078] The second fault diagnosis module: establishes a fault location model for each component area, and determines the faulty parts of the corresponding component area according to the fault detection data of the fault detection node of the component area;

[0079] Fault location module: compares the similar fault data parts grouping corresponding to the operation data of the first fault diagnosis module with the faulty parts of the corresponding component area determined by the second fault diagnosis module, and screens out the overlapping parts as high-risk fault parts.

[0080] Specifically, this embodiment divides the components of the drum granulator into a drum component area, a motor transmission component area, a hydraulic control component area, a power supply component area and a cooling component area through a regional data acquisition module, and monitors the operating data of each component area; and divides the parts in the component area into several similar fault data part groups through a first fault diagnosis module, performs statistics on the fault data of the similar fault data part groups, establishes similar fault analysis models for each component area, analyzes the collected operating data of the component area through the model, and groups them into corresponding similar fault data part groups; by grouping the component areas and establishing similar fault analysis models, similar fault data part groups with faults can be diagnosed more quickly, which is convenient for narrowing the scope of troubleshooting of faulty parts; and by dividing the components into different similar fault data part groups for statistics and analysis, the specific area where the fault occurs can be located more accurately.

[0081] The first fault diagnosis module divides the parts in the component area into several similar fault data part groups, collects the fault data of the similar fault data part groups, and establishes similar fault analysis models for each component area. The collected operation data of the component area is analyzed by the model and grouped into corresponding similar fault data part groups. Grouping the parts according to similar faults facilitates more accurate determination of the fault location, which helps to quickly locate and repair the fault. At the same time, by analyzing similar fault data, the potential failure mode of the parts can be discovered in time, and then preventive maintenance can be implemented to avoid the occurrence of faults and improve equipment reliability. By collecting the operation data of the component area and performing group analysis, it is helpful to have a deep understanding of the operation status of the equipment and determine the scope of the fault. At the same time, through model analysis and grouping, the speed of fault diagnosis is accelerated.

[0082] The node setting module is used to obtain historical data on faults in different component areas of the drum granulator, and the component with the highest probability of failure in the component area is used as the fault detection node; the second fault diagnosis module establishes a fault location model for each component area, and determines the faulty parts in the corresponding component area based on the fault detection data of the fault detection node in the component area; by using the component with the highest probability of failure in the component area as the fault detection node, data collection is performed, which usually detects faults inside the equipment faster than the data of the entire component area; at the same time, by establishing a fault location model for each component area and analyzing the fault detection data of the fault detection node in the component area, it is convenient to find potential fault points in time, so as to avoid the inability to determine the potential fault point when the first fault diagnosis module fails to detect the faulty part, and to perform preventive maintenance, thereby improving the reliability and stability of the equipment.

[0083] The fault location module compares the similar fault data parts grouping corresponding to the operation data of the first fault diagnosis module with the faulty parts of the corresponding component area determined by the second fault diagnosis module, and selects the overlapping parts as high-risk faulty parts; through comparison and analysis, it is convenient to quickly select the overlapping faulty parts and accurately locate the high-risk faulty parts, saving time and energy for troubleshooting. By identifying high-risk faulty parts, these parts can be monitored and warned in advance, reducing the probability of faults and improving the reliability and stability of the equipment. Targeted maintenance and inspection of high-risk faulty parts, giving priority to parts with potential risks, avoiding the occurrence of faults, and reducing losses caused by faults.

[0084] In one embodiment of the present invention, the regional data acquisition module monitors the operation data of each component area including the following steps:

[0085] For the drum component area and motor transmission component area, monitor vibration, temperature and speed data;

[0086] For the hydraulic control component area, monitor the oil pressure and oil temperature data;

[0087] For the power supply component area, monitor the power and voltage data of each component area;

[0088] For cooling component areas, monitor cooling water flow, pressure and temperature.

[0089] In one embodiment of the present invention, the first fault diagnosis module includes:

[0090] Similarity analysis unit: obtaining the fault monitoring data of the corresponding component area when each part in different component areas fails in the historical data, and evaluating the similarity of the fault monitoring data in the component area according to the fault monitoring data;

[0091] Parts grouping unit: according to the similarity of the fault monitoring data in the component area, the parts with similarity evaluation values ​​less than a preset threshold are grouped into the same similar fault data parts group, and the parts in the component area are divided into several similar fault data parts groups;

[0092] Similar fault analysis model building unit: statistics the fault data of similar fault data parts grouping in the component area, and establish similar fault analysis models for each component area;

[0093] Fault detection unit: Through similar fault analysis model, similar fault analysis is performed on the operating data collected in real time from each component area, and based on the similar fault analysis results, it is determined whether there is a fault in each component area of ​​the drum granulator;

[0094] Fault location unit: If a fault occurs, the fault data of the faulty component area is grouped into corresponding similar fault data parts groups according to the similar fault analysis results of the faulty component area.

[0095] In one embodiment of the present invention, evaluating the similarity of fault monitoring data in a component area according to the fault monitoring data comprises the following steps:

[0096] According to the fault monitoring data of the corresponding component area when each part fails in the component area, the similarity of the fault monitoring data in the component area is evaluated by evaluating the similarity of the fault monitoring data of any two parts in the component area;

[0097] For the drum component area and the motor transmission component area, the similarity of the fault monitoring data between the parts in the component area is evaluated by the following formula:

[0098]

[0099] Among them, K 1 For the drum component area or the motor transmission component area, the similarity of the fault monitoring data between the parts, a 1 and a2 is the vibration data of any two parts fault monitoring data in the drum component area or the motor transmission component area, b 1 and b 2 is the temperature data of any two parts fault monitoring data in the drum component area or the motor transmission component area, c 1 and c 2 The rotation speed data of any two parts fault monitoring data in the drum component area or the motor transmission component area;

[0100] For the hydraulic control component area, the similarity of the fault monitoring data in the component area is evaluated by the following formula:

[0101]

[0102] Among them, K 2 is the hydraulic control component area, the similarity of fault monitoring data between parts, d 1 and d 2 is the oil pressure data of any two parts fault monitoring data in the hydraulic control component area, f 1 and f 2 The oil temperature data of any two parts fault monitoring data in the hydraulic control component area;

[0103] For the power supply component area, the similarity of the fault monitoring data in the component area is evaluated by the following formula:

[0104]

[0105] Among them, K 3 is the power supply component area, the similarity of fault monitoring data between parts, p 1 and p 2 is the power data of the fault monitoring data of any two parts in the hydraulic control component area, u 1 and u 2 The voltage data of the fault monitoring data of any two parts in the hydraulic control component area;

[0106] For the cooling component area, the similarity of the fault monitoring data in the component area is evaluated by the following formula:

[0107]

[0108] Among them, K 4 is the similarity of fault monitoring data between parts in the cooling component area, g 1 and g 2 is the cooling water flow data of the fault monitoring data of any two parts in the cooling component area, h 1 and h 2is the cooling water pressure data of the fault monitoring data of any two parts in the cooling component area, t 1 and t 2 It is the cooling water temperature data of the fault monitoring data of any two parts in the cooling component area.

[0109] In one embodiment of the present invention, according to the similarity of the fault monitoring data in the component area, the parts with similarity evaluation values ​​less than a preset threshold are grouped into the same similar fault data part group, and the parts in the component area are divided into several similar fault data part groups, including the following steps:

[0110] According to the similarity of the fault monitoring data in the component area, for the drum component area and the motor transmission component area, when the similarity of the fault monitoring data is greater than 0.7, the fault monitoring data is divided into the same similar fault data part group;

[0111] For the hydraulic control component area, when the similarity of fault monitoring data is greater than 0.6, the corresponding parts are divided into the same similar fault data parts group;

[0112] For the power supply component area, when the similarity of the fault monitoring data is greater than 0.6, the corresponding parts are divided into the same similar fault data part group;

[0113] For the cooling component area, when the similarity of the fault monitoring data is greater than 0.7, the corresponding parts are divided into the same similar fault data part group.

[0114] In one embodiment of the present invention, statistics are collected on the failure data of the parts grouped by similar failure data in the component area to establish a similar failure analysis model for each component area, including the following steps:

[0115] Performing statistics on the failure data of the parts grouped with similar failure data in the component area, and calculating the average number of the failure data of the parts grouped with similar failure data in each component area;

[0116] According to the evaluation formula of the similarity of fault monitoring data in the component area, the similarity of fault monitoring data between the real-time collected operation data of each component area and the average number of fault data of the similar fault data part grouping in the corresponding component area is calculated, and a similar fault analysis model for each component area is established.

[0117] In one embodiment of the present invention, similar fault analysis is performed on the operating data collected in real time from each component area through a similar fault analysis model, and according to the similar fault analysis results, it is determined whether each component area of ​​the drum granulator has a fault, including the following steps:

[0118] By establishing similar fault analysis models for each component area, the similarity of fault monitoring data between the real-time collected operation data and the average number of fault data of each similar fault data part group in the corresponding component area is calculated, and similar fault analysis is performed on the real-time collected operation data;

[0119] Obtain similar fault analysis results. If the fault monitoring data similarity between the real-time collected operation data and the average number of fault data of the similar fault data part group in the corresponding component area is greater than 0.5, the corresponding similar fault data part group is screened out, the fault is located in the screened similar fault data part group, and the real-time collected operation data is used as the fault monitoring data. Otherwise, no division is performed.

[0120] If the fault monitoring data similarity between the real-time collected operating data and the average number of fault data of each similar fault data part group in the corresponding component area is less than or equal to 0.5, the similar fault analysis model of the component area determines that there is no fault in each component area of ​​the drum granulator.

[0121] In one embodiment of the present invention, the second fault diagnosis module establishes a fault location model for each component area, and determines the faulty parts of the corresponding component area according to the fault detection data of the fault detection node of the component area, including the following steps:

[0122] Obtain historical data of failures in different component areas of the drum granulator, use the component with the highest probability of failure in the component area as a fault detection node, and detect vibration, temperature and pressure data at the fault detection node;

[0123] Numbering other components in different component areas in sequence according to their distances from the fault detection node;

[0124] For the drum component area, motor transmission component area, hydraulic control component area and cooling component area, according to the number of the parts in each component area, the vibration, temperature and pressure data of the fault detection node when each component in the component area fails are detected in turn;

[0125] For the power supply component area, when each component in the component area fails, the current and voltage data of the fault detection node are detected in turn;

[0126] Correlate the detection data of each component detected by the fault detection node with the corresponding part number, and establish a fault location model for each detection component area;

[0127] According to the real-time fault detection data of the fault detection node in the component area, the part number is screened through the established fault location model, and the screened part number is determined as the fault part number of the corresponding component area.

[0128] In one embodiment of the present invention, the detection data of each component detected by the fault detection node when the fault occurs is correlated with the corresponding part number, and a fault location model for each detection component area is established, including the following steps:

[0129] When each component fails, the detection data of the fault detection node is correlated with the corresponding part number;

[0130] For the drum component area, motor transmission component area, hydraulic control component area and cooling component area, the fault location model is established by the following formula. According to the number of the parts in the component area, the deviation between the real-time fault detection data of the fault detection node in the component area and the fault detection data when each part in the component area fails is calculated in turn:

[0131]

[0132] Among them, E 1 is the deviation between the real-time fault detection data in the drum component area, motor transmission component area, hydraulic control component area and cooling component area and the fault detection data when the i-th part in the component area fails. F is the vibration data of the real-time fault detection data of the fault detection node in the component area. F i is the vibration data in the fault detection data when the i-th part fails, T is the temperature data of the real-time fault detection data of the fault detection node in the component area, and T i is the temperature data in the fault detection data when the i-th part fails, P is the pressure data of the real-time fault detection data of the fault detection node in the component area, and P i is the pressure data in the fault detection data when the i-th part fails;

[0133] When the deviation between the real-time fault detection data and the fault detection data when each part in the component area fails is less than 0.3, the corresponding part number is selected in the drum component area, the motor transmission component area, the hydraulic control component area and the cooling component area;

[0134] For the power supply component area, the fault location model is established by the following formula. According to the number of the parts in the component area, the deviation between the real-time fault detection data of the fault detection node in the component area and the fault detection data when each part in the component area fails is calculated in sequence:

[0135]

[0136] Among them, E 2 is the deviation between the real-time fault detection data in the power supply component area and the fault detection data when the i-th part in the component area fails, I is the current data of the real-time fault detection data of the fault detection node in the component area, and I iis the current data in the fault detection data when the i-th part fails, U is the voltage data of the real-time fault detection data of the fault detection node in the component area, and U i is the voltage data in the fault detection data when the i-th component fails;

[0137] When the deviation between the real-time fault detection data screened out in the drum component area, motor transmission component area, hydraulic control component area and cooling component area and the fault detection data when each part in the component area fails is less than 0.2, the corresponding part number.

[0138] In one embodiment of the present invention, the fault location module compares the similar fault data part group corresponding to the operation data of the first fault diagnosis module with the faulty parts of the corresponding component area determined by the second fault diagnosis module, and selects the overlapping parts as high-risk faulty parts, including the following steps:

[0139] The parts with similar fault data corresponding to the operation data of the first fault diagnosis module are grouped as a set, and the parts corresponding to the part numbers of the corresponding component areas determined by the second fault diagnosis module are grouped as another set. The overlapping parts of the two sets are compared, and the parts corresponding to the overlapping parts of the two sets are grouped as high-risk fault parts, and the parts in the non-overlapping parts of the two sets are grouped as medium-risk fault parts.

[0140] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A fault diagnosis system for a drum granulator, characterized in that: include: Regional data acquisition module: divides the components of the drum granulator into drum component area, motor transmission component area, hydraulic control component area, power supply component area and cooling component area, and monitors the operating data of each component area; Node setting module: obtain the historical data of failures in different component areas of the drum granulator, and use the component with the highest probability of failure in the component area as the fault detection node; The first fault diagnosis module: divides the parts in the component area into several similar fault data part groups, collects statistics on the fault data of the similar fault data part groups, establishes similar fault analysis models for each component area, analyzes the collected operation data of the component area through the model, and groups the collected operation data into corresponding similar fault data part groups; The second fault diagnosis module: establishes a fault location model for each component area, and determines the faulty parts of the corresponding component area according to the fault detection data of the fault detection node of the component area; Fault location module: compares the similar fault data parts group corresponding to the operation data of the first fault diagnosis module with the fault parts of the corresponding component area determined by the second fault diagnosis module, and selects the overlapping parts as high-risk fault parts; Wherein, the first fault diagnosis module includes: Similarity analysis unit: obtaining the fault monitoring data of the corresponding component area when each part in different component areas fails in the historical data, and evaluating the similarity of the fault monitoring data in the component area according to the fault monitoring data; Parts grouping unit: according to the similarity of the fault monitoring data in the component area, the parts with similarity evaluation values ​​less than a preset threshold are grouped into the same similar fault data parts group, and the parts in the component area are divided into several similar fault data parts groups; Similar fault analysis model building unit: statistics the fault data of similar fault data parts grouping in the component area, and establish similar fault analysis models for each component area; Fault detection unit: Through similar fault analysis model, similar fault analysis is performed on the operating data collected in real time from each component area, and based on the similar fault analysis results, it is determined whether there is a fault in each component area of ​​the drum granulator; Fault location unit: If a fault occurs, the fault data of the faulty component area is grouped into corresponding similar fault data parts groups according to the similar fault analysis results of the faulty component area.

2. A fault diagnosis system for a drum granulator according to claim 1, characterized in that: The regional data acquisition module monitors the operation data of each component area and includes the following steps: For the drum component area and motor transmission component area, monitor vibration, temperature and speed data; For the hydraulic control component area, monitor the oil pressure and oil temperature data; For the power supply component area, monitor the power and voltage data of each component area; For cooling component areas, monitor cooling water flow, pressure and temperature.

3. A fault diagnosis system for a drum granulator according to claim 1, characterized in that: Evaluating the similarity of the fault monitoring data in the component area according to the fault monitoring data includes the following steps: According to the fault monitoring data of the corresponding component area when each part fails in the component area, the similarity of the fault monitoring data in the component area is evaluated by evaluating the similarity of the fault monitoring data of any two parts in the component area; For the drum component area and the motor transmission component area, the similarity of the fault monitoring data between the parts in the component area is evaluated by the following formula: Wherein, K1 is the similarity of fault monitoring data between parts in the drum component area or the motor transmission component area, a1 and a2 are the vibration data of any two parts fault monitoring data in the drum component area or the motor transmission component area, b1 and b2 are the temperature data of any two parts fault monitoring data in the drum component area or the motor transmission component area, c1 and c2 are the speed data of any two parts fault monitoring data in the drum component area or the motor transmission component area; For the hydraulic control component area, the similarity of the fault monitoring data in the component area is evaluated by the following formula: Among them, K2 is the similarity of fault monitoring data between parts in the hydraulic control component area, d1 and d2 are the oil pressure data of any two parts fault monitoring data in the hydraulic control component area, and f1 and f2 are the oil temperature data of any two parts fault monitoring data in the hydraulic control component area; For the power supply component area, the similarity of the fault monitoring data in the component area is evaluated by the following formula: Among them, K3 is the similarity of fault monitoring data between parts in the power supply component area, p1 and p2 are the power data of fault monitoring data of any two parts in the hydraulic control component area, and u1 and u2 are the voltage data of fault monitoring data of any two parts in the hydraulic control component area; For the cooling component area, the similarity of the fault monitoring data in the component area is evaluated by the following formula: Among them, K4 is the similarity of fault monitoring data between parts in the cooling component area, g1 and g2 are the cooling water flow data of the fault monitoring data of any two parts in the cooling component area, h1 and h2 are the cooling water pressure data of the fault monitoring data of any two parts in the cooling component area, t1 and t2 are the cooling water temperature data of the fault monitoring data of any two parts in the cooling component area.

4. A fault diagnosis system for a drum granulator according to claim 1, characterized in that: According to the similarity of the fault monitoring data in the component area, the parts with similarity evaluation values ​​less than a preset threshold are grouped into the same similar fault data part group, and the parts in the component area are divided into several similar fault data part groups, including the following steps: According to the similarity of the fault monitoring data in the component area, for the drum component area and the motor transmission component area, when the similarity of the fault monitoring data is greater than 0.7, the fault monitoring data is divided into the same similar fault data part group; For the hydraulic control component area, when the similarity of fault monitoring data is greater than 0.6, the corresponding parts are divided into the same similar fault data parts group; For the power supply component area, when the similarity of the fault monitoring data is greater than 0.6, the corresponding parts are divided into the same similar fault data part group; For the cooling component area, when the similarity of the fault monitoring data is greater than 0.7, the corresponding parts are divided into the same similar fault data part group.

5. A fault diagnosis system for a drum granulator according to claim 1, characterized in that: The failure data of the parts grouped by similar failure data in the component area are statistically analyzed to establish a similar failure analysis model for each component area, including the following steps: Performing statistics on the failure data of the parts grouped with similar failure data in the component area, and calculating the average number of the failure data of the parts grouped with similar failure data in each component area; According to the evaluation formula of the similarity of fault monitoring data in the component area, the similarity of fault monitoring data between the real-time collected operation data of each component area and the average number of fault data of the similar fault data part grouping in the corresponding component area is calculated, and a similar fault analysis model for each component area is established.

6. A fault diagnosis system for a drum granulator according to claim 1, characterized in that: Through the similar fault analysis model, the operation data collected in real time from each component area is analyzed for similar faults, and according to the similar fault analysis results, it is judged whether each component area of ​​the drum granulator has a fault, including the following steps: By establishing similar fault analysis models for each component area, the similarity of fault monitoring data between the real-time collected operation data and the average number of fault data of each similar fault data part group in the corresponding component area is calculated, and similar fault analysis is performed on the real-time collected operation data; Obtain similar fault analysis results. If the fault monitoring data similarity between the real-time collected operation data and the average number of fault data of the similar fault data part group in the corresponding component area is greater than 0.5, the corresponding similar fault data part group is screened out, the fault is located in the screened similar fault data part group, and the real-time collected operation data is used as the fault monitoring data. Otherwise, no division is performed. If the fault monitoring data similarity between the real-time collected operating data and the average number of fault data of each similar fault data part group in the corresponding component area is less than or equal to 0.5, the similar fault analysis model of the component area determines that there is no fault in each component area of ​​the drum granulator.

7. A fault diagnosis system for a drum granulator according to claim 1, characterized in that: The second fault diagnosis module establishes a fault location model for each component area, and determines the faulty parts of the corresponding component area according to the fault detection data of the fault detection node of the component area, including the following steps: Obtain historical data of failures in different component areas of the drum granulator, use the component with the highest probability of failure in the component area as a fault detection node, and detect vibration, temperature and pressure data at the fault detection node; Numbering other components in different component areas in sequence according to their distances from the fault detection node; For the drum component area, motor transmission component area, hydraulic control component area and cooling component area, according to the number of the parts in each component area, the vibration, temperature and pressure data of the fault detection node when each component in the component area fails are detected in turn; For the power supply component area, when each component in the component area fails, the current and voltage data of the fault detection node are detected in turn; Correlate the detection data of each component detected by the fault detection node with the corresponding part number, and establish a fault location model for each detection component area; According to the real-time fault detection data of the fault detection node in the component area, the part number is screened through the established fault location model, and the screened part number is determined as the fault part number of the corresponding component area.

8. A fault diagnosis system for a drum granulator according to claim 7, characterized in that: Correlating the detection data of each component detected by the fault detection node with the corresponding part number, and establishing a fault location model for each detection component area, including the following steps: When each component fails, the detection data of the fault detection node is correlated with the corresponding part number; For the drum component area, motor transmission component area, hydraulic control component area and cooling component area, the fault location model is established by the following formula. According to the number of the parts in the component area, the deviation between the real-time fault detection data of the fault detection node in the component area and the fault detection data when each part in the component area fails is calculated in turn: Among them, E1 is the deviation between the real-time fault detection data in the drum component area, motor transmission component area, hydraulic control component area and cooling component area and the fault detection data when the i-th part in the component area fails, F is the vibration data of the real-time fault detection data of the fault detection node in the component area, and F i is the vibration data in the fault detection data when the i-th part fails, T is the temperature data of the real-time fault detection data of the fault detection node in the component area, and T i is the temperature data in the fault detection data when the i-th part fails, P is the pressure data of the real-time fault detection data of the fault detection node in the component area, and P i is the pressure data in the fault detection data when the i-th part fails; When the deviation between the real-time fault detection data and the fault detection data when each part in the component area fails is less than 0.3, the corresponding part number is selected in the drum component area, the motor transmission component area, the hydraulic control component area and the cooling component area; For the power supply component area, the fault location model is established by the following formula. According to the number of the parts in the component area, the deviation between the real-time fault detection data of the fault detection node in the component area and the fault detection data when each part in the component area fails is calculated in sequence: Where E2 is the deviation between the real-time fault detection data in the power supply component area and the fault detection data when the i-th part in the component area fails, I is the current data of the real-time fault detection data of the fault detection node in the component area, and I i is the current data in the fault detection data when the i-th part fails, U is the voltage data of the real-time fault detection data of the fault detection node in the component area, and U i is the voltage data in the fault detection data when the i-th component fails; When the deviation between the real-time fault detection data screened out in the drum component area, motor transmission component area, hydraulic control component area and cooling component area and the fault detection data when each part in the component area fails is less than 0.2, the corresponding part number.

9. A fault diagnosis system for a drum granulator according to claim 1, characterized in that: The fault location module compares the similar fault data part group corresponding to the operation data of the first fault diagnosis module with the faulty parts of the corresponding component area determined by the second fault diagnosis module, and selects the overlapping parts as high-risk faulty parts, including the following steps: The parts with similar fault data corresponding to the operation data of the first fault diagnosis module are grouped as a set, and the parts corresponding to the part numbers of the corresponding component areas determined by the second fault diagnosis module are grouped as another set. The overlapping parts of the two sets are compared, and the parts corresponding to the overlapping parts of the two sets are grouped as high-risk fault parts, and the parts in the non-overlapping parts of the two sets are grouped as medium-risk fault parts.

Images