Intelligent lubricator lubricating system and control method

Abstract

The invention discloses an intelligent lubricator lubricating system, and relates to the technical field of lubricator control, and the intelligent lubricator lubricating system comprises a lubricator which is used for receiving a command issued by a system management platform and executing corresponding operation; the system management platform communicates with the lubricator and a user, and is responsible for receiving parameters set by the user, issuing a command to the lubricator, receiving field information uploaded by the lubricator and feeding back information to the user in real time; the user sets parameters through the system management platform and receives real-time information fed back by the system management platform, through analysis of the lubrication monitoring model, the lubrication state of the equipment can be accurately judged, the lubrication strategy can be adjusted in time, the defects of fixed time intervals and empirical lubrication in a traditional lubrication mode are overcome, the lubrication effect is improved, and the service life of the equipment is prolonged. Equipment faults caused by the lubrication problem are effectively reduced, excessive abrasion and damage of parts are avoided, the equipment fault rate is reduced, and the service life of equipment is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of lubricator control, and in particular to an intelligent lubricator lubrication system and a control method. Background Art

[0002] Lubricators are indispensable tools in the maintenance of mechanical equipment. Their core function is to reduce friction, wear, and overheat by providing lubricants (such as oil or grease) to moving parts, thereby extending equipment life and improving operating efficiency.

[0003] Traditional lubricator systems typically rely on lubrication at fixed intervals. However, this approach has some significant disadvantages:

[0004] 1. The fixed refueling interval cannot be flexibly adjusted, which easily leads to inaccurate lubricating oil quantity, thereby causing problems of over- or under-refueling and affecting equipment operation;

[0005] 2. Traditional lubricator systems fail to take into account the actual operating status of the equipment or environmental changes, and therefore cannot flexibly respond to different working conditions;

[0006] 3. Fixed refueling cycles often cannot be optimized, which may lead to waste of lubricating oil and cannot be reasonably adjusted according to the needs of the equipment;

[0007] Based on the above shortcomings, an intelligent lubricator lubrication system and control method are proposed. Summary of the Invention

[0008] The purpose of the present invention is to solve the problems in the prior art and to propose an intelligent lubricator lubrication system and a control method.

[0009] An intelligent lubricator lubrication system, comprising:

[0010] Lubricator: used to receive commands issued by the system management platform and perform corresponding operations;

[0011] System management platform: communicates with the lubricator and the user respectively, is responsible for receiving the parameters set by the user, issuing commands to the lubricator, receiving the field information uploaded by the lubricator, and providing real-time feedback to the user;

[0012] User: Set parameters through the system management platform and receive real-time information feedback from the system management platform.

[0013] Preferably, the command issued by the system management platform includes the lubrication cycle of the lubricator and the oil output of the oil cup.

[0014] A lubrication control method for an intelligent lubricator is also proposed, comprising the following steps:

[0015] S1. Collect real-time operating data of the lubricated equipment through sensors;

[0016] S2, building a lubrication monitoring model;

[0017] S3, using the historical operating data of the lubricated equipment as a training set to train the lubrication monitoring model;

[0018] S4. Input the collected real-time operation data into the trained lubrication monitoring model, output the feature vector Tensor1, input the normal operation data of the lubricated equipment into the trained lubrication monitoring model, output the feature vector Tensor2, and calculate the similarity θ between the two sets of vectors;

[0019] S5. Adjust the command to the lubricator according to the similarity θ.

[0020] Preferably, in step S1, the operating data includes the operating speed, the effective value of the operating vibration value and the oil output of the lubricator oil cup.

[0021] Preferably, in step S2, the lubrication monitoring model includes: an input layer for receiving input operation data;

[0022] At least one large block for processing input data;

[0023] A ReLU activation layer to introduce nonlinearity;

[0024] An SE-Large block for further processing of the data;

[0025] Output layer, used to output the processed feature vector;

[0026] Multiple large blocks and SE-Large blocks are arranged in a stacked manner, wherein the large block includes an expanded re-parameter block, at least one batch normalization layer, a feed-forward network and a residual connection, and the SE-Large block includes an expanded re-parameter block, at least one batch normalization layer and an SE module.

[0027] Preferably, in step S3, the historical operating data of the lubricated equipment includes normal state data and abnormal state data of the equipment.

[0028] Preferably, in step S4, Tensor1 = [T1, T2, .... Tn], Tensor2 = [T1', T2', .... Tn'], then the similarity θ calculation formula is:

[0029]

[0030] Compared with the existing technology, the advantages of the present invention are:

[0031] 1. The present invention can accurately judge the lubrication status of the equipment through lubrication monitoring model analysis and adjust the lubrication strategy in time, avoiding the shortcomings of fixed time intervals and empirical lubrication in traditional lubrication methods and improving the lubrication effect.

[0032] 2. The present invention accurately identifies the abnormal operating status of the equipment through the lubrication monitoring model and takes timely measures, effectively reducing equipment failures caused by lubrication problems, avoiding excessive wear and damage to parts, reducing equipment failure rates, and extending the service life of the equipment.

[0033] 3. The lubricator of the present invention can dynamically adjust the oil output according to actual needs, thereby avoiding over-lubrication, saving the use of lubricating oil, and reducing production costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic structural diagram of the lubrication system of the present invention.

[0035] Figure 2 It is a structural diagram of the lubrication monitoring model in the present invention.

[0036] Figure 3 Schematic diagram of the structure of the Dilated Re-param Block in the present invention. DETAILED DESCRIPTION

[0037] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0038] Reference Figure 1 As shown, an intelligent lubricator lubrication system includes:

[0039] Lubricator: used to receive commands issued by the system management platform and perform corresponding operations;

[0040] System management platform: communicates with the lubricator and the user respectively, is responsible for receiving the parameters set by the user, issuing commands to the lubricator, receiving the field information uploaded by the lubricator, and providing real-time feedback to the user;

[0041] User: Set parameters through the system management platform and receive real-time information feedback from the system management platform.

[0042] The commands issued by the system management platform include the lubrication cycle of the lubricator and the oil output of the oil cup. Users can set the lubrication cycle of the lubricator and the oil output of the oil cup. For example, the lubrication cycle is 1 month, and the oil output each time is about 4.17 ml. When the equipment installed with the lubricator stops running, the system management platform can detect that the equipment has stopped running through the equipment management system and issue instructions. The lubricator will stop automatic lubrication to avoid unnecessary lubrication operations.

[0043] An intelligent lubricator lubrication control method comprises the following steps:

[0044] S1. Collect real-time operating data of the lubricated equipment through sensors;

[0045] S2, building a lubrication monitoring model;

[0046] S3, using the historical operating data of the lubricated equipment as a training set to train the lubrication monitoring model;

[0047] S4. Input the collected real-time operation data into the trained lubrication monitoring model, output the feature vector Tensor1, input the normal operation data of the lubricated equipment into the trained lubrication monitoring model, output the feature vector Tensor2, and calculate the similarity θ between the two sets of vectors;

[0048] S5. Adjust the command to the lubricator according to the similarity θ.

[0049] In step S1 , the operating data includes the operating speed, the effective value of the operating vibration value, and the oil output of the lubricator oil cup.

[0050] In step S2, Figure 2 As shown, the lubrication monitoring model includes: an input layer for receiving input operation data;

[0051] At least one large block for processing input data;

[0052] A ReLU activation layer to introduce nonlinearity;

[0053] An SE-Large block for further processing of the data;

[0054] Output layer, used to output the processed feature vector;

[0055] Multiple large blocks and SE-Large blocks are arranged in a stacked manner, wherein the large block includes a dilated re-param block, at least one batch normalization layer (BN), a feedforward network (FNN) and a residual connection, and the SE-Large block includes a dilated re-param block, at least one batch normalization layer (BN), an SE module and a residual connection.

[0056] The dilated re-param block is composed of multiple parallel dilated convolutions combined into a larger dilated convolution. For example, 9x9 convolution Figure 3 As shown in

[15] , enhancing the ability of large kernels to capture sparse patterns (i.e., a pixel on the feature map may be more distantly related to its neighboring pixels) may produce higher quality features.

[0057] The feedforward network (FNN) consists of two fully connected layers with a GELU activation function in between, which is used to perform further nonlinear transformations on the input data. The commonly used ReLU activation function is replaced with the GELU activation function. Compared with the ReLU activation function, the GELU activation line number is smoother, which helps to improve the convergence speed and performance of the training process.

[0058] The SE module includes an average pooling layer (avg-pool), a RELU activation function, and a sigmoid activation function. The SE module focuses on the channel dimension of the feature map and assigns a weight to each channel, thereby strengthening important channels and suppressing irrelevant channels. It extracts the global information of each channel through avg-pool global average pooling and uses a scalar to describe the importance of each channel. By generating the weight of each channel through the activation function and two fully connected layers, the number of parameters and computations are relatively small, but the network's expressive power is significantly improved.

[0059] In step S3, the historical operating data of the lubricated equipment includes normal state data and abnormal state data of the equipment.

[0060] In step S4, Tensor1 = [T1, T2, .... Tn], Tensor2 = [T1', T2', .... Tn'], then the similarity θ calculation formula is:

[0061]

[0062] When using multiple sets of data for empirical value testing, when θ is within the normal range, it is judged that the bearing condition is within the normal range and the oil output of the lubricator's oil cup is normal. When the two sets of eigenvectors are very close, θ is outside the normal range and there is a problem with the operating data, reflecting that there is a problem with the oil output of the lubricator's oil cup and the oil output strategy needs to be changed.

[0063] Example

[0064] In this embodiment, the lubricated equipment is selected as the motor bearing. The system hardware includes a lubricator, a system management platform server, and a sensor (used to collect the operating speed of the motor bearing, the effective value of the operating vibration value, and the oil output of the lubricator oil cup). In terms of software, the system management platform runs on the server and uses a professional database management system to store equipment operation data.

[0065] The model is learned and processed for normal data, and through field data testing, it is found that when the similarity θ distance is between 0 and 5, it is within the normal range.

[0066] Real-time operating data of the motor bearings was collected for three periods of time, with each period lasting 5 minutes. A set of operating data was recorded every 30 seconds. The 10 sets of operating data for each period were input into the lubrication monitoring model to calculate the similarity θ. The average similarity θ for each period was then calculated. The results are shown in the following table.

[0067] Table 1: Operation data result judgment table

[0068] time distance Situation determination Lubrication real situation Paragraph 1 22.5 Abnormal operation Grease blockage Second paragraph 40.32 Abnormal operation Too little oil Paragraph 3 3.2 Normal operation Normal oil content

[0069] According to the experimental results in the table above, the average similarity distance in the first period was 22.5. The system judged it to be abnormal operation, but the actual condition was grease blockage. This shows that when the bearing is grease blocked, its operating data is quite different from the normal operating data. The model can keenly capture this difference and make an accurate judgment. The average similarity distance in the second period was 40.32. The system judged it to be abnormal operation, but the actual condition was too little grease. This shows that insufficient grease will significantly affect the operating characteristics of the bearing. The model can identify this abnormal state by analyzing the operating data. The average similarity distance in the third period was 3.2. The system judged it to be normal operation, but the actual condition was normal grease. This verifies that when the bearing is in good lubrication condition, its operating data is highly similar to the normal operating data.

[0070] It is understood from common technical knowledge that the present invention may be implemented by other embodiments that do not depart from its spirit or essential features. Therefore, the embodiments disclosed above are, in all respects, merely illustrative and not exclusive. All modifications within the scope of the present invention or equivalent to the scope of the present invention are intended to be encompassed by the present invention.

Claims

1. An intelligent lubricator lubrication system, characterized in that: include: Lubricator: used to receive commands issued by the system management platform and perform corresponding operations; System management platform: communicates with the lubricator and the user respectively, is responsible for receiving the parameters set by the user, issuing commands to the lubricator, receiving the field information uploaded by the lubricator, and providing real-time feedback to the user; User: Set parameters through the system management platform and receive real-time information feedback from the system management platform.

2. The intelligent lubricator lubrication system according to claim 1, characterized in that: The commands issued by the system management platform include the lubrication cycle of the lubricator and the oil output of the oil cup.

3. An intelligent lubricator lubrication control method, characterized in that: The following steps are involved: S1. Collect real-time operating data of the lubricated equipment through sensors; S2, building a lubrication monitoring model; S3, using the historical operating data of the lubricated equipment as a training set to train the lubrication monitoring model; S4. Input the collected real-time operation data into the trained lubrication monitoring model, output the feature vector Tensor1, input the normal operation data of the lubricated equipment into the trained lubrication monitoring model, output the feature vector Tensor2, and calculate the similarity θ between the two sets of vectors; S5. Adjust the command to the lubricator according to the similarity θ.

4. The intelligent lubricator lubrication control method according to claim 3, characterized in that: In step S1 , the operating data includes the operating speed, the effective value of the operating vibration value, and the oil output of the lubricator oil cup.

5. The intelligent lubricator lubrication control method according to claim 3, characterized in that: In step S2, the lubrication monitoring model includes: an input layer for receiving input operation data; At least one large block for processing input data; A ReLU activation layer to introduce nonlinearity; An SE-Large block for further processing of the data; Output layer, used to output the processed feature vector; Multiple large blocks and SE-Large blocks are arranged in a stacked manner, wherein the large block includes an expanded re-parameter block, at least one batch normalization layer, a feed-forward network and a residual connection, and the SE-Large block includes an expanded re-parameter block, at least one batch normalization layer and an SE module.

6. The intelligent lubricator lubrication control method according to claim 3, characterized in that: In step S3, the historical operating data of the lubricated equipment includes normal state data and abnormal state data of the equipment.

7. The intelligent lubricator lubrication control method according to claim 3, characterized in that: In step S4, Tensor1 = [T1, T2, .... Tn], Tensor2 = [T1', T2', .... Tn'], then the similarity θ calculation formula is:

Images