An intelligent cleaning evaluation method for belt conveyors
Through multiple image acquisition sensors and random forest models, the shape and type of residues on the belt conveyor are detected, and the appropriate cleaning mode is selected, which solves the problem of poor cleaning effect in the prior art and improves the cleaning accuracy of the belt conveyor.
Patent Information
- Application Number
- CN202411545423.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-01
- Publication Date
- 2025-01-03
- Estimated Expiration
- 2044-11-01
AI Technical Summary
The prior art cannot effectively detect the shape and type of residues on belt conveyors, resulting in poor cleaning effect and low cleaning accuracy.
Multiple image acquisition sensors (one located directly above and one located on the side) collect the shape and color information of the residue. Combined with the random forest model, the control center selects the appropriate cleaning mode based on the collected data for cleaning, and improves through continuous learning. Accuracy of cleaning mode selection.
Accurate detection of the shape and type of residues of belt conveyors is achieved, and the appropriate cleaning mode is selected, which improves the cleaning effect and cleaning accuracy.
Smart Images
Figure CN119038090B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of cleaning of belt conveyors, and in particular to an intelligent cleaning evaluation method for belt conveyors. Background Art
[0002] Existing cleaning systems for belt conveyors mostly rely on manual inspection or simple sensor feedback, which has the following disadvantages: 1. High dependence on manual inspection. Manual inspection is not only time-consuming, but may also lead to inconsistent cleaning results due to human factors. 2. Poor real-time performance. Traditional systems are usually unable to monitor cleaning results in real time and cannot detect problems in time. 3. Limited data processing capabilities. Existing technologies often only use simple threshold judgments and fail to fully utilize data for in-depth analysis.
[0003] In the patent publication number CN118183229A, an intelligent cleaning evaluation system for a belt conveyor is disclosed. This application performs real-time monitoring during daily cleaning, and when the monitoring result is unqualified, timely adjustment is made through the second cleaning module to perform a suppressive secondary operation. The suppressive cleaning reduces the interference caused by belt slack, and makes the belt as straight as possible before cleaning, thereby improving the cleaning efficiency. This application improves the cleaning effect and the evaluation efficiency and accuracy of the cleaning effect through the integrated structural design of daily cleaning (first cleaning module), real-time monitoring, and emergency cleaning (second cleaning module), and realizes real-time monitoring.
[0004] The inventors found that the prior art has the following problems:
[0005] In the prior art, it is impossible to detect the shape and type of the residue, resulting in poor cleaning effect and poor cleaning accuracy of the belt conveyor. Summary of the invention
[0006] In view of the above problems, the present application provides an intelligent cleaning evaluation method for a belt conveyor, which is used to solve the technical problem in the prior art that the shape and type of residues cannot be detected, resulting in poor cleaning effect and poor cleaning accuracy of the belt conveyor.
[0007] To achieve the above objectives, the present application provides an intelligent cleaning evaluation method for a belt conveyor, comprising the following steps:
[0008] S1. Build an intelligent cleaning evaluation system for belt conveyors
[0009] Construct a data acquisition module, a data transmission module, a control center and a belt conveyor cleaning module. The data acquisition module includes more than two image acquisition sensors, one of which is located directly above the belt conveyor and is used to collect the horizontal shape and color information of the residue on the belt conveyor. The other image acquisition sensor is located on the side of the belt conveyor and is used to collect the vertical shape and color information of the residue on the belt conveyor. The belt conveyor cleaning module includes a scraper cleaning mode, a rotating brush cleaning mode, a hydraulic flushing cleaning mode, a vibration cleaning mode and a negative pressure cleaning mode;
[0010] S2. Selection of cleaning mode
[0011] The data collected by the data acquisition module is sent to the control center through the data transmission module. The control center connects to and controls the belt conveyor cleaning module. The control center determines the amount and type of residues on the belt conveyor according to the data collected by the data acquisition module. The control center selects different cleaning modes of the belt conveyor cleaning module according to the amount and type of residues on the belt conveyor. One cleaning mode or a combination of multiple cleaning modes can be selected at the same time to clean the residues on the belt conveyor.
[0012] S3. Saving of cleaning results
[0013] After cleaning the residues on the belt conveyor, image information of the residues on the belt conveyor is collected again, and compared with the image information of the residues on the belt conveyor before cleaning, to obtain cleaning effect information, and save the cleaning effect information for future use;
[0014] S4. Build a random forest model
[0015] A random forest model is established, and the image information of the residues on the belt conveyor before cleaning, the image information of the residues on the belt conveyor after cleaning, the cleaning effect information, and the corresponding belt conveyor cleaning module information collected by the data acquisition module are input into the random forest model for training to obtain a trained random forest model;
[0016] S5. Store the random forest model in the control center
[0017] The trained random forest model is saved in the control center. Next time when the data acquisition module collects image information of residues on the belt conveyor, the control center selects the cleaning mode of the belt conveyor cleaning module through the trained random forest model.
[0018] As an embodiment of the present invention, a belt conveyor cleaning module is installed on one side of the belt conveyor. The belt conveyor cleaning module is set relative to the contact position between the belt and the support wheel. The belt conveyor cleaning module includes a scraper mechanism, a rotating brush head, a hydraulic flushing head, a vibration mechanism and a negative pressure mechanism.
[0019] As an embodiment of the present invention, the intelligent cleaning evaluation system also includes a cleaning frequency monitor, a belt speed sensor and an inclination sensor. The cleaning frequency monitor is used to record the cleaning frequency of the belt conveyor cleaning module, the belt speed sensor is used to detect the running speed of the belt conveyor, and the inclination sensor is used to detect the inclination angle of the belt.
[0020] As an embodiment of the present invention, in the step of selecting a cleaning mode, image information of the residues and type information of the residues are pre-stored in a database, and based on the image information of the residues and the type information of the residues, a corresponding cleaning mode is stored, and the corresponding relationship between the residues and the cleaning mode is stored in a control center.
[0021] As an embodiment of the present invention, in the step of selecting the cleaning mode, the control center processes the image information transmitted by the data acquisition module, extracts the edges of the residues, filters out the interference information according to the size, shape, and color characteristics, retains the image information of the residues, and compares the detected image information of the residues with the image information of the residues in the database, thereby selecting the cleaning mode through the control center.
[0022] As an implementation mode of the present invention, in the step of constructing a random forest model:
[0023] The sample is established, the intelligent cleaning evaluation system of the belt conveyor is started, and the cleaning task of the first predetermined number of times is completed. The data acquisition module records the mechanical and environmental parameters of the intelligent cleaning evaluation system during cleaning and stores them in the control center as a cleaning sample set;
[0024] Feature selection when generating a single decision tree: the data collected by the data acquisition module is used as the feature set of the decision tree to set the tree depth value of the single decision tree.
[0025] As an embodiment of the present invention, in the step of constructing a random forest model,
[0026] Randomly selecting a second preset number of samples from the cleaning sample set with replacement as a cleaning subset, and repeatedly constructing a decision tree with differences;
[0027] When each node of the decision tree needs to be split, three feature attributes are randomly selected from the eight feature attributes of the sweeping subset, and one attribute is selected from these three attributes using the information gain strategy as the splitting attribute of the node;
[0028] Each decision tree produces an output, which is determined by parliamentary voting.
[0029] As an implementation mode of the present invention, according to the cleaning classification result, whether it is qualified or unqualified, it is determined whether the data in the cleaning sample set is a qualified sample or an unqualified sample, and the unqualified samples are deleted.
[0030] As an implementation mode of the present invention, a human-computer interaction system is established, and the human-computer interaction system includes a display module, an alarm module and a database, and the display module is used to display the real-time data collected by the data collection module and the cleaning effect data;
[0031] The alarm module is used in the intelligent cleaning evaluation system. When the cleaning result is evaluated as unqualified, an alarm is issued and stored in the control center;
[0032] The database is used to save all the cleaning data, and the random forest model is updated through the database.
[0033] Different from the prior art, the above technical solution uses more than two image acquisition sensors, one of which is located directly above the belt conveyor, for collecting the horizontal shape and color information of the residue on the belt conveyor, and the other image acquisition sensor is located on the side of the belt conveyor, for collecting the vertical shape and color information of the residue on the belt conveyor. The belt conveyor cleaning module includes a scraper cleaning mode, a rotating brush cleaning mode, a hydraulic flushing cleaning mode, a vibration cleaning mode and a negative pressure cleaning mode; the control center selects different cleaning modes of the belt conveyor cleaning module according to the amount and type of residue on the belt conveyor, and can select one cleaning mode or a combination of multiple cleaning modes at the same time, save the cleaning effect information, train the cleaning effect through a random forest model, and improve the accuracy of the control center's selection of the cleaning mode through continuous learning to achieve intelligent cleaning; in this way, different belt conveyor cleaning modes can be selected by detecting the shape and type of residues, thereby improving the cleaning effect and cleaning accuracy of the belt conveyor.
[0034] The above-mentioned records related to the invention content are only an overview of the technical solution of the present application. In order to enable ordinary technicians in the field to more clearly understand the technical solution of the present application, and then implement it according to the text of the specification and the contents recorded in the drawings, and to make the above-mentioned purpose and other purposes, features and advantages of the present application easier to understand, the following is an explanation in combination with the specific implementation mode and drawings of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] The drawings are only used to illustrate the principles, implementation methods, applications, characteristics and effects of the specific embodiments of the present application and other related contents, and shall not be considered as limitations of the present application.
[0036] In the drawings of the specification:
[0037] Figure 1 A top view of a data acquisition module in a belt conveyor according to an embodiment of the present application;
[0038] Figure 2 This is a front view of a data acquisition module in a belt conveyor according to an embodiment of the present application;
[0039] Figure 3 A schematic diagram of a single decision tree according to an embodiment of the present application;
[0040] Figure 4 A schematic diagram of the random forest modeling process of one embodiment of the present application.
[0041] The reference numerals in the above drawings are described as follows:
[0042] 10. Belt,
[0043] 20. Image acquisition sensor,
[0044] 30. Residues,
[0045] 40. Belt conveyor cleaning module. DETAILED DESCRIPTION
[0046] In order to explain in detail the possible application scenarios, technical principles, specific schemes that can be implemented, and the purposes and effects that can be achieved, the following is a detailed description of the specific embodiments listed in conjunction with the accompanying drawings. The embodiments described herein are only used to more clearly illustrate the technical solutions of the present application, and are therefore only used as examples, and cannot be used to limit the scope of protection of the present application.
[0047] Reference to "embodiment" herein means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present application. The term "embodiment" appearing in various places in the specification does not necessarily refer to the same embodiment, nor does it particularly limit its independence or association with other embodiments. In principle, in the present application, as long as there is no technical contradiction or conflict, the various technical features mentioned in the embodiments can be combined in any way to form a corresponding implementable technical solution.
[0048] Unless otherwise defined, the technical terms used in this document have the same meanings as those generally understood by those skilled in the art to which this application belongs; the use of relevant terms in this document is only for describing specific embodiments and is not intended to limit this application.
[0049] In the description of this application, the term "and / or" is an expression used to describe the logical relationship between objects, indicating that three relationships may exist, for example, A and / or B, which means: A exists, B exists, and A and B exist at the same time. In addition, the character " / " in this article generally indicates that the objects before and after are in an "or" logical relationship.
[0050] In the present application, terms such as “first” and “second” are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship of quantity, priority or sequence between these entities or operations.
[0051] Without further limitations, in this application, the words "include", "comprises", "has" or other similar expressions used in the sentences are intended to cover non-exclusive inclusion. These expressions do not exclude the presence of additional elements in the process, method or product including the elements, so that the process, method or product including a series of elements may include not only those limited elements, but also other elements not explicitly listed, or also include elements inherent to such process, method or product.
[0052] Similar to the understanding in the Examination Guidelines, in this application, expressions such as "greater than", "less than", and "exceed" are understood to exclude the number itself; expressions such as "above", "below", and "within" are understood to include the number itself.
[0053] In addition, in the description of the embodiments of the present application, the meaning of "multiple" is more than two (including two), and similar expressions related to "multiple" should also be understood in this way, such as "multiple groups", "multiple times", etc., unless otherwise clearly and specifically defined.
[0054] According to some embodiments of the present application, please refer to Figures 1 to 4 , this embodiment relates to an intelligent cleaning evaluation method for a belt conveyor, comprising the following steps:
[0055] S1. Build an intelligent cleaning evaluation system for belt conveyors
[0056] Construct a data acquisition module, a data transmission module, a control center and a belt conveyor cleaning module. The data acquisition module includes more than two image acquisition sensors, one of which is located directly above the belt conveyor and is used to collect the horizontal shape and color information of the residue on the belt conveyor. The other image acquisition sensor is located on the side of the belt conveyor and is used to collect the vertical shape and color information of the residue on the belt conveyor. The belt conveyor cleaning module includes a scraper cleaning mode, a rotating brush cleaning mode, a hydraulic flushing cleaning mode, a vibration cleaning mode and a negative pressure cleaning mode;
[0057] S2. Selection of cleaning mode
[0058] The data collected by the data acquisition module is sent to the control center through the data transmission module. The control center connects to and controls the belt conveyor cleaning module. The control center determines the amount and type of residues on the belt conveyor according to the data collected by the data acquisition module. The control center selects different cleaning modes of the belt conveyor cleaning module according to the amount and type of residues on the belt conveyor. One cleaning mode or a combination of multiple cleaning modes can be selected at the same time to clean the residues on the belt conveyor.
[0059] S3. Saving of cleaning results
[0060] After cleaning the residues on the belt conveyor, image information of the residues on the belt conveyor is collected again, and compared with the image information of the residues on the belt conveyor before cleaning, to obtain cleaning effect information, and save the cleaning effect information for future use;
[0061] S4. Build a random forest model
[0062] A random forest model is established, and the image information of the residues on the belt conveyor before cleaning, the image information of the residues on the belt conveyor after cleaning, the cleaning effect information, and the corresponding belt conveyor cleaning module information collected by the data acquisition module are input into the random forest model for training to obtain a trained random forest model;
[0063] S5. Store the random forest model in the control center
[0064] The trained random forest model is saved in the control center. Next time when the data acquisition module collects image information of residues on the belt conveyor, the control center selects the cleaning mode of the belt conveyor cleaning module through the trained random forest model.
[0065] In this embodiment, Figure 1 and Figure 2As shown, an image acquisition sensor 20 is provided on each of the left and right sides of the belt 10, facing the left and right sides of the belt 10 respectively, and can collect the height and length of the residue 30 on the belt 10. An image acquisition sensor 20 is provided on each of the left and right sides of the belt 10, and can collect from the left and right sides of the residue 30 respectively, thereby improving the accuracy of the image acquisition sensor 20;
[0066] Directly above the belt 10, there are more than two image acquisition sensors 20, which face the belt 10 from top to bottom, and can collect the length and width of the residue 30 on the belt 10. Through at least three image acquisition sensors 20, two image acquisition sensors 20 on the left and right, plus an image acquisition sensor 20 directly above, the number, shape, and color of the residue 30 on the belt 10 can be accurately detected, and the type of residue can be analyzed, providing strong support for the subsequent selection of the belt conveyor cleaning module 40.
[0067] In this embodiment, the positions of the plurality of belt conveyor cleaning modules 40 may be arranged in a horizontal direction (specifically, in a width direction of the residue 30 ) or in a vertical direction.
[0068] In this embodiment, the types of materials transported (ore, coal, sand, gravel and other residues or dust) and the characteristics of different materials affect the adhesion of the residues. Different cleaning modes are used for different residues.
[0069] The image acquisition sensor uses a high-resolution industrial camera with a focal length of 20mm. USB3.0 is used to connect the PLC and the camera, and code is written to control the camera for image acquisition.
[0070] According to some embodiments of the present application, optionally, a belt conveyor cleaning module is installed on one side of the belt conveyor, and the belt conveyor cleaning module is arranged relative to the contact position between the belt and the support wheel. The belt conveyor cleaning module includes a scraper mechanism, a rotating brush head, a hydraulic flushing head, a vibration mechanism and a negative pressure mechanism.
[0071] In this embodiment, considering that the belt is tightened on both sides and loose in the middle after being used for a long time, forming a concave shape, the belt conveyor cleaning module is installed at the contact position between the belt and the support wheel. The bottom of the belt is supported by the support wheel at this position, so that the originally loose middle part of the belt is propped up, which can avoid the problem of poor cleaning of the loose middle part of the belt;
[0072] In this way, the belt conveyor cleaning module is installed on one side of the belt conveyor, the belt conveyor cleaning module is set relative to the contact position between the belt and the supporting wheel, and the belt conveyor cleaning module has different cleaning mechanisms to improve the cleaning effect of the belt conveyor cleaning module.
[0073] According to some embodiments of the present application, optionally, the intelligent cleaning evaluation system also includes a cleaning frequency monitor, a belt speed sensor and an inclination sensor. The cleaning frequency monitor is used to record the cleaning frequency of the belt conveyor cleaning module, the belt speed sensor is used to detect the running speed of the belt conveyor, and the inclination sensor is used to detect the inclination angle of the belt.
[0074] In this embodiment, an environmental sensor is also included, which monitors factors affecting the cleaning effect such as humidity and temperature of the surrounding environment. According to the conditions detected by the cleaning frequency monitor, the belt speed sensor, the inclination sensor, and the environmental sensor, the cleaning mode and the cleaning time are selected. The cleaning can be carried out by means of a scraper, a water flushing, a vibration, or an adsorption method, which do not interfere with each other and can be carried out separately or in combination.
[0075] According to some embodiments of the present application, optionally, in the step of selecting a cleaning mode, image information of the residues and type information of the residues are pre-stored in a database, and based on the image information of the residues and the type information of the residues, a corresponding cleaning mode is stored, and the corresponding relationship between the residues and the cleaning mode is stored in a control center.
[0076] In this way, the fixed types of transported materials (ore, coal, sand and gravel, residues or dust), the corresponding image information, and material types can be stored in the database first, and the cleaning methods for different material types can be stored accordingly. For example, large ores are cleaned by vibration, coal is cleaned by negative pressure adsorption, and small sand and gravel are cleaned by high-pressure water guns.
[0077] According to some embodiments of the present application, optionally, in the step of selecting the cleaning mode, the control center processes the image information transmitted by the data acquisition module, extracts the edges of the residues, filters out interference information based on size, shape, and color features, retains the image information of the residues, and compares the detected image information of the residues with the image information of the residues in the database, thereby selecting the cleaning mode through the control center.
[0078] The control center receives and stores data from sensors in a unified format, that is, text data and logical data are converted into numerical data. For example, the adhesion generated by different types of conveyed materials is normalized and divided into 1 to 10. The installation angle of the belt conveyor in the inclination sensor is set to 0 to 90°. Then, outliers and missing values are explored and processed, and data that is not within the normal range is deleted. The missing values are predicted using the spline fitting method.
[0079] In this way, the accuracy of data processing can be improved, and the images during the shooting process can be stored in the random forest model for training.
[0080] According to some embodiments of the present application, optionally, in the step of building a random forest model:
[0081] The sample is established, the intelligent cleaning evaluation system of the belt conveyor is started, and the cleaning task of the first predetermined number of times is completed. The data acquisition module records the mechanical and environmental parameters of the intelligent cleaning evaluation system during cleaning and stores them in the control center as a cleaning sample set;
[0082] Feature selection when generating a single decision tree: the data collected by the data acquisition module is used as the feature set of the decision tree to set the tree depth value of the single decision tree.
[0083] In this embodiment, the data collected by the sensor is used as the feature set of the decision tree. It is necessary to consider the performance of a single decision tree and the differences between trees. Therefore, the feature set for generating a single decision tree is set to 5. Tree depth: If the tree depth is too large, each subtree will over-learn, which may lead to overfitting. Therefore, the tree depth is set to 3. According to the parameter setting, a single decision tree Figure 3 shown.
[0084] According to some embodiments of the present application, optionally, Figure 4 As shown, in the step of building a random forest model,
[0085] Randomly selecting a second preset number of samples from the cleaning sample set with replacement as a cleaning subset, and repeatedly constructing a decision tree with differences;
[0086] Specifically, 80 samples were randomly selected from the sample set as a subset using bagging, and the process was repeated 20 times to construct 20 different decision trees.
[0087] When each node of the decision tree needs to be split, three feature attributes are randomly selected from the eight feature attributes of the sweeping subset, and one attribute is selected from these three attributes using the information gain strategy as the splitting attribute of the node;
[0088] Each decision tree produces an output, which is determined by parliamentary voting.
[0089] "Parliamentary system" voting method: Referring to the "parliamentary system", the generated decision tree is divided into Speaker A and Member B according to its false negative rate (the proportion of samples that fail to correctly identify as actually qualified) and accuracy (if the classification accuracy of a decision tree is higher than the set threshold and the false negative rate is lower than the threshold, the tree is divided into A; if the classification accuracy of a decision tree is higher than the set threshold, but the false negative rate is not lower than the threshold, the tree is divided into B), and different weight coefficients are assigned to each member: the weight coefficient of member A is the false negative rate of each member Ai, and the weight coefficient of member B is the accuracy rate of each member Bj. The final judgment result of a sample is as follows:
[0090] 1) If the members of B consider this sample to be unqualified after weight addition, this result shall be regarded as the judgment result of this sample;
[0091] 2) If the members of B consider this sample to be a qualified sample after weighted addition, the weighted addition judgment result of the members of A on this sample will be used as the judgment result of this sample.
[0092] According to some embodiments of the present application, optionally, based on the cleaning classification result, whether it is qualified or unqualified, it is determined whether the data in the cleaning sample set is a qualified sample or an unqualified sample, and the unqualified samples are deleted.
[0093] The random forest model in this embodiment is a continuous optimization process. Samples with qualified cleaning effects are retained, and samples with unqualified cleaning effects are deleted, so as to achieve the purpose of controlling only intelligent cleaning.
[0094] In this way, the accuracy of the random forest model is improved.
[0095] According to some embodiments of the present application, optionally, a human-computer interaction system is established, the human-computer interaction system includes a display module, an alarm module and a database, the display module is used to display the real-time data collected by the data collection module and the cleaning effect data;
[0096] The alarm module is used in the intelligent cleaning evaluation system. When the cleaning result is evaluated as unqualified, an alarm is issued and stored in the control center;
[0097] The database is used to save all the cleaning data, and the random forest model is updated through the database.
[0098] In this embodiment, the display module has a real-time data display function: the control center sends a call instruction to the database to view the real-time monitoring data of various sensors; at the same time, when too much residue is detected (the number of residue images identified on the same belt is greater than 10), the operator is reminded to start the cleaning operation immediately and set the parameters of this cleaning task as high as possible to improve the cleaning accuracy.
[0099] Cleaning task parameters: including setting cleaning time, cleaning frequency, suction strength of cleaning equipment, belt running speed, and belt conveyor inclination.
[0100] Cleaning effect evaluation: The random forest model stores the classified cleaning results in a table in the database, and the control center calls this data to display it to the operator.
[0101] Adjustment of cleaning strategy: For "unqualified" cleaning results, the control center analyzes historical data and adjusts the cleaning strategy based on the false negative rate and accuracy rate. For example, a high false negative rate means that the model fails to identify certain areas that need to be cleaned. The belt speed can be reduced, and the room temperature can be appropriately lowered (to reduce adsorption force); a low accuracy rate means that the cleaning force is not strong enough, and the suction strength and cleaning time of the cleaning equipment can be appropriately increased.
[0102] For cleaning tasks with an evaluation result of "unqualified" by the alarm module, the control center will mark the task and issue an alarm. After receiving the alarm, the operator will review the cleaning task. The operator can manually modify the cleaning strategy or adjust the cleaning strategy according to the recommended cleaning suggestions.
[0103] After each cleaning task is completed, the database creates a table for the saved cleaning data, which contains all the characteristics of that cleaning. This can be used to expand the sample set of the random forest model to make the model more accurate, and descriptive analysis methods and visualization tools can be used to help operators quickly understand the implementation effect of the cleaning task.
[0104] Different from the prior art, the above technical solution uses more than two image acquisition sensors, one of which is located directly above the belt conveyor, for collecting the horizontal shape and color information of the residue on the belt conveyor, and the other image acquisition sensor is located on the side of the belt conveyor, for collecting the vertical shape and color information of the residue on the belt conveyor. The belt conveyor cleaning module includes a scraper cleaning mode, a rotating brush cleaning mode, a hydraulic flushing cleaning mode, a vibration cleaning mode and a negative pressure cleaning mode; the control center selects different cleaning modes of the belt conveyor cleaning module according to the amount and type of residue on the belt conveyor, and can select one cleaning mode or a combination of multiple cleaning modes at the same time, save the cleaning effect information, train the cleaning effect through a random forest model, and improve the accuracy of the control center's selection of the cleaning mode through continuous learning to achieve intelligent cleaning; in this way, different belt conveyor cleaning modes can be selected by detecting the shape and type of residues, thereby improving the cleaning effect and cleaning accuracy of the belt conveyor.
[0105] This embodiment also relates to an intelligent cleaning evaluation method for a belt conveyor;
[0106] Different from the prior art, the intelligent cleaning evaluation system of the technical solution of the present application can select different belt conveyor cleaning modes by detecting the shape and type of residues, establish a random forest model by collecting the cleaning effects, continuously adjust the weight distribution, improve the accuracy, and ultimately improve the accuracy of the control center's selection of cleaning modes to achieve intelligent cleaning evaluation.
[0107] Those skilled in the art should understand that the above embodiments can be provided as methods, devices, or computer program products. These embodiments can be in the form of complete hardware embodiments, complete software embodiments, or embodiments combining software and hardware. All or part of the steps in the methods involved in the above embodiments can be completed by instructing the relevant hardware through a program, and the program can be stored in a storage medium readable by a computer device for executing all or part of the steps described in the above embodiments. The computer device includes, but is not limited to: personal computers, servers, general-purpose computers, special-purpose computers, network devices, embedded devices, programmable devices, smart mobile terminals, smart home devices, wearable smart devices, in-vehicle smart devices, etc.; the storage medium includes, but is not limited to: RAM, ROM, disk, tape, CD, flash memory, U disk, mobile hard disk, memory card, memory stick, network server storage, network cloud storage, etc.
[0108] The above embodiments are described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a computer device to generate a machine, so that the instructions executed by the processor of the computer device generate instructions for implementing the process in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0109] These computer program instructions may also be stored in a computer device readable memory capable of directing a computer device to operate in a specific manner, so that the instructions stored in the computer device readable memory produce a manufactured product including an instruction device, the instruction device implementing the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0110] These computer program instructions can also be loaded onto a computer device so that a series of operating steps are executed on the computer device to produce a computer-implemented process, whereby the instructions executed on the computer device provide for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0111] It should be noted that, although the above embodiments have been described in this article, the patent protection scope of the present invention is not limited thereby. Therefore, based on the innovative concept of the present invention, changes and modifications made to the embodiments described herein, or equivalent structures or equivalent process changes made using the contents of the present invention specification and drawings, directly or indirectly applying the above technical solutions to other related technical fields, are all included in the patent protection scope of the present invention.
Claims
1. An intelligent cleaning evaluation method for a belt conveyor, characterized in that: The following steps are involved: S1. Build an intelligent cleaning evaluation system for belt conveyors Construct a data acquisition module, a data transmission module, a control center and a belt conveyor cleaning module. The data acquisition module includes more than two image acquisition sensors, one of which is located directly above the belt conveyor and is used to collect the horizontal shape and color information of the residue on the belt conveyor. The other image acquisition sensor is located on the side of the belt conveyor and is used to collect the vertical shape and color information of the residue on the belt conveyor. The belt conveyor cleaning module includes a scraper cleaning mode, a rotating brush cleaning mode, a hydraulic flushing cleaning mode, a vibration cleaning mode and a negative pressure cleaning mode; S2. Selection of cleaning mode The data collected by the data acquisition module is sent to the control center through the data transmission module. The control center connects to and controls the belt conveyor cleaning module. The control center determines the amount and type of residues on the belt conveyor according to the data collected by the data acquisition module. The control center selects different cleaning modes of the belt conveyor cleaning module according to the amount and type of residues on the belt conveyor. One cleaning mode or a combination of multiple cleaning modes can be selected at the same time to clean the residues on the belt conveyor. S3. Preservation of cleaning results After cleaning the residues on the belt conveyor, image information of the residues on the belt conveyor is collected again, and compared with the image information of the residues on the belt conveyor before cleaning, to obtain cleaning effect information, and save the cleaning effect information for future use; S4. Build a random forest model A random forest model is established, and the image information of the residues on the belt conveyor before cleaning, the image information of the residues on the belt conveyor after cleaning, the cleaning effect information, and the corresponding belt conveyor cleaning module information collected by the data acquisition module are input into the random forest model for training to obtain a trained random forest model; S5. Store the random forest model in the control center The trained random forest model is saved in the control center. Next time when the data acquisition module collects image information of residues on the belt conveyor, the control center selects the cleaning mode of the belt conveyor cleaning module according to the trained random forest model. In the step of selecting the cleaning mode, image information of the residue and type information of the residue are stored in the database in advance, and the corresponding cleaning mode is stored according to the image information of the residue and the type information of the residue, and the corresponding relationship between the residue and the cleaning mode is stored in the control center; In the step of selecting the cleaning mode, the control center processes the image information transmitted by the data acquisition module, extracts the edge of the residue, filters out the interference information according to the size, shape and color characteristics, retains the image information of the residue, and compares the image information of the detected residue with the image information of the residue in the database, thereby selecting the cleaning mode through the control center; In the step of building a random forest model: The sample is established, the intelligent cleaning evaluation system of the belt conveyor is started, and the cleaning task of the first predetermined number of times is completed. The data acquisition module records the mechanical and environmental parameters of the intelligent cleaning evaluation system during cleaning and stores them in the control center as a cleaning sample set; Feature selection when generating a single decision tree: the data collected by the data collection module is used as the feature set of the decision tree to set the tree depth value of the single decision tree; In the step of building a random forest model, Randomly selecting a second preset number of samples from the cleaning sample set with replacement as a cleaning subset, and repeatedly constructing a decision tree with differences; When each node of the decision tree needs to be split, three feature attributes are randomly selected from the eight feature attributes of the sweeping subset, and one attribute is selected from these three attributes using the information gain strategy as the splitting attribute of the node; Each decision tree produces an output, which is determined by parliamentary voting.
2. The intelligent cleaning evaluation method for a belt conveyor according to claim 1, characterized in that: The belt conveyor cleaning module is installed on one side of the belt conveyor. The belt conveyor cleaning module is set relative to the contact position between the belt and the supporting wheel. The belt conveyor cleaning module includes a scraper mechanism, a rotating brush head, a hydraulic flushing head, a vibration mechanism and a negative pressure mechanism.
3. The intelligent cleaning evaluation method for a belt conveyor according to claim 1, characterized in that: The intelligent cleaning evaluation system also includes a cleaning frequency monitor, a belt speed sensor and an inclination sensor. The cleaning frequency monitor is used to record the cleaning frequency of the belt conveyor cleaning module, the belt speed sensor is used to detect the running speed of the belt conveyor, and the inclination sensor is used to detect the inclination angle of the belt.
4. The intelligent cleaning evaluation method for a belt conveyor according to claim 1, characterized in that: According to the cleaning classification results, whether it is qualified or unqualified, it is determined whether the data in the cleaning sample set is a qualified sample or an unqualified sample, and the unqualified samples are deleted.
5. The intelligent cleaning evaluation method for a belt conveyor according to claim 1, characterized in that: Establish a human-computer interaction system, which includes a display module, an alarm module and a database. The display module is used to display the real-time data collected by the data acquisition module and the cleaning effect data; The alarm module is used in the intelligent cleaning evaluation system. When the cleaning result is evaluated as unqualified, an alarm is issued and stored in the control center; The database is used to save all the cleaning data, and the random forest model is updated through the database.