Valve remote monitoring analysis system and method based on cloud platform

Abstract

The invention relates to the technical field of data processing, and discloses a valve remote monitoring analysis system and method based on a cloud platform, and the method comprises the steps: collecting the corrosion degree feature data of each valve; calling a first deep neural network model and a second deep neural network model; inputting the corrosion degree characteristic data into a first deep neural network model to obtain a corrosion degree coefficient of each valve; obtaining failure reflection feature data of each valve, and inputting the failure reflection feature data into the second deep neural network model to obtain a failure probability of each valve at a future moment; all target valves and the failure time of each target valve are determined according to the size of the failure probability, the remaining service life of all the target valves is calculated according to the failure time, and operation state grading is conducted according to the remaining service life; the corrosion degree, the failure probability and the residual life of the valve can be dynamically evaluated, and then hierarchical management and differentiated maintenance are achieved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more specifically, to a valve remote monitoring and analysis system and method based on a cloud platform. Background Art

[0002] As the core equipment in the industrial pipeline system, the operating status of the valve directly affects the safety and stability of the system. Under high pressure, high temperature or corrosive environment, the valve is subjected to complex chemical and mechanical effects for a long time, and is prone to problems such as corrosion and fatigue. This cumulative damage will cause the valve performance to gradually decline or even fail. Valve failure may cause leakage, interruption or even safety accidents, bringing huge risks and challenges to industrial production and equipment management.

[0003] Traditional valve monitoring and maintenance methods mainly rely on manual inspections or fixed-cycle maintenance. This method not only has the problem of insufficient real-time performance, but is also prone to excessive maintenance or maintenance delays due to inaccurate status assessment, increasing operating costs and reducing maintenance efficiency. At the same time, the existing technology fails to fully utilize the complex correlation between multi-dimensional feature data in the prediction of the remaining life and failure risk of the valve, and thus cannot accurately predict the degree of corrosion of the valve. Furthermore, it is no longer possible to determine the failure time, remaining service life and operating status level of the valve based on the degree of corrosion. As a result, it is difficult to provide reliable technical support for the efficient operation and maintenance and intelligent upgrading of industrial pipeline systems.

[0004] Therefore, how to use advanced cloud computing, big data analysis and artificial intelligence technologies to achieve remote real-time monitoring of valve operating status and accurate life prediction has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a valve remote monitoring and analysis system and method based on a cloud platform.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A valve remote monitoring and analysis method based on a cloud platform, the method comprising:

[0008] Collect the corrosion characteristic data of each valve in the pipeline and upload it to the cloud platform;

[0009] Calling a first deep neural network model pre-configured in the cloud platform for predicting the corrosion degree coefficient of the valve, and a second deep neural network model pre-configured for predicting the failure probability of the valve;

[0010] Inputting the corrosion degree characteristic data into the first deep neural network model to obtain the corrosion degree coefficient of each valve;

[0011] Obtain failure reflection characteristic data of each valve in the range from the past time TN to the current time T, and input the failure reflection characteristic data into the second deep neural network model to obtain the failure probability of each valve at the future time R, wherein the failure reflection characteristic data includes material characteristic data, fatigue degree characteristic data and corrosion degree coefficient at each time, and T, N and R are integers greater than zero;

[0012] The failure time of all target valves and each target valve is determined according to the failure probability, the remaining service life of all target valves is calculated according to the failure time, and the operating status is graded according to the remaining service life to obtain the operating status classification results of all target valves.

[0013] Furthermore, the corrosion degree characteristic data includes but is not limited to signal propagation time, signal natural frequency change, signal amplitude change and damping ratio.

[0014] Furthermore, the training method of the first deep neural network model is as follows:

[0015] Acquire historical corrosion degree training data, and divide the historical corrosion degree training data into a corrosion degree training set and a corrosion degree test set; the historical corrosion degree training data includes corrosion degree characteristic data and its corresponding corrosion degree coefficient;

[0016] Constructing a first deep neural network, taking the corrosion degree feature data in the corrosion degree training set as the input of the first deep neural network, taking the corrosion degree coefficient in the corrosion degree training set as the output of the first deep neural network, and taking minimizing the sum of the first prediction accuracies of all training samples as the training goal, training the first deep neural network until the sum of the first prediction accuracies reaches convergence, and stopping the training to obtain an initial first deep neural network;

[0017] The calculation formula of the sum of the first prediction accuracy is as follows:

[0018] ;

[0019] Where: represents the sum of the first prediction accuracy, represents the actual corrosion degree coefficient of the i-th group of training samples, represents the predicted corrosion degree coefficient of the i-th group of training samples, is the number of training samples in the corrosion degree training set;

[0020] The initial first deep neural network is verified using the corrosion degree test set. When the initial first deep neural network is less than or equal to the first prediction accuracy, it is used as the first deep neural network model for predicting the corrosion degree coefficient of the valve.

[0021] Furthermore, the method for obtaining the corrosion degree coefficient in the historical corrosion degree training data is as follows:

[0022] Obtain internal damage images of valves;

[0023] Grayscale the internal damage image to obtain a grayscale image;

[0024] Acquire a damage type reference image of the valve, and use the grayscale image as the target image, and use the damage type reference image as the matching image, wherein the damage types include cracks, holes, and bubbles;

[0025] Set the step size to 1, and perform cross-correlation calculation between the matching image and the target image through a sliding window method to obtain the similarity of each overlapping part;

[0026] The overlapping parts with similarity greater than a preset similarity threshold are regarded as damage areas, the number of damage areas of each damage type is counted, and the pixel area of ​​the damage area of ​​each damage type is obtained;

[0027] The total pixel area of ​​the target image is obtained, and the corrosion degree coefficient of the valve is calculated according to the number of damaged areas, the pixel area of ​​the damaged areas and the total pixel area;

[0028] The specific calculation formula of the corrosion degree coefficient is as follows:

[0029] ;

[0030] Where: Represents the corrosion severity coefficient, represents the pixel area of ​​the damaged area of ​​the vth damage type, represents the number of damaged areas of the vth damage type, represents the weight factor of the vth damage type, Indicates the total number of damage types, Represents the total pixel area of ​​the target image.

[0031] Furthermore, the training method of the second deep neural network model is as follows:

[0032] Acquire historical failure probability training data, and divide the historical failure probability training data into a failure probability training set and a failure probability test set; the historical failure probability training data includes failure reflection characteristic data and its corresponding failure probability;

[0033] The failure reflection characteristic data include material characteristic data, fatigue characteristic data and corrosion degree coefficient at each moment; the material characteristic data include material density, material hardness, material strength and material toughness; the fatigue characteristic data include the number of opening and closing times, operating time and average stress value;

[0034] Constructing a second deep neural network, taking the failure reflection feature data in the failure probability training set as the input of the second deep neural network, taking the failure probability in the failure probability training set as the output of the second deep neural network, and taking minimizing the sum of the second prediction accuracies of all training samples as the training goal, training the second deep neural network until the sum of the second prediction accuracies reaches convergence, and stopping the training to obtain an initial second deep neural network;

[0035] The calculation formula of the sum of the second prediction accuracy is as follows:

[0036] ;

[0037] Where: represents the sum of the second prediction accuracy, represents the actual failure probability of the jth group of training samples, represents the predicted failure probability of the jth group of training samples, is the number of samples in the corrosion degree training set;

[0038] The initial second deep neural network is verified using a failure probability test set, and when the initial second deep neural network is less than or equal to the second prediction accuracy, it is used as the second deep neural network model for predicting the failure probability of the valve.

[0039] Further, determining the failure time of each valve according to the failure probability includes:

[0040] Extract the failure probability of all valves at the future time R;

[0041] The failure probability of each valve is compared with the preset failure probability threshold. If the failure probability is greater than or equal to the preset failure probability threshold, the corresponding valve is marked as the target valve, and the future time R is used as the failure time of the target valve; if the failure probability is less than the preset failure probability threshold, the corresponding valve is marked as a non-target valve;

[0042] By counting all valves whose failure probability is greater than or equal to the failure probability threshold and recording the future time R, the failure time of all target valves and each target valve is obtained.

[0043] Furthermore, the remaining service life of all target valves is calculated based on the failure time, including:

[0044] Extract the failure time of each target valve;

[0045] Calculate the difference between the failure time and the current time T, and take the difference between the failure time and the current time T as the remaining service life;

[0046] The difference between the failure time of each target valve and the current time T is counted to obtain the remaining service life of all target valves.

[0047] Furthermore, the operation status classification according to the remaining service life includes:

[0048] Set remaining service life thresholds Td1 and Td2, where Td1>Td2>0;

[0049] comparing the remaining useful life of each target valve with a preset remaining useful life threshold;

[0050] If the remaining service life is greater than or equal to the remaining service life threshold value Td1, it is determined that the corresponding target valve belongs to the first-level operating state level, and the first-level operating state level is output as the operating state classification result;

[0051] If the remaining service life is less than the remaining service life threshold value Td1 and greater than the remaining service life threshold value Td2, it is determined that the corresponding target valve belongs to the second-level operating state level, and the second-level operating state level is output as the operating state classification result;

[0052] If the remaining service life is less than or equal to the remaining service life threshold value Td2, it is determined that the corresponding target valve belongs to the third-level operating state level, and the third-level operating state level is output as the operating state classification result.

[0053] A valve remote monitoring and analysis system based on a cloud platform, which is implemented based on any one of the above-mentioned valve remote monitoring and analysis methods based on a cloud platform, and the system comprises:

[0054] The data acquisition module is used to collect the corrosion characteristic data of each valve in the pipeline and upload it to the cloud platform;

[0055] A model retrieval module, used to call a first deep neural network model pre-configured in the cloud platform for predicting the corrosion degree coefficient of the valve, and a second deep neural network model pre-configured for predicting the failure probability of the valve;

[0056] A corrosion prediction module, used for inputting the corrosion degree characteristic data into the first deep neural network model to obtain the corrosion degree coefficient of each valve;

[0057] A failure prediction module is used to obtain failure reflection characteristic data of each valve within the range from the past time TN to the current time T, and input the failure reflection characteristic data into the second deep neural network model to obtain the failure probability of each valve at the future time R, wherein the failure reflection characteristic data includes material characteristic data, fatigue degree characteristic data and corrosion degree coefficient at each time, and T, N and R are integers greater than zero;

[0058] The state analysis module is used to determine the failure time of all target valves and each target valve according to the failure probability, calculate the remaining service life of all target valves according to the failure time, and classify the operating status according to the remaining service life to obtain the operating status classification results of all target valves.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] The present application discloses a valve remote monitoring and analysis system and method based on a cloud platform, including: collecting corrosion degree characteristic data of each valve; calling a first deep neural network model and a second deep neural network model; inputting the corrosion degree characteristic data into the first deep neural network model to obtain the corrosion degree coefficient of each valve; obtaining the failure reflection characteristic data of each valve, and inputting the failure reflection characteristic data into the second deep neural network model to obtain the failure probability of each valve at a future moment; determining the failure time of all target valves and each target valve according to the failure probability, calculating the remaining service life of all target valves according to the failure time, and grading the operating status according to the remaining service life; based on the above technical features, the present invention overcomes the shortcomings of the traditional method of poor real-time performance, inaccurate evaluation and difficulty in predicting failure risks; at the same time, with the help of cloud computing and artificial intelligence technology, the present invention can dynamically evaluate the corrosion degree, failure probability and remaining life of the valve, thereby realizing hierarchical management and differentiated maintenance of the operating status, reducing maintenance costs, and improving system safety and operating efficiency; it is conducive to providing reliable technical support for the efficient operation and maintenance and intelligent upgrading of industrial pipeline systems, and has significant application value and industry promotion potential. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A flow chart of a valve remote monitoring and analysis method based on a cloud platform provided by the present invention;

[0062] Figure 2 A schematic diagram of the module structure of a cloud platform-based valve remote monitoring and analysis system provided by the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not 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.

[0064] Example 1

[0065] See also Figure 1 As shown, this embodiment discloses a valve remote monitoring and analysis method based on a cloud platform, the method comprising:

[0066] S101: Collect the corrosion characteristic data of each valve in the pipeline and upload it to the cloud platform;

[0067] It should be understood that valves are widely used in various types of pipelines, such as water supply pipelines and gas supply pipelines. They are key equipment in fluid control systems and are used to adjust and control parameters such as flow, pressure, and temperature of fluids to ensure safe and efficient operation of the system. However, since valves are in harsh scenarios for a long time, especially in high-pressure, high-temperature or corrosive environments, and are subject to chemical reactions and mechanical loads, long-term operation will lead to valve failure, and valve failure may cause leakage, explosion and other serious accidents. Therefore, it is necessary to determine the remaining available time of the valve through life estimation, achieve targeted maintenance, avoid unnecessary frequent maintenance, reduce emergency repairs caused by sudden failures, and reduce the difficulty of maintenance work and emergency procurement costs.

[0068] The corrosion degree characteristic data include but are not limited to signal propagation time, signal natural frequency change, signal amplitude change and damping ratio, etc.;

[0069] It should be noted that: the corrosion degree characteristic data is measured by an acoustic sensor, and the acoustic sensor includes but is not limited to a vibration sensor and an ultrasonic sensor.

[0070] S102: calling a first deep neural network model pre-configured in the cloud platform for predicting a corrosion degree coefficient of a valve, and a second deep neural network model pre-configured in the cloud platform for predicting a failure probability of a valve;

[0071] In implementation, the training method of the first deep neural network model is as follows:

[0072] Acquire historical corrosion degree training data, and divide the historical corrosion degree training data into a corrosion degree training set and a corrosion degree test set; the historical corrosion degree training data includes corrosion degree characteristic data and its corresponding corrosion degree coefficient;

[0073] It should be noted that the corrosion degree characteristic data and corrosion degree coefficient in the historical corrosion degree training data are collected, recorded or calculated by technicians based on actual scenarios or experimental data;

[0074] The method for obtaining the corrosion degree coefficient in the historical corrosion degree training data is as follows:

[0075] Obtain internal damage images of valves;

[0076] It should be noted that the acquisition of the internal damage image of the valve is achieved through non-contact non-destructive testing technology, for example, high-resolution internal images can be generated through ultrasonic testing (UT) (such as phased array ultrasonic (PAUT) or total focusing method (TFM), etc.), and can also be obtained by non-destructive methods such as X-ray imaging (Radiographic Testing, RT);

[0077] Grayscale the internal damage image to obtain a grayscale image;

[0078] Acquire a damage type reference image of the valve, and use the grayscale image as the target image, and use the damage type reference image as the matching image, wherein the damage types include cracks, holes, and bubbles;

[0079] It should be understood that damage type reference images including crack, hole and bubble damage types are stored in the system database. By matching the damage type reference image of each damage type with the internal damage image, the internal damage condition of the valve can be obtained.

[0080] Set the step size to 1, and perform cross-correlation calculation between the matching image and the target image through a sliding window method to obtain the similarity of each overlapping part;

[0081] It should be understood that cross-correlation calculation is a method of image processing, and its basic idea is to use one image as a template and another image as a matching object, place the matching object on the template, slide the matching object in the template, and calculate the similarity of each overlapping part to obtain the same area of ​​the two images. The calculation formula of the similarity includes but is not limited to one of cosine similarity or Euclidean distance;

[0082] The overlapping parts with similarity greater than a preset similarity threshold are regarded as damage areas, the number of damage areas of each damage type is counted, and the pixel area of ​​the damage area of ​​each damage type is obtained;

[0083] The total pixel area of ​​the target image is obtained, and the corrosion degree coefficient of the valve is calculated according to the number of damaged areas, the pixel area of ​​the damaged areas and the total pixel area;

[0084] The specific calculation formula of the corrosion degree coefficient is as follows:

[0085] ;

[0086] Where: Represents the corrosion severity coefficient, represents the pixel area of ​​the damaged area of ​​the vth damage type, represents the number of damaged areas of the vth damage type, represents the weight factor of the vth damage type, Indicates the total number of damage types, Represents the total pixel area of ​​the target image; wherein the weight factor of each damage type is manually designed and determined by technicians based on experimental data or historical statistical data;

[0087] Constructing a first deep neural network, taking the corrosion degree feature data in the corrosion degree training set as the input of the first deep neural network, taking the corrosion degree coefficient in the corrosion degree training set as the output of the first deep neural network, and taking minimizing the sum of the first prediction accuracies of all training samples as the training goal, training the first deep neural network until the sum of the first prediction accuracies reaches convergence, and stopping the training to obtain an initial first deep neural network;

[0088] The calculation formula of the sum of the first prediction accuracy is as follows:

[0089] ;

[0090] Where: represents the sum of the first prediction accuracy, represents the actual corrosion degree coefficient of the i-th group of training samples, represents the predicted corrosion degree coefficient of the i-th group of training samples, is the number of training samples in the corrosion degree training set;

[0091] The initial first deep neural network is verified using a corrosion degree test set, and when the initial first deep neural network is less than or equal to the first prediction accuracy, it is used as a first deep neural network model for predicting the corrosion degree coefficient of the valve;

[0092] In implementation, the training method of the second deep neural network model is as follows:

[0093] Acquire historical failure probability training data, and divide the historical failure probability training data into a failure probability training set and a failure probability test set; the historical failure probability training data includes failure reflection characteristic data and its corresponding failure probability;

[0094] The failure reflection characteristic data include material characteristic data, fatigue characteristic data and corrosion degree coefficient at each moment; the material characteristic data include but are not limited to material density, material hardness, material strength and material toughness, etc.; the fatigue characteristic data include but are not limited to the number of opening and closing times, operating time and average stress value, etc.;

[0095] It should be noted that the failure reflection characteristic data and failure probability in the historical failure probability training data are collected, recorded or calculated by technicians based on actual scenarios or experimental data; wherein the material characteristic data has been manually entered into the system database in advance;

[0096] Constructing a second deep neural network, taking the failure reflection feature data in the failure probability training set as the input of the second deep neural network, taking the failure probability in the failure probability training set as the output of the second deep neural network, and taking minimizing the sum of the second prediction accuracies of all training samples as the training goal, training the second deep neural network until the sum of the second prediction accuracies reaches convergence, and stopping the training to obtain an initial second deep neural network;

[0097] The calculation formula of the sum of the second prediction accuracy is as follows:

[0098] ;

[0099] Where: represents the sum of the second prediction accuracy, represents the actual failure probability of the jth group of training samples, represents the predicted failure probability of the jth group of training samples, is the number of samples in the corrosion degree training set;

[0100] The initial second deep neural network is verified using a failure probability test set, and when the initial second deep neural network is less than or equal to the second prediction accuracy, it is used as the second deep neural network model for predicting the failure probability of the valve.

[0101] S103: Inputting the corrosion degree characteristic data into the first deep neural network model to obtain the corrosion degree coefficient of each valve.

[0102] S104: Obtain failure reflection characteristic data of each valve in the range from the past time TN to the current time T, and input the failure reflection characteristic data into the second deep neural network model to obtain the failure probability of each valve at the future time R, wherein the failure reflection characteristic data includes material characteristic data, fatigue degree characteristic data and corrosion degree coefficient at each time, and T, N and R are integers greater than zero.

[0103] S105: determining the failure time of all target valves and each target valve according to the failure probability, calculating the remaining service life of all target valves according to the failure time, and classifying the operating status according to the remaining service life to obtain the operating status classification results of all target valves;

[0104] In implementation, determining the failure time of each valve according to the failure probability includes:

[0105] Extract the failure probability of all valves at the future time R;

[0106] The failure probability of each valve is compared with the preset failure probability threshold. If the failure probability is greater than or equal to the preset failure probability threshold, the corresponding valve is marked as the target valve, and the future time R is used as the failure time of the target valve; if the failure probability is less than the preset failure probability threshold, the corresponding valve is marked as a non-target valve;

[0107] Count all valves whose failure probability is greater than or equal to the failure probability threshold, and record the future time R, then obtain all target valves and the failure time of each target valve;

[0108] In implementation, the remaining service life of all target valves is calculated based on the failure time, including:

[0109] Extract the failure time of each target valve;

[0110] Calculate the difference between the failure time and the current time T, and take the difference between the failure time and the current time T as the remaining service life;

[0111] The difference between the failure time of each target valve and the current time T is counted to obtain the remaining service life of all target valves;

[0112] In implementation, the operation status classification according to the remaining service life includes:

[0113] Set remaining service life thresholds Td1 and Td2, where Td1>Td2>0;

[0114] comparing the remaining useful life of each target valve with a preset remaining useful life threshold;

[0115] If the remaining service life is greater than or equal to the remaining service life threshold value Td1, it is determined that the corresponding target valve belongs to the first-level operating state level, and the first-level operating state level is output as the operating state classification result;

[0116] If the remaining service life is less than the remaining service life threshold value Td1 and greater than the remaining service life threshold value Td2, it is determined that the corresponding target valve belongs to the second-level operating state level, and the second-level operating state level is output as the operating state classification result;

[0117] If the remaining service life is less than or equal to the remaining service life threshold value Td2, it is determined that the corresponding target valve belongs to the third-level operating state level, and the third-level operating state level is output as the operating state classification result;

[0118] It should be understood that: when the target valve belongs to the first level of operation status, it means that the corresponding target valve is still a long way from complete failure, its current performance is good, and no additional maintenance operation is required. It only needs to monitor its operation status regularly according to the regular plan to ensure continuous safe and efficient operation; when the target valve belongs to the second level of operation status, it means that the target valve has entered the early warning stage, its remaining life is limited, and it may fail in the short to medium term. At this time, the monitoring frequency of the valve operation status should be increased; when the target valve belongs to the third level of operation status, it means that the target valve has entered the high-risk stage, its remaining life is extremely short, and it may fail at any time. At this time, the valve maintenance or replacement should be arranged as a priority to avoid its sudden failure causing system interruption or safety accidents;

[0119] By collecting multi-dimensional feature data in real time and combining it with a deep neural network model for precise analysis, the present invention overcomes the shortcomings of traditional methods such as poor real-time performance, inaccurate evaluation, and difficulty in predicting failure risks. With the help of cloud computing and artificial intelligence technology, the present invention can dynamically evaluate the corrosion degree, failure probability, and remaining life of the valve, thereby realizing hierarchical management and differentiated maintenance of the operating status, reducing maintenance costs, and improving system safety and operating efficiency. It is conducive to providing reliable technical support for the efficient operation and maintenance and intelligent upgrading of industrial pipeline systems, and has significant application value and industry promotion potential.

[0120] Example 2

[0121] See also Figure 2 As shown, based on the same inventive concept, this embodiment discloses a valve remote monitoring and analysis system based on a cloud platform. For details not provided in this embodiment, please refer to the description of the relevant parts in Embodiment 1. The system includes:

[0122] The data acquisition module 210 is used to collect the corrosion characteristic data of each valve in the pipeline and upload it to the cloud platform;

[0123] A model retrieval module 220, used to call a first deep neural network model pre-configured in the cloud platform for predicting the corrosion degree coefficient of the valve, and a second deep neural network model pre-configured in the cloud platform for predicting the failure probability of the valve;

[0124] A corrosion prediction module 230, for inputting the corrosion degree characteristic data into the first deep neural network model to obtain the corrosion degree coefficient of each valve;

[0125] A failure prediction module 240 is used to obtain failure reflection characteristic data of each valve within the range from the past time TN to the current time T, and input the failure reflection characteristic data into the second deep neural network model to obtain the failure probability of each valve at the future time R, wherein the failure reflection characteristic data includes material characteristic data, fatigue degree characteristic data and corrosion degree coefficient at each time, and T, N and R are integers greater than zero;

[0126] The state analysis module 250 is used to determine the failure time of all target valves and each target valve according to the failure probability, calculate the remaining service life of all target valves according to the failure time, and classify the operating status according to the remaining service life to obtain the operating status classification results of all target valves.

[0127] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters, weights and thresholds in the formula are set by technicians in this field according to actual conditions.

[0128] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired network or a wireless network. The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD) or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0129] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0130] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0131] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0132] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0133] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0134] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0135] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A valve remote monitoring and analysis method based on a cloud platform, characterized in that: The method comprises: Collect the corrosion characteristic data of each valve in the pipeline and upload it to the cloud platform; Calling a first deep neural network model pre-configured in the cloud platform for predicting the corrosion degree coefficient of the valve, and a second deep neural network model pre-configured for predicting the failure probability of the valve; Inputting the corrosion degree characteristic data into the first deep neural network model to obtain the corrosion degree coefficient of each valve; Obtain failure reflection characteristic data of each valve in the range from the past time TN to the current time T, and input the failure reflection characteristic data into the second deep neural network model to obtain the failure probability of each valve at the future time R, wherein the failure reflection characteristic data includes material characteristic data, fatigue degree characteristic data and corrosion degree coefficient at each time, and T, N and R are integers greater than zero; The failure time of all target valves and each target valve is determined according to the failure probability, the remaining service life of all target valves is calculated according to the failure time, and the operating status is graded according to the remaining service life to obtain the operating status classification results of all target valves.

2. The cloud platform-based valve remote monitoring and analysis method according to claim 1 is characterized in that: The corrosion degree characteristic data include but are not limited to signal propagation time, signal natural frequency change, signal amplitude change and damping ratio.

3. The cloud platform-based valve remote monitoring and analysis method according to claim 2 is characterized in that: The training method of the first deep neural network model is as follows: Acquire historical corrosion degree training data, and divide the historical corrosion degree training data into a corrosion degree training set and a corrosion degree test set; the historical corrosion degree training data includes corrosion degree characteristic data and its corresponding corrosion degree coefficient; Constructing a first deep neural network, taking the corrosion degree feature data in the corrosion degree training set as the input of the first deep neural network, taking the corrosion degree coefficient in the corrosion degree training set as the output of the first deep neural network, and taking minimizing the sum of the first prediction accuracies of all training samples as the training goal, training the first deep neural network until the sum of the first prediction accuracies reaches convergence, and stopping the training to obtain an initial first deep neural network; The calculation formula of the sum of the first prediction accuracy is as follows: ； Where: represents the sum of the first prediction accuracy, represents the actual corrosion degree coefficient of the i-th group of training samples, represents the predicted corrosion degree coefficient of the i-th group of training samples, is the number of training samples in the corrosion degree training set; The initial first deep neural network is verified using the corrosion degree test set. When the initial first deep neural network is less than or equal to the first prediction accuracy, it is used as the first deep neural network model for predicting the corrosion degree coefficient of the valve.

4. The cloud platform-based valve remote monitoring and analysis method according to claim 3 is characterized in that: The method for obtaining the corrosion degree coefficient in the historical corrosion degree training data is as follows: Obtain internal damage images of valves; Grayscale the internal damage image to obtain a grayscale image; Acquire a damage type reference image of the valve, and use the grayscale image as the target image, and use the damage type reference image as the matching image, wherein the damage types include cracks, holes, and bubbles; Set the step size to 1, and perform cross-correlation calculation between the matching image and the target image through a sliding window method to obtain the similarity of each overlapping part; The overlapping parts with similarity greater than a preset similarity threshold are regarded as damage areas, the number of damage areas of each damage type is counted, and the pixel area of ​​the damage area of ​​each damage type is obtained; The total pixel area of ​​the target image is obtained, and the corrosion degree coefficient of the valve is calculated according to the number of damaged areas, the pixel area of ​​the damaged areas and the total pixel area; The specific calculation formula of the corrosion degree coefficient is as follows: ； Where: Represents the corrosion severity coefficient, represents the pixel area of ​​the damaged area of ​​the vth damage type, represents the number of damaged areas of the vth damage type, represents the weight factor of the vth damage type, Indicates the total number of damage types, Represents the total pixel area of ​​the target image.

5. The cloud platform-based valve remote monitoring and analysis method according to claim 4 is characterized in that: The training method of the second deep neural network model is as follows: Acquire historical failure probability training data, and divide the historical failure probability training data into a failure probability training set and a failure probability test set; the historical failure probability training data includes failure reflection characteristic data and its corresponding failure probability; The failure reflection characteristic data include material characteristic data, fatigue characteristic data and corrosion degree coefficient at each moment; the material characteristic data include material density, material hardness, material strength and material toughness; the fatigue characteristic data include the number of opening and closing times, operating time and average stress value; Constructing a second deep neural network, taking the failure reflection feature data in the failure probability training set as the input of the second deep neural network, taking the failure probability in the failure probability training set as the output of the second deep neural network, and taking minimizing the sum of the second prediction accuracies of all training samples as the training goal, training the second deep neural network until the sum of the second prediction accuracies reaches convergence, and stopping the training to obtain an initial second deep neural network; The calculation formula of the sum of the second prediction accuracy is as follows: ； Where: represents the sum of the second prediction accuracy, represents the actual failure probability of the jth group of training samples, represents the predicted failure probability of the jth group of training samples, is the number of samples in the corrosion degree training set; The initial second deep neural network is verified using a failure probability test set, and when the initial second deep neural network is less than or equal to the second prediction accuracy, it is used as the second deep neural network model for predicting the failure probability of the valve.

6. The cloud platform-based valve remote monitoring and analysis method according to claim 5 is characterized in that: Determining the failure time of each valve according to the failure probability includes: Extract the failure probability of all valves at the future time R; The failure probability of each valve is compared with the preset failure probability threshold. If the failure probability is greater than or equal to the preset failure probability threshold, the corresponding valve is marked as the target valve, and the future time R is used as the failure time of the target valve; if the failure probability is less than the preset failure probability threshold, the corresponding valve is marked as a non-target valve; By counting all valves whose failure probability is greater than or equal to the failure probability threshold and recording the future time R, the failure time of all target valves and each target valve is obtained.

7. The cloud platform-based valve remote monitoring and analysis method according to claim 6 is characterized in that: The remaining service life of all target valves is calculated based on the failure time, including: Extract the failure time of each target valve; Calculate the difference between the failure time and the current time T, and take the difference between the failure time and the current time T as the remaining service life; The difference between the failure time of each target valve and the current time T is counted to obtain the remaining service life of all target valves.

8. The cloud platform-based valve remote monitoring and analysis method according to claim 7 is characterized in that: The operating status classification according to the remaining service life includes: Set remaining service life thresholds Td1 and Td2, where Td1>Td2>0; comparing the remaining useful life of each target valve with a preset remaining useful life threshold; If the remaining service life is greater than or equal to the remaining service life threshold value Td1, it is determined that the corresponding target valve belongs to the first-level operating state level, and the first-level operating state level is output as the operating state classification result; If the remaining service life is less than the remaining service life threshold value Td1 and greater than the remaining service life threshold value Td2, it is determined that the corresponding target valve belongs to the second-level operating state level, and the second-level operating state level is output as the operating state classification result; If the remaining service life is less than or equal to the remaining service life threshold value Td2, it is determined that the corresponding target valve belongs to the third-level operating state level, and the third-level operating state level is output as the operating state classification result.

9. A valve remote monitoring and analysis system based on a cloud platform, characterized in that: The system is implemented based on the cloud platform-based valve remote monitoring and analysis method according to any one of claims 1 to 8, and the system comprises: The data acquisition module is used to collect the corrosion characteristic data of each valve in the pipeline and upload it to the cloud platform; A model retrieval module, used to call a first deep neural network model pre-configured in the cloud platform for predicting the corrosion degree coefficient of the valve, and a second deep neural network model pre-configured for predicting the failure probability of the valve; A corrosion prediction module, used for inputting the corrosion degree characteristic data into the first deep neural network model to obtain the corrosion degree coefficient of each valve; A failure prediction module is used to obtain failure reflection characteristic data of each valve within the range from the past time TN to the current time T, and input the failure reflection characteristic data into the second deep neural network model to obtain the failure probability of each valve at the future time R, wherein the failure reflection characteristic data includes material characteristic data, fatigue degree characteristic data and corrosion degree coefficient at each time, and T, N and R are integers greater than zero; The state analysis module is used to determine the failure time of all target valves and each target valve according to the failure probability, calculate the remaining service life of all target valves according to the failure time, and classify the operating status according to the remaining service life to obtain the operating status classification results of all target valves.

Images