A method for cleaning oil pipes by combining medium frequency heating with mechanical cleaning

Through the combination of medium frequency heating and mechanical cleaning and combined with artificial intelligence control, the problems of low cleaning efficiency and environmental pollution of traditional oil pipes are solved, and efficient and environmentally friendly oil pipe cleaning effects are achieved.

CN118751619BActive Publication Date: 2024-12-03XINJIANG JIN CHENG PETROLEUM CHEM EQUIP CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411242092.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-05
Publication Date
2024-12-03
Estimated Expiration
2044-09-05

AI Technical Summary

Technical Problem

Traditional oil pipe cleaning methods are inefficient and labor-intensive, which can easily damage the material of the oil pipe and produce a large amount of oil-containing sewage and seriously pollute the environment.

Method used

Using a combination of intermediate frequency heating and mechanical cleaning, the oil pipe is heated through an induction heating system and physically cleaned by an external brushing device. Combined with data processing and image analysis algorithm based on artificial intelligence and deep learning, the power supply voltage frequency is controlled in real time, and adaptively Adjust the heating temperature.

Benefits of technology

It improves the oil pipe cleaning efficiency, avoids damage to the oil pipe material, reduces the generation of sewage, and achieves a more environmentally friendly and efficient cleaning effect.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118751619B_ABST
    Figure CN118751619B_ABST
Patent Text Reader

Abstract

The present application relates to a method for cleaning an oil pipe that combines medium frequency heating with mechanical cleaning. The method comprises: conveying the oil pipe into a cleaning box; in the cleaning box, heating the oil pipe by an induction heating system while feeding the oil pipe and cleaning the oil pipe by an external brushing device; purging the oil pipe by compressed air in the cleaning box; and driving the oil pipe to exit the cleaning box. Thus, in the method for cleaning an oil pipe that combines medium frequency heating with mechanical cleaning, the oil pipe is heated by an induction heating system, and the magnetic field generated by the medium frequency power supply voltage is used to make the metal molecules on the inner wall of the oil pipe move at high speed, generate heat, and thus soften or melt the dirt. At the same time, the oil pipe is physically cleaned by an external brushing device to remove the dirt on the surface of the heated oil pipe.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of oil pipe cleaning, and in particular to an oil pipe cleaning method combining medium frequency heating with mechanical cleaning. Background Art

[0002] At present, a large number of oil pipes used in oil fields often accumulate a lot of oil pollution after leaving the well. In order to reduce costs, these used oil pipes are subsequently reused by cleaning and repairing, so that the repaired qualified oil pipes can be recycled. However, the effect of traditional cleaning methods is often unsatisfactory, which will damage the metallographic structure of the oil pipe body, thus posing a certain degree of quality risk.

[0003] At present, the oil pipe cleaning generally adopts the process of water heating combined with chemical agents to clean the inner and outer walls of the oil pipe. Although this method can remove the surface dirt of the oil pipe to a certain extent, it has low production efficiency and is accompanied by high labor intensity. The cleaning effect is often not ideal and sometimes damages the oil pipe material, affecting the reuse and safe use of the oil pipe. In addition, another significant disadvantage of the traditional cleaning method is that it will produce a large amount of oily wastewater, which will cause serious pollution to the environment.

[0004] Therefore, an optimized oil pipe cleaning solution is desired. Summary of the invention

[0005] This summary is provided to introduce concepts in a brief form that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0006] In the first aspect, the present application provides an oil pipe cleaning method combining medium frequency heating and mechanical cleaning, the method comprising: conveying the oil pipe into a cleaning box; in the cleaning box, heating the oil pipe by an induction heating system while feeding the oil pipe and cleaning the oil pipe by an external brushing device, wherein the induction heating system comprises: a power supply, an electric heating coil electrically connected to the power supply, and a controller electrically connected to the power supply, the controller being used to control the voltage frequency of the power supply; purging the oil pipe by compressed air in the cleaning box; and driving the oil pipe to exit the cleaning box.

[0007] Optionally, in the cleaning box, while feeding the oil pipe, the oil pipe is heated by an induction heating system and cleaned by an external brushing device, including: obtaining a surface state image of the cleaned oil pipe collected by a camera; obtaining a time series of temperature values ​​of the induction heating system; performing feature extraction on the surface state image of the cleaned oil pipe to obtain an oil pipe surface state feature map; passing the oil pipe surface state feature map through a foreground mask enhancer based on a gated convolutional feedforward network to obtain an oil pipe surface state foreground enhancement feature map; passing the time series of temperature values ​​of the induction heating system through a temperature time series encoder based on a Bi-LSTM model to obtain a temperature time series associated feature vector; passing the oil pipe surface state foreground enhancement feature map and the temperature time series associated feature vector through a heterogeneous Transformer cross-domain joint encoder based on multimodal assistance to obtain a temperature time series modulated oil pipe surface state optimized expression feature map; based on the temperature time series modulated oil pipe surface state optimized expression feature map, determining a control result of the power supply voltage frequency, the control result being used to indicate whether the power supply voltage frequency at the current time point should increase, decrease or remain unchanged.

[0008] Optionally, feature extraction is performed on the surface state image of the cleaned oil pipe to obtain an oil pipe surface state feature map, including: passing the surface state image of the cleaned oil pipe through an oil pipe surface state feature extractor based on a hole convolutional neural network model to obtain the oil pipe surface state feature map.

[0009] Optionally, the oil pipe surface state feature map is passed through a foreground mask enhancer based on a gated convolution feedforward network to obtain an oil pipe surface state foreground enhancement feature map, including: performing layer normalization processing on the oil pipe surface state feature map to obtain a normalized oil pipe surface state feature map; performing channel expansion based on point convolution and deep convolution encoding based on a hole convolution layer on the normalized oil pipe surface state feature map to obtain an oil pipe surface state deep convolution backup feature map and an oil pipe surface state deep convolution original feature map; inputting the oil pipe surface state deep convolution original feature map into a network based on The foreground gated mask module of the function is used to obtain a deep convolution gated mask weight feature map of the oil pipe surface state; the position point multiplication between the deep convolution gated mask weight feature map of the oil pipe surface state and the deep convolution backup feature map of the oil pipe surface state is calculated to obtain a foreground highlighting feature map of the oil pipe surface state gated mask; and the channel contraction based on point convolution is performed on the foreground highlighting feature map of the oil pipe surface state gated mask to obtain the foreground strengthening feature map of the oil pipe surface state.

[0010] Optionally, the oil pipe surface state deep convolution original feature map is input based on The foreground gated mask module of the function is used to obtain a deep convolution gated mask weight feature map of the surface state of the oil pipe, including: inputting the deep convolution original feature map of the surface state of the oil pipe into After activation in the function, mask processing is performed on the feature values ​​of each position in the activated deep convolution original feature map of the oil pipe surface state to obtain the deep convolution gated mask weight feature map of the oil pipe surface state, wherein, in response to the feature value in the activated deep convolution original feature map of the oil pipe surface state being greater than a predetermined hyperparameter, the feature value is set to 1, otherwise it is set to 0.

[0011] Optionally, the oil pipe surface state prospect enhancement feature map and the temperature time series associated feature vector are passed through a heterogeneous Transformer cross-domain joint encoder based on multi-modal assistance to obtain a temperature time series modulated oil pipe surface state optimization expression feature map, including: performing linear transformation and feature shape reshaping on the temperature time series associated feature vector to obtain a modulated temperature time series associated feature matrix; performing a first linear transformation and a second linear transformation on the oil pipe surface state prospect enhancement feature map to obtain a first linear transformation oil pipe surface state prospect enhancement feature map and a second linear transformation oil pipe surface state prospect enhancement feature map; performing feature shape reshaping on the first linear transformation oil pipe surface state prospect enhancement feature map and the second linear transformation oil pipe surface state prospect enhancement feature map to obtain a first linear transformation oil pipe surface state prospect enhancement feature matrix and a second linear transformation oil pipe surface state prospect enhancement feature matrix; using the first linear transformation oil pipe surface state prospect enhancement feature matrix and the second linear transformation oil pipe surface state prospect enhancement feature matrix as a key matrix and a value matrix and using the modulated temperature time series associated feature matrix as a query matrix, inputting the query matrix, the key matrix and the value matrix into a Transformer structure to obtain a temperature time series attention query oil pipe surface state feature matrix; reshaping and point convolution encoding are performed on the temperature time series attention query oil pipe surface state feature matrix to obtain a temperature time series attention query oil pipe surface state feature map; performing position point multiplication on the temperature time series attention query oil pipe surface state feature map and the oil pipe surface state prospect enhancement feature map to obtain the temperature time series modulated oil pipe surface state optimized expression feature map.

[0012] Optionally, the temperature timing association feature vector is linearly transformed and feature-reshaped to obtain a modulated temperature timing association feature matrix, including: calculating the matrix multiplication between the temperature timing association feature vector and the vector modulation weight matrix, and then adding the matrix multiplication with the vector modulation bias vector by position to obtain a modulated temperature timing association feature vector; activating the modulated temperature timing association feature vector through a Sigmoid function to obtain a linearly transformed temperature timing association feature vector; and reshaping the linearly transformed temperature timing association feature vector to obtain the modulated temperature timing association feature matrix.

[0013] Optionally, performing a first linear transformation and a second linear transformation on the oil pipe surface state prospect enhancement feature map to obtain a first linear transformation oil pipe surface state prospect enhancement feature map and a second linear transformation oil pipe surface state prospect enhancement feature map, including: calculating the matrix multiplication between each feature matrix along the channel dimension in the oil pipe surface state prospect enhancement feature map and the first feature map modulation weight matrix, and then adding the matrix multiplication with the first feature map modulation bias matrix in a positional manner to obtain the first modulated oil pipe surface state prospect enhancement feature map; activating the first modulated oil pipe surface state prospect enhancement feature map by a Sigmoid function to obtain the first linear transformation oil pipe surface state prospect enhancement feature map; calculating the matrix multiplication between each feature matrix along the channel dimension in the oil pipe surface state prospect enhancement feature map and the second feature map modulation weight matrix, and then adding the matrix multiplication with the second feature map modulation bias matrix in a positional manner to obtain the second modulated oil pipe surface state prospect enhancement feature map; activating the second modulated oil pipe surface state prospect enhancement feature map by a Sigmoid function to obtain the second linear transformation oil pipe surface state prospect enhancement feature map.

[0014] Optionally, the first linear transformation oil pipe surface state prospect enhancement feature matrix and the second linear transformation oil pipe surface state prospect enhancement feature matrix are used as a key matrix and a value matrix and the modulated temperature time series association feature matrix is ​​used as a query matrix, and the query matrix, the key matrix and the value matrix are input into a Transformer structure to obtain a temperature time series attention query oil pipe surface state feature matrix, including: after calculating the matrix multiplication between the transposed vector of the modulated temperature time series association feature matrix and the first linear transformation oil pipe surface state prospect enhancement feature matrix, the obtained feature matrix is ​​probabilistically activated through a softmax function to obtain a probabilistic oil pipe surface state prospect enhancement weight matrix; and the matrix multiplication between the probabilistic oil pipe surface state prospect enhancement weight matrix and the second linear transformation oil pipe surface state prospect enhancement feature matrix is ​​calculated to obtain the temperature time series attention query oil pipe surface state feature matrix.

[0015] Optionally, based on the temperature-series timing-modulated oil pipe surface state optimization expression characteristic graph, a control result of the power supply voltage frequency is determined, and the control result is used to indicate whether the power supply voltage frequency at the current time point should increase, decrease, or remain unchanged, including: passing the temperature-series timing-modulated oil pipe surface state optimization expression characteristic graph through a classifier-based power supply voltage frequency controller to obtain a control result, and the control result is used to indicate whether the power supply voltage frequency at the current time point should increase, decrease, or remain unchanged.

[0016] By adopting the above technical solution, the oil pipe is transported into the cleaning box; in the cleaning box, the oil pipe is heated by the induction heating system while being fed and cleaned by the external brushing device; the oil pipe is purged by compressed air in the cleaning box; and the oil pipe is driven to exit the cleaning box. In this way, in the oil pipe cleaning method combining medium frequency heating with mechanical cleaning, the oil pipe is heated by the induction heating system, and the magnetic field generated by the medium frequency power supply voltage is used to make the metal molecules on the inner wall of the oil pipe move at high speed, generate heat, and thus soften or melt the dirt. At the same time, the oil pipe is physically cleaned by the external brushing device to remove the dirt on the surface of the heated oil pipe.

[0017] Other features and advantages of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above and other features, advantages and aspects of the embodiments of the present application will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0019] In the attached picture: Figure 1 It is a flow chart of an oil pipe cleaning method combining medium frequency heating and mechanical cleaning according to an exemplary embodiment.

[0020] Figure 2 is based on Figure 1 The illustrated embodiment shows a flow chart of step S102 of an oil pipe cleaning method combining medium frequency heating with mechanical cleaning.

[0021] Figure 3 It is a block diagram of an oil pipe cleaning system combining medium frequency heating and mechanical cleaning according to an exemplary embodiment.

[0022] Figure 4 It is a schematic structural diagram of an oil pipe combining medium frequency heating and mechanical cleaning according to an exemplary embodiment.

[0023] Figure 5It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0024] The embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be construed as being limited to the embodiments described herein. Instead, these embodiments are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are only for exemplary purposes and are not intended to limit the scope of protection of the present application.

[0025] It should be understood that the various steps described in the method implementation of the present application can be performed in different orders and / or performed in parallel. In addition, the method implementation may include additional steps and / or omit the steps shown. The scope of the present application is not limited in this respect.

[0026] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0027] It should be noted that the concepts such as "first" and "second" mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0028] It should be noted that the modifications of "one" and "plurality" mentioned in the present application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0029] The names of the messages or information exchanged between multiple devices in the embodiments of the present application are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0030] As the most basic and critical part of the oil pipe repair process, the cleaning quality of the oil pipe has a significant impact on the use cost of the oil pipe and the quality of the construction work. At the same time, the cleaning method also affects the cleaning effect and cost, repair quality, operational safety and environmental protection. At present, there are several ways to clean the oil pipe: physical cleaning is a way to remove dirt on the surface of an object by relying on external energy, such as mechanical friction, ultrasound, negative pressure, high pressure, and ultraviolet light. At present, the physical cleaning method used for oil field oil pipe cleaning is mainly high-pressure water jet cleaning. This technology uses clean water as the medium, and generates high-pressure and high-speed water jets through high-pressure plunger pumps and specially designed nozzles. The jets are continuously shot at the inner and outer surfaces of the oil pipe at a certain angle with a strong impact force, thereby directly peeling off and flushing the scale on the oil pipe wall to make it fall off, but its operating cost is high and the efficiency is low.

[0031] Chemical cleaning is a method of removing dirt from the surface of an object by using chemicals or other solvents, relying on the action of chemical reactions. This method uses boiler heating and high-temperature chemical cleaning, that is, the oil pipes recovered from the oil well are placed in a pool for soaking, the temperature is maintained at 80-90℃, and then chemical cleaning agents are added for boiling.

[0032] Diesel heating soaking is to soak the oil pipes recovered from the well site in high-temperature diesel (usually above 80°C) for a certain period of time to remove the colloid, scale and oil stains on the inner and outer walls of the oil pipes. This method has simple equipment and low investment, but high energy consumption, consuming hundreds of tons of diesel every year, and has high cleaning costs, low efficiency and poor quality.

[0033] High-temperature alkaline solution cleaning is a chemical cleaning method that uses boiler steam or gas to directly heat the cleaning agent and oil pipes to 80-90°C, soak them for a certain period of time, and then rinse them with hot water to remove the oil stains on the surface of the oil pipes. The cleaning quality is affected by comprehensive factors such as the cleaning liquid temperature, concentration, and flushing pressure, and the cleaning quality is difficult to control.

[0034] Heating cleaning includes: (1) Induction heating cleaning technology, which uses the medium frequency induction principle to instantly heat the oil pipe, so that the temperature of the inner and outer surfaces of the oil pipe rises rapidly, and the oil stains and wax deposits melt instantly. This technology can use the inverter to adjust the power supply voltage frequency according to the dirt on the surface of the oil pipe, so as to properly adjust the oil pipe feed speed. By adjusting the appropriate speed and temperature to induction heat the oil pipe, the metallographic structure of the oil pipe body is avoided from being damaged. (2) Far infrared heating technology, as the temperature decreases, the oil stains on the oil pipe quickly harden, wax and solidify. When the temperature drops below 18°C, the cleaning resistance inside the oil pipe increases, affecting normal production. In order to effectively clean it, it is usually necessary to preheat the oil pipe. When the ambient temperature is below 18°C, the far infrared oil pipe preheating system automatically starts, and the fully automatic natural gas furnace is ignited. The burning flame is converted into infrared radiation through the special far infrared tube, which is reflected by the preheating cover and irradiates the oil pipe to be cleaned for preheating.

[0035] In order to solve the above problems, the present application provides an oil pipe cleaning method combining medium frequency heating with mechanical cleaning, which is as follows: the oil pipe is transported into a cleaning box; in the cleaning box, the oil pipe is heated by an induction heating system while being fed into the cleaning box and cleaned by an external brushing device; the oil pipe is purged by compressed air in the cleaning box; and the oil pipe is driven to exit the cleaning box. In this way, in the oil pipe cleaning method combining medium frequency heating with mechanical cleaning, the oil pipe is heated by an induction heating system, and the magnetic field generated by the medium frequency power supply voltage is used to make the metal molecules on the inner wall of the oil pipe move at high speed, generate heat, and thus soften or melt the dirt. At the same time, the oil pipe is physically cleaned by an external brushing device to remove the dirt on the surface of the heated oil pipe.

[0036] In this application, the medium frequency heating cleaning equipment solution effectively combines the two cleaning methods together based on the previous research on medium frequency heating cleaning technology and mechanical cleaning technology. The equipment is mainly composed of loading and unloading parts, oil pipe drive system, straightening system, tool rod scraping system, induction heating system, sewage discharge system, purification and dust removal system, and automatic control system.

[0037] The technical route of this application is: Based on the previous research and discussion, the technical route of medium frequency heating and cleaning of the oil pipe has been formed: This cleaning technology is to remove impurities inside and outside the oil pipe by high-speed scraping of mechanical tools and steel brushes while induction heating the oil pipe, and discharge the heated free matter and scraped off matter to the outside of the oil pipe by compressed air sprayed from the scraper. The oil pipe is transported by the conveying line outside the box and the pipe inlet wheel group inside the box. The oil pipe passes through the feed hole of the end plate of the box body to reach the box. After entering the inclined conveying wheel group, the upper inclined conveying wheel group presses down to drive the oil pipe spiral forward, and passes through the external brushing device and the electric heating coil in turn. The air injection solenoid valve is opened, and a large amount of compressed air passes through the hollow rotating knife rod and sprays from the internal washing steel brush and scraper to blow away the melted crude oil and scraped dirt in the oil pipe until the entire oil pipe completely passes through the scraper and the internal washing steel brush. The inclined conveying wheel group reverses and drives the oil pipe to exit the box. It consists of loading and unloading parts, oil pipe driving system, straightening system, tool bar scraping system, induction heating system, induction cooling system, sewage discharge system, purification and dust removal system, and automatic control system.

[0038] The specific implementation of the present application is described in detail below with reference to the accompanying drawings.

[0039] Medium frequency heating technology is a technology that uses the principle of electromagnetic induction with a frequency range of 1 kHz to 100 kHz to heat metal. It can quickly heat the oil pipe, thereby softening and decomposing stubborn dirt and improving the subsequent cleaning effect. Applying medium frequency heating technology to oil pipe cleaning can effectively remove dirt and rust on the inner wall of the oil pipe without damaging the material of the oil pipe, thereby improving the cleaning efficiency of the oil pipe.

[0040] However, a single medium frequency heating technology may not fully meet the needs of oil pipe cleaning, especially when facing different types and degrees of dirt. Therefore, combining medium frequency heating technology with mechanical cleaning technology can give full play to the advantages of both and achieve a more efficient and thorough cleaning effect.

[0041] Figure 1 FIG. 1 is a flow chart showing a method for cleaning an oil pipe by combining medium frequency heating with mechanical cleaning according to an exemplary embodiment. Figure 1 As shown, the method includes: step S101, transporting the oil pipe into the cleaning box.

[0042] Step S102: In the cleaning box, while feeding the oil pipe, the oil pipe is heated by an induction heating system and cleaned by an external brushing device, wherein the induction heating system includes: a power supply, an electric heating coil electrically connected to the power supply, and a controller electrically connected to the power supply, and the controller is used to control the voltage frequency of the power supply.

[0043] Step S103: purge the oil pipe in the cleaning box using compressed air.

[0044] Step S104, driving the oil pipe to exit the cleaning box.

[0045] In the oil pipe cleaning method combining medium frequency heating with mechanical cleaning, the oil pipe is heated by an induction heating system, and the magnetic field generated by the medium frequency power supply voltage is used to make the metal molecules on the inner wall of the oil pipe move at high speed, generate heat, and thus soften or melt the dirt. At the same time, the oil pipe is physically cleaned by an external brushing device to remove the dirt on the surface of the heated oil pipe. In particular, in the oil pipe cleaning method, it is considered that in the process of heating the oil pipe using medium frequency heating technology, too high a temperature will destroy the metallographic structure of the oil pipe body, thereby affecting the safe use of the oil pipe, while too low a temperature will cause the dirt on the surface of the oil pipe to not be effectively cleaned and removed.

[0046] Based on this, in the technical solution of the present application, a controller is introduced, which can adaptively adjust the voltage frequency of the power supply according to the dirt cleaning situation on the surface of the oil pipe, thereby adjusting the heating temperature of the oil pipe accordingly to perform induction heating on the oil pipe, so as to avoid damage to the metallographic structure and material of the oil pipe body due to excessive temperature. In this way, the oil pipe can be cleaned without affecting the quality of the oil pipe.

[0047] Figure 2 is based on Figure 1 The embodiment shown is a flow chart of step S102 of a method for cleaning an oil pipe by combining medium frequency heating with mechanical cleaning. Figure 2 As shown, step S102, in the cleaning box, while feeding the oil pipe, the oil pipe is heated by the induction heating system and cleaned by the external brushing device, including: step S1021, obtaining the surface state image of the cleaned oil pipe collected by the camera.

[0048] Step S1022: Acquire the time series of the temperature values ​​of the induction heating system.

[0049] Step S1023: extracting features from the surface state image of the cleaned oil pipe to obtain an oil pipe surface state feature map.

[0050] Step S1024: Pass the oil pipe surface state feature map through a foreground mask enhancer based on a gated convolutional feedforward network to obtain an oil pipe surface state foreground enhanced feature map.

[0051] Step S1025 , passing the time series of the temperature values ​​of the induction heating system through a temperature time series encoder based on a Bi-LSTM model to obtain a temperature time series associated feature vector.

[0052] Step S1026: The oil pipe surface state foreground enhancement feature map and the temperature time series associated feature vector are passed through a heterogeneous Transformer cross-domain joint encoder based on multi-modal assistance to obtain a temperature time series modulated oil pipe surface state optimized expression feature map.

[0053] Step S1027, based on the temperature time series modulation oil pipe surface state optimization expression characteristic diagram, determine the control result of the power supply voltage frequency, the control result is used to indicate whether the power supply voltage frequency at the current time point should increase, decrease or remain unchanged.

[0054] Specifically, the technical concept of the present application is to collect the surface state image of the oil pipe during the cleaning process through a camera during the induction heating stage, and monitor and collect the temperature value of the induction heating system in real time, and then introduce data processing and image analysis algorithms based on artificial intelligence and deep learning in the back end to characterize the surface state image of the oil pipe and extract the time series characteristics of the heating temperature, so as to learn the semantic representation of the surface state of the oil pipe under the influence of temperature, so as to control the power supply voltage frequency in real time based on the surface state of the oil pipe, so as to adaptively adjust the induction heating temperature. In this way, intelligent heating control of oil pipe cleaning can be realized, which helps to avoid damage to the metallographic structure and material of the oil pipe body due to excessive temperature. In this way, the oil pipe can be cleaned without affecting the quality of the oil pipe.

[0055] More specifically, in the process of heating the oil pipe by the induction heating system and cleaning the oil pipe by the external brushing device, first, the surface state image of the cleaned oil pipe collected by the camera is obtained, and the time series of the temperature value of the induction heating system is obtained. Then, the surface state image of the cleaned oil pipe is subjected to feature mining in the oil pipe surface state feature extractor based on the hole convolutional neural network model to extract the implicit feature distribution information of the surface state of the cleaned oil pipe, which can reflect the current surface state of the oil pipe in the oil pipe cleaning process, and provide a basis for the subsequent heating temperature control task, thereby obtaining the surface state feature map of the oil pipe.

[0056] In one embodiment of the present application, feature extraction is performed on the surface state image of the cleaned oil pipe to obtain an oil pipe surface state feature map, including: passing the surface state image of the cleaned oil pipe through an oil pipe surface state feature extractor based on a void convolutional neural network model to obtain the oil pipe surface state feature map.

[0057] Furthermore, considering that it is crucial to understand the state of the oil pipe surface during the oil pipe cleaning process, and there may be various dirt and sediments on the oil pipe surface, these different features are presented differently at different positions and scales in the surface state image, and there are also some background interference features and redundant information in the image that are irrelevant to dirt identification and subsequent temperature control tasks, which will directly affect the cleaning effect. Therefore, in order to enhance the expression ability of the oil pipe foreground dirt features in the surface state image and help the model better focus on key features to perform subsequent control tasks, in the technical solution of the present application, the oil pipe surface state feature map is further passed through a foreground mask enhancer based on a gated convolution feedforward network to obtain a foreground enhanced feature map of the oil pipe surface state.

[0058] Specifically, since the oil pipe surface state feature map may have scale differences between different layers, which affects the expression of the oil pipe surface state features and dirt recognition, the foreground mask enhancer based on the gated convolution feedforward network first normalizes the oil pipe surface state feature map to eliminate the scale differences between different layers, making the subsequent feature processing more consistent and effective, thereby improving the quality of the feature map. Then, the normalized oil pipe surface state feature map is channel expanded through point convolution to enhance the channel expression ability of the feature map, so that the model can better capture the relationship between different channels. This is particularly important for the recognition and understanding of the complex state of the oil pipe surface, and helps to extract more representative features. At the same time, the introduction of the hole convolution layer can adjust the coverage of the convolution kernel and obtain richer spatial and depth information in the feature map. This deep convolutional encoding can provide more contextual information and help the model understand the surface state of the oil pipe more comprehensively. In the core step, the original deep convolution feature map of the oil pipe surface state is sent to the foreground gated mask module, which dynamically generates foreground mask weights based on functions, and realizes the adaptive selection and enhancement of the foreground features of the oil pipe surface state in the feature map. This step is the key to realizing the saliency of foreground information. Through the gating mechanism, the model can focus on the foreground area of ​​the oil pipe surface state, while suppressing background noise, and improving the recognition ability of lesions. Subsequently, by calculating the positional point multiplication between the deep convolution gated mask weight feature map and the deep convolution backup feature map, the gated mask foreground highlighting of the feature map is realized. This operation effectively combines the foreground mask weights with the original oil pipe surface state feature map to generate a feature map that highlights the foreground information of the oil pipe surface state, providing a clearer basis for subsequent image fusion and analysis. Finally, the foreground highlighting feature map of the gated mask of the oil pipe surface state is channel-contracted through point convolution, completing the final modulation of the feature and generating a foreground-saliency oil pipe surface state feature map, ensuring that the output feature map reaches the best state in terms of expression ability and information content. This process enables the model to perform subsequent processing and analysis with greater efficiency and accuracy.

[0059] In one embodiment of the present application, the oil pipe surface state feature map is passed through a foreground mask enhancer based on a gated convolution feedforward network to obtain an oil pipe surface state foreground enhancement feature map, including: performing layer normalization processing on the oil pipe surface state feature map to obtain a normalized oil pipe surface state feature map; performing channel expansion based on point convolution and deep convolution encoding based on a hole convolution layer on the normalized oil pipe surface state feature map to obtain an oil pipe surface state deep convolution backup feature map and an oil pipe surface state deep convolution original feature map; inputting the oil pipe surface state deep convolution original feature map into a network based on The foreground gated mask module of the function is used to obtain a deep convolution gated mask weight feature map of the oil pipe surface state; the position point multiplication between the deep convolution gated mask weight feature map of the oil pipe surface state and the deep convolution backup feature map of the oil pipe surface state is calculated to obtain a foreground highlighting feature map of the oil pipe surface state gated mask; and the channel contraction based on point convolution is performed on the foreground highlighting feature map of the oil pipe surface state gated mask to obtain the foreground strengthening feature map of the oil pipe surface state.

[0060] Further, in one embodiment of the present application, the oil pipe surface state deep convolution original feature map is input based on The foreground gated mask module of the function is used to obtain a deep convolution gated mask weight feature map of the surface state of the oil pipe, including: inputting the deep convolution original feature map of the surface state of the oil pipe into After activation in the function, mask processing is performed on the feature values ​​of each position in the activated deep convolution original feature map of the oil pipe surface state to obtain the deep convolution gated mask weight feature map of the oil pipe surface state, wherein, in response to the feature value in the activated deep convolution original feature map of the oil pipe surface state being greater than a predetermined hyperparameter, the feature value is set to 1, otherwise it is set to 0.

[0061] Specifically, the oil pipe surface state feature map is processed by the foreground mask enhancer based on the gated convolution feedforward network using the following foreground mask enhancement formula to obtain the oil pipe surface state foreground enhanced feature map; wherein the foreground mask enhancement formula is: ;in, is the characteristic diagram of the surface state of the oil pipe, Indicates the layer normalization operation on the feature map. is the normalized oil pipe surface state characteristic map, is the point convolution operation, The convolution kernel is The dilated convolution operation, It is the deep convolution backup feature map of the oil pipe surface state. It is the original feature map of the deep convolution of the oil pipe surface state. for Activation function, For masking, is the deep convolution gated mask weight feature map of the oil pipe surface state, is the feature value of each position in the feature map, To predetermine the hyperparameters, For point-by-position multiplication, The gated mask foreground highlight feature map for the surface state of the oil pipe. It is a prospect enhancement feature map of the surface state of the oil pipe.

[0062] Then, the time series of the temperature values ​​of the induction heating system is encoded through a temperature time series encoder based on a Bi-LSTM model to extract the time series context association features of the induction heating temperature in the time dimension, thereby obtaining a temperature time series association feature vector.

[0063] It should be understood that since the oil pipe surface state prospect enhancement feature map and the temperature time series association feature vector respectively contain the surface state prospect enhancement features and temperature time series association features related to the oil pipe, there is an implicit correlation and interaction between the two features. However, since the oil pipe surface state features and the temperature time series features are data features of two different modes, in order to capture the interactive implicit correlation information between the two features to help the system understand and control the cleaning task more accurately, in the technical solution of the present application, the oil pipe surface state prospect enhancement feature map and the temperature time series association feature vector are further combined through a heterogeneous Transformer cross-domain joint encoder based on multi-modal assistance to obtain a temperature time series modulated oil pipe surface state optimization expression feature map.

[0064] In one embodiment of the present application, the oil pipe surface state prospect enhancement feature map and the temperature time series associated feature vector are subjected to a multi-modal assisted heterogeneous Transformer cross-domain joint encoder to obtain a temperature time series modulated oil pipe surface state optimized expression feature map, including: performing a linear transformation and a feature shape reshaping on the temperature time series associated feature vector to obtain a modulated temperature time series associated feature matrix; performing a first linear transformation and a second linear transformation on the oil pipe surface state prospect enhancement feature map to obtain a first linear transformation oil pipe surface state prospect enhancement feature map and a second linear transformation oil pipe surface state prospect enhancement feature map; performing feature shape reshaping on the first linear transformation oil pipe surface state prospect enhancement feature map and the second linear transformation oil pipe surface state prospect enhancement feature map to obtain a first linear transformation oil pipe surface state prospect enhancement feature map. The first linear transformation oil pipe surface state foreground enhancement feature matrix and the second linear transformation oil pipe surface state foreground enhancement feature matrix are used as key matrix and value matrix and the modulated temperature time series associated feature matrix is ​​used as query matrix, and the query matrix, the key matrix and the value matrix are input into the Transformer structure to obtain the temperature time series attention query oil pipe surface state feature matrix; the temperature time series attention query oil pipe surface state feature matrix is ​​reshaped and point convolution encoded to obtain the temperature time series attention query oil pipe surface state feature map; the temperature time series attention query oil pipe surface state feature map is multiplied by the position point with the oil pipe surface state foreground enhancement feature map to obtain the temperature time series modulation oil pipe surface state optimized expression feature map.

[0065] Specifically, through the processing of the heterogeneous Transformer cross-domain joint encoder based on multimodal assistance, the effective fusion and optimization of different modal features of the oil pipe surface state semantics and the heating temperature time series semantics can be achieved. The heterogeneous Transformer cross-domain joint encoder based on multimodal assistance first performs linear transformation and feature shape reshaping on the temperature time series associated feature vector to generate a modulated temperature time series associated feature matrix. At the same time, the oil pipe surface state foreground enhancement feature map is linearly transformed twice and the feature shape is reshaped to obtain two sets of linear transformation oil pipe surface state foreground enhancement feature matrices. Then, the two sets of linear transformation oil pipe surface state foreground enhancement feature matrices are used as key matrices and value matrices, and the modulated temperature time series associated feature matrix is ​​used as a query matrix, and input into the Transformer structure. This process uses the self-attention mechanism of the Transformer to realize the dynamic association and fusion of cross-modal features between the oil pipe surface state semantic features and the heating temperature time series semantic features, and capture the complex relationship between the two different modalities. Subsequently, the obtained temperature time series attention query oil pipe surface state feature matrix is ​​reshaped and point convolution encoded to generate a temperature time series attention query oil pipe surface state feature map. Finally, by calculating the positional dot product between the temperature time series attention query oil pipe surface state feature map and the oil pipe surface state foreground enhancement feature map, the temperature time series modulated oil pipe surface state optimized expression feature map is obtained to integrate the information of cross-modal features and original oil pipe surface state features, and generate the final multi-modal joint information optimization feature map.

[0066] In one embodiment of the present application, the temperature timing association feature vector is linearly transformed and reshaped to obtain a modulated temperature timing association feature matrix, including: calculating the matrix multiplication between the temperature timing association feature vector and the vector modulation weight matrix, and then adding the matrix multiplication with the vector modulation bias vector by position to obtain the modulated temperature timing association feature vector; activating the modulated temperature timing association feature vector through a Sigmoid function to obtain a linearly transformed temperature timing association feature vector; and reshaping the linearly transformed temperature timing association feature vector to obtain the modulated temperature timing association feature matrix.

[0067] Further, in one embodiment of the present application, the first linear transformation and the second linear transformation are performed on the oil pipe surface state prospect enhancement feature map to obtain the first linear transformation oil pipe surface state prospect enhancement feature map and the second linear transformation oil pipe surface state prospect enhancement feature map, including: calculating the matrix multiplication between each feature matrix along the channel dimension in the oil pipe surface state prospect enhancement feature map and the first feature map modulation weight matrix, and then adding the matrix multiplication with the first feature map modulation bias matrix in position to obtain the first modulated oil pipe surface state prospect enhancement feature map; activating the first modulated oil pipe surface state prospect enhancement feature map by using a Sigmoid function to obtain the first linear transformation oil pipe surface state prospect enhancement feature map; calculating the matrix multiplication between each feature matrix along the channel dimension in the oil pipe surface state prospect enhancement feature map and the second feature map modulation weight matrix, and then adding the matrix multiplication with the second feature map modulation bias matrix in position to obtain the second modulated oil pipe surface state prospect enhancement feature map; activating the second modulated oil pipe surface state prospect enhancement feature map by using a Sigmoid function to obtain the second linear transformation oil pipe surface state prospect enhancement feature map.

[0068] Furthermore, in one embodiment of the present application, the first linear transformation oil pipe surface state foreground enhancement feature matrix and the second linear transformation oil pipe surface state foreground enhancement feature matrix are used as a key matrix and a value matrix and the modulated temperature time series association feature matrix is ​​used as a query matrix, and the query matrix, the key matrix and the value matrix are input into a Transformer structure to obtain a temperature time series attention query oil pipe surface state feature matrix, including: after calculating the matrix multiplication between the transposed vector of the modulated temperature time series association feature matrix and the first linear transformation oil pipe surface state foreground enhancement feature matrix, the obtained feature matrix is ​​probabilistically activated through a softmax function to obtain a probabilistic oil pipe surface state foreground enhancement weight matrix; and the matrix multiplication between the probabilistic oil pipe surface state foreground enhancement weight matrix and the second linear transformation oil pipe surface state foreground enhancement feature matrix is ​​calculated to obtain the temperature time series attention query oil pipe surface state feature matrix.

[0069] Specifically, the oil pipe surface state prospect enhancement feature map and the temperature time series associated feature vector are processed by the heterogeneous Transformer cross-domain joint encoder based on multimodal assistance using the following cross-domain joint encoding formula to obtain the temperature time series modulated oil pipe surface state optimized expression feature map; wherein the cross-domain joint encoding formula is: ;in, is the prospect enhancement feature map of the oil pipe surface state, is the temperature time series associated feature vector, is the vector modulation weight matrix, is the vector modulation bias vector, is the linear transformation temperature time series correlation feature vector, is the Sigmoid function, For the reshape operation, is the modulation temperature time series correlation characteristic matrix, and are the first feature map modulation weight matrix and the first feature map modulation bias matrix, respectively. is the first linear transformation pipeline surface state prospect enhancement feature map, and are the second feature map modulation weight matrix and the second feature map modulation bias matrix, respectively. is the second linear transformation pipeline surface state prospect enhancement feature map, is the first linear transformation pipeline surface state prospect enhancement feature matrix, is the transposed matrix of the modulation temperature time series correlation characteristic matrix, is the second linear transformation pipeline surface state prospect enhancement feature matrix, for function, is the probabilistic pipeline surface state prospect enhancement weight matrix, is matrix multiplication, Query the surface state feature matrix of the oil pipe for temperature time series attention, is the point convolution operation, Query the characteristic graph of the surface state of the oil pipe for the temperature time series attention, For point-by-position multiplication, The temperature time series modulation oil pipe surface state optimization expression characteristic diagram.

[0070] Then, the temperature time series modulated oil pipe surface state optimization expression characteristic diagram is passed through a power supply voltage frequency controller based on a classifier to obtain a control result, and the control result is used to indicate whether the power supply voltage frequency at the current time point should be increased, decreased, or unchanged. In other words, the optimized expression characteristics of the oil pipe surface state after the heating temperature time series characteristics are modulated are classified and processed, so as to adaptively control the power supply voltage frequency at the current time point. In this way, the voltage frequency of the power supply can be adaptively adjusted according to the cleaning condition of the dirt on the oil pipe surface, so as to adjust the heating temperature of the oil pipe accordingly to perform induction heating on the oil pipe, so as to avoid damage to the metallographic structure and material of the oil pipe body due to excessive temperature. In this way, the oil pipe can be cleaned without affecting the quality of the oil pipe.

[0071] In one embodiment of the present application, based on the temperature-time-modulated oil pipe surface state optimization expression characteristic graph, a control result of the power supply voltage frequency is determined, and the control result is used to indicate whether the power supply voltage frequency at the current time point should increase, decrease, or remain unchanged, including: passing the temperature-time-modulated oil pipe surface state optimization expression characteristic graph through a classifier-based power supply voltage frequency controller to obtain a control result, and the control result is used to indicate whether the power supply voltage frequency at the current time point should increase, decrease, or remain unchanged.

[0072] Accordingly, considering that the oil pipe surface state foreground enhancement feature map and the temperature time series association feature vector respectively express the foreground enhancement features of the surface state image of the cleaned oil pipe based on the image semantic feature spatial distribution convolution gating and the short-range-long-range bidirectional time series association features of the temperature value of the induction heating system, when performing heterogeneous context association cross-domain joint encoding based on multimodal assistance, the obtained temperature time series modulated oil pipe surface state optimization expression feature map will also cause insufficient feature joint encoding coverage due to the structured differences in cross-modal feature distribution under time series, thereby causing the loss of aggregated key information, affecting the expression effect of the temperature time series modulated oil pipe surface state optimization expression feature map.

[0073] Based on this, in the preferred example, when the temperature-series modulation oil pipe surface state optimization expression feature map is input into the classifier-based power supply voltage frequency controller to obtain the control result, the temperature-series modulation oil pipe surface state optimization expression feature map is optimized. Specifically, the optimization process includes the following steps: expanding the temperature-series modulation oil pipe surface state optimization expression feature map into a temperature-series modulation oil pipe surface state optimization expression feature vector; clustering all eigenvalues ​​of the temperature-series modulation oil pipe surface state optimization expression feature vector based on the distance between eigenvalues, such as the difference absolute value distance, and determining the ratio of the number of features of the clustering features to the number of features of the temperature-series modulation oil pipe surface state optimization expression feature vector; arranging the clustering features into a temperature-series modulation oil pipe surface state optimization clustering vector, and dividing the second norm of the temperature-series modulation oil pipe surface state optimization clustering vector by the second norm of the temperature-series modulation oil pipe surface state optimization expression feature vector to obtain a first temperature-series modulation oil pipe surface state optimization clustering related weight value; and calculating the temperature-series modulation oil pipe surface state optimization clustering vector. A first power value of a norm of which is exponentially the scale value is divided by a second power value of a norm of which is exponentially the scale value for the temperature time series modulation of the oil pipe surface state optimization expression feature vector to obtain a second temperature time series modulation of the oil pipe surface state optimization clustering related weight value; for each eigenvalue of the temperature time series modulation of the oil pipe surface state optimization clustering vector, multiply it by the inverse of the difference between the first and second temperature time series modulation of the oil pipe surface state optimization clustering related weight values ​​to obtain the optimized eigenvalue of the temperature time series modulation of the oil pipe surface state optimization clustering vector; for each eigenvalue outside the cluster in the temperature time series modulation of the oil pipe surface state optimization expression feature vector, multiply it by the inverse of the sum of the first and second temperature time series modulation of the oil pipe surface state optimization clustering related weight values ​​to obtain the optimized out-of-class eigenvalue of the temperature time series modulation of the oil pipe surface state optimization expression feature vector; and the optimized eigenvalue of the temperature time series modulation of the oil pipe surface state optimization clustering vector and the optimized out-of-class eigenvalue of the temperature time series modulation of the oil pipe surface state optimization expression feature vector are combined into an optimized temperature time series modulation of the oil pipe surface state optimization expression feature vector.

[0074] Then, the optimized temperature timing modulation oil pipe surface state optimized expression feature vector is input into the classifier-based power supply voltage frequency controller to obtain the control result.

[0075] That is, the optimization process is expressed as: ;in, is the optimized characteristic vector expressing the surface state of the oil pipe modulated by the temperature time series, is the number of eigenvalues ​​of the characteristic vector for optimizing the expression of the surface state of the oil pipe modulated by the temperature time series, is the optimized clustering vector of the temperature time series modulation of the oil pipe surface state, is the number of eigenvalues ​​of the optimized clustering vector of the temperature time series modulation oil pipe surface state, is the ratio value, represents the clustering feature set corresponding to the optimized clustering vector of the temperature time series modulation oil pipe surface state, and Represents the binary norm and the uninorm of the vector respectively. Power, is the characteristic value of each position in the characteristic vector of the optimized expression of the surface state of the oil pipe modulated by the temperature time series, It is the characteristic value of each position in the characteristic vector of the optimized temperature time series modulation oil pipe surface state optimization expression.

[0076] Here, in order to avoid the loss of key suffix semantic information relative to the original feature set as a whole due to aggregation conflict caused by the fact that the temperature time series modulation oil pipe surface state optimization expression feature graph is based on aggregation features, the clustering ratio of the number of features of the temperature time series modulation oil pipe surface state optimization clustering vector relative to the number of features of the temperature time series modulation oil pipe surface state optimization expression feature vector is used as a decision function to perform an adversarial decision on the set absolute representation of the one-norm of the temperature time series modulation oil pipe surface state optimization clustering vector and the temperature time series modulation oil pipe surface state optimization expression feature vector, and respectively compare the temperature time series modulation oil pipe surface state optimization clustering vector and the temperature time series modulation oil pipe surface state optimization expression feature vector with each other. The clustering inner conflict representation of the two norms of the optimized expression feature vector of the surface state of the oil pipe is positively and negatively interacted to construct the solid alignment guardrail of the optimized expression feature vector of the surface state of the oil pipe after the temperature sequence modulation is based on the aggregated features and the original feature set as a whole, so as to achieve the intentional mitigation of the harmful information loss of the optimized expression feature vector of the surface state of the oil pipe after the temperature sequence modulation is based on the aggregated risk transferability, and improve the expression effect of the optimized expression feature vector of the surface state of the oil pipe after the temperature sequence modulation is improved, so as to improve the accuracy of the control result obtained by the power supply voltage frequency controller based on the classifier input of the optimized expression feature map of the surface state of the oil pipe after the temperature sequence modulation is improved. In this way, the voltage frequency of the power supply can be adjusted more accurately according to the cleaning condition of the dirt on the surface of the oil pipe, so as to adjust the heating temperature of the oil pipe accordingly to perform induction heating on the oil pipe, and avoid damage to the metallographic structure and material of the oil pipe body due to excessive temperature. In this way, the oil pipe can be cleaned without affecting the quality of the oil pipe.

[0077] In summary, the above scheme is adopted. During the induction heating stage, the surface state image of the oil pipe during the cleaning process is collected by the camera, and the temperature value of the induction heating system is collected and monitored in real time. Then, the data processing and image analysis algorithm based on artificial intelligence and deep learning is introduced at the back end to characterize the surface state image of the oil pipe and extract the time series characteristics of the heating temperature, so as to learn the semantic representation of the surface state of the oil pipe under the influence of temperature, and thus control the power supply voltage frequency in real time based on the surface state of the oil pipe, so as to adaptively adjust the induction heating temperature. In this way, intelligent heating control of the oil pipe cleaning can be realized, which helps to avoid damage to the metallographic structure and material of the oil pipe body due to excessive temperature. In this way, the oil pipe can be cleaned without affecting the quality of the oil pipe.

[0078] Figure 3 FIG. 1 is a block diagram of an oil pipe cleaning system combining medium frequency heating and mechanical cleaning according to an exemplary embodiment. Figure 3 As shown, the system 200 includes: an oil pipe delivery module 201, which is used to deliver the oil pipe into the cleaning tank.

[0079] The heating and cleaning module 202 is used to heat the oil pipe through an induction heating system and clean the oil pipe through an external brushing device in the cleaning box while feeding the oil pipe, wherein the induction heating system includes: a power supply, an electric heating coil electrically connected to the power supply, and a controller electrically connected to the power supply, and the controller is used to control the voltage frequency of the power supply.

[0080] The purge module 203 is used to purge the oil pipe in the cleaning box by using compressed air.

[0081] The oil pipe exit module 204 is used to drive the oil pipe to exit the cleaning tank.

[0082] Figure 4 FIG. 1 is a schematic diagram of a structure of an oil pipe combining medium frequency heating and mechanical cleaning according to an exemplary embodiment. Figure 4 As shown, there is provided an oil pipe combining medium frequency heating and mechanical cleaning, comprising: an oil pipe 1 and a cleaning box 2; the oil pipe 1 is conveyed into the cleaning box 2; in the cleaning box 2, while the oil pipe 1 is fed, the oil pipe 1 is heated by an induction heating system 3 and the oil pipe 1 is cleaned by an external brushing device, wherein the induction heating system 3 comprises: a power supply 31, an electric heating coil 32 electrically connected to the power supply 31, and a controller 33 electrically connected to the power supply 31, the controller 33 is used to control the voltage frequency of the power supply 31; the oil pipe 1 is purged by compressed air in the cleaning box 2; and the oil pipe 1 is driven to exit the cleaning box 2.

[0083] Reference below Figure 5 , which shows a schematic diagram of the structure of an electronic device 600 suitable for implementing the embodiment of the present application. The terminal device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0084] like Figure 5 As shown, the electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0085] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 5 The electronic device 600 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0086] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of the embodiment of the present application are executed.

[0087] It should be noted that the computer-readable medium mentioned above in the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0088] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0089] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0090] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0091] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0092] The modules involved in the embodiments described in this application can be implemented by software or hardware. The name of the module does not limit the module itself in some cases. For example, the test parameter acquisition module can also be described as a "module for acquiring device test parameters corresponding to the target device".

[0093] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0094] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0095] In the application scenario of the oil pipe cleaning method combining medium frequency heating with mechanical cleaning, first, a surface state image of the cleaned oil pipe captured by a camera is obtained; a time series of temperature values ​​of an induction heating system is obtained; and then, the obtained surface state image and the time series of temperature values ​​are input into a server deployed with an oil pipe cleaning algorithm combining medium frequency heating with mechanical cleaning, wherein the server can process the surface state image and the time series of temperature values ​​based on the oil pipe cleaning algorithm combining medium frequency heating with mechanical cleaning to determine a control result of the power supply voltage frequency, wherein the control result is used to indicate whether the power supply voltage frequency at the current time point should increase, decrease, or remain unchanged.

[0096] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in this application (but not limited to) by each other to form a technical solution.

[0097] In addition, although each operation is described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or to be performed in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the application. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0098] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely exemplary forms of implementation. With respect to the apparatus in the above-mentioned embodiments, the specific manner in which each module performs the operation has been described in detail in the embodiments related to the method, and will not be described in detail here.

Claims

1. A method for cleaning oil pipes by combining medium frequency heating with mechanical cleaning. It is characterized in that include: Transport the oil pipe into the cleaning tank; In the cleaning box, the oil pipe is heated by an induction heating system while being fed and cleaned by an external brushing device, wherein the induction heating system comprises: a power supply, an electric heating coil electrically connected to the power supply, and a controller electrically connected to the power supply, the controller being used to control the voltage frequency of the power supply; Purging the oil pipe by compressed air in the cleaning box; Driving the oil pipe to exit the cleaning box; Wherein, in the cleaning box, while feeding the oil pipe, the oil pipe is heated by an induction heating system and cleaned by an external brushing device, including: Acquire a surface state image of the cleaned oil pipe collected by a camera; Obtaining a time series of temperature values ​​of an induction heating system; Performing feature extraction on the surface state image of the cleaned oil pipe to obtain an oil pipe surface state feature map; The oil pipe surface state feature map is passed through a foreground mask enhancer based on a gated convolutional feedforward network to obtain an oil pipe surface state foreground enhanced feature map; The time series of the temperature values ​​of the induction heating system is passed through a temperature time series encoder based on a Bi-LSTM model to obtain a temperature time series associated feature vector; The oil pipe surface state foreground enhancement feature map and the temperature time series associated feature vector are passed through a heterogeneous Transformer cross-domain joint encoder based on multi-modal assistance to obtain a temperature time series modulated oil pipe surface state optimized expression feature map; Based on the temperature time series modulation oil pipe surface state optimization expression characteristic diagram, determining the control result of the power supply voltage frequency, the control result is used to indicate whether the power supply voltage frequency at the current time point should increase, decrease or remain unchanged; The step of extracting features from the surface state image of the cleaned oil pipe to obtain an oil pipe surface state feature map comprises: passing the surface state image of the cleaned oil pipe through an oil pipe surface state feature extractor based on a hole convolutional neural network model to obtain the oil pipe surface state feature map; The oil pipe surface state feature map is passed through a foreground mask enhancer based on a gated convolution feedforward network to obtain an oil pipe surface state foreground enhanced feature map, including: Performing layer normalization processing on the oil pipe surface state characteristic map to obtain a normalized oil pipe surface state characteristic map; Performing point convolution-based channel expansion and deep convolution coding based on a hole convolution layer on the normalized oil pipe surface state feature map to obtain an oil pipe surface state deep convolution backup feature map and an oil pipe surface state deep convolution original feature map; The deep convolution original feature map of the oil pipe surface state is input based on The foreground gated mask module of the function is used to obtain the deep convolution gated mask weight feature map of the surface state of the oil pipe; Calculate the position point multiplication between the oil pipe surface state deep convolution gated mask weight feature map and the oil pipe surface state deep convolution backup feature map to obtain the oil pipe surface state gated mask foreground highlight feature map; Performing point convolution-based channel contraction on the oil pipe surface state gated mask foreground highlighting feature map to obtain the oil pipe surface state foreground strengthening feature map; The oil pipe surface state deep convolution original feature map is input based on The foreground gated mask module of the function is used to obtain a deep convolution gated mask weight feature map of the surface state of the oil pipe, including: inputting the deep convolution original feature map of the surface state of the oil pipe into After activation in the function, mask processing is performed on the feature values ​​of each position in the activated deep convolution original feature map of the oil pipe surface state to obtain the deep convolution gated mask weight feature map of the oil pipe surface state, wherein, in response to the feature value in the activated deep convolution original feature map of the oil pipe surface state being greater than a predetermined hyperparameter, the feature value is set to 1, otherwise it is set to 0.

2. The oil pipe cleaning method combining medium frequency heating and mechanical cleaning according to claim 1, It is characterized in that The oil pipe surface state prospect enhancement feature map and the temperature time series associated feature vector are passed through a heterogeneous Transformer cross-domain joint encoder based on multi-modal assistance to obtain a temperature time series modulated oil pipe surface state optimized expression feature map, including: Performing linear transformation and characteristic shape reshaping on the temperature time series associated characteristic vector to obtain a modulated temperature time series associated characteristic matrix; Performing a first linear transformation and a second linear transformation on the oil pipe surface state prospect enhancement feature map to obtain a first linear transformation oil pipe surface state prospect enhancement feature map and a second linear transformation oil pipe surface state prospect enhancement feature map; Reshaping the first linear transformation tubing surface state prospect enhancement feature map and the second linear transformation tubing surface state prospect enhancement feature map to obtain a first linear transformation tubing surface state prospect enhancement feature matrix and a second linear transformation tubing surface state prospect enhancement feature matrix; Taking the first linear transformation oil pipe surface state prospect enhancement feature matrix and the second linear transformation oil pipe surface state prospect enhancement feature matrix as a key matrix and a value matrix and taking the modulated temperature time series association feature matrix as a query matrix, inputting the query matrix, the key matrix and the value matrix into a Transformer structure to obtain a temperature time series attention query oil pipe surface state feature matrix; Reshaping and point convolution encoding are performed on the temperature time series attention query oil pipe surface state feature matrix to obtain a temperature time series attention query oil pipe surface state feature map; The temperature time series attention query oil pipe surface state characteristic map and the oil pipe surface state foreground enhancement characteristic map are multiplied by position points to obtain the temperature time series modulation oil pipe surface state optimization expression characteristic map.

3. The oil pipe cleaning method combining medium frequency heating and mechanical cleaning according to claim 2, It is characterized in that The temperature time series correlation feature vector is linearly transformed and the feature shape is reshaped to obtain a modulated temperature time series correlation feature matrix, including: Calculating the matrix multiplication between the temperature time series associated characteristic vector and the vector modulation weight matrix and then adding the result to the vector modulation bias vector according to the position to obtain the modulated temperature time series associated characteristic vector; Activating the modulation temperature time series associated feature vector by a Sigmoid function to obtain a linearly transformed temperature time series associated feature vector; The linear transformation temperature time series correlation feature vector is reshaped to obtain the modulated temperature time series correlation feature matrix.

4. The oil pipe cleaning method combining medium frequency heating and mechanical cleaning according to claim 3, It is characterized in that Performing a first linear transformation and a second linear transformation on the oil pipe surface state prospect enhancement feature map to obtain a first linear transformation oil pipe surface state prospect enhancement feature map and a second linear transformation oil pipe surface state prospect enhancement feature map, including: Calculating matrix multiplication between each feature matrix along the channel dimension in the oil pipe surface state prospect enhancement feature map and the first feature map modulation weight matrix, and then adding the matrix matrix to the first feature map modulation bias matrix according to position to obtain a first modulated oil pipe surface state prospect enhancement feature map; Activating the first modulated oil pipe surface state foreground enhancement feature map by using a Sigmoid function to obtain the first linearly transformed oil pipe surface state foreground enhancement feature map; Calculating matrix multiplication between each feature matrix along the channel dimension in the oil pipe surface state prospect enhancement feature map and the second feature map modulation weight matrix, and then adding the matrix matrix to the second feature map modulation bias matrix according to position to obtain a second modulated oil pipe surface state prospect enhancement feature map; The second modulated oil pipe surface state foreground enhancement feature map is activated by a Sigmoid function to obtain the second linearly transformed oil pipe surface state foreground enhancement feature map.

5. The oil pipe cleaning method combining medium frequency heating and mechanical cleaning according to claim 4, It is characterized in that The first linear transformation oil pipe surface state prospect enhancement feature matrix and the second linear transformation oil pipe surface state prospect enhancement feature matrix are used as a key matrix and a value matrix and the modulated temperature time series association feature matrix is ​​used as a query matrix, and the query matrix, the key matrix and the value matrix are input into a Transformer structure to obtain a temperature time series attention query oil pipe surface state feature matrix, including: After calculating the matrix multiplication between the transposed vector of the modulation temperature time series correlation feature matrix and the first linear transformation oil pipe surface state prospect enhancement feature matrix, the obtained feature matrix is ​​probabilistically activated through a softmax function to obtain a probabilistic oil pipe surface state prospect enhancement weight matrix; Calculate the matrix multiplication between the probabilistic oil pipe surface state prospect enhancement weight matrix and the second linear transformation oil pipe surface state prospect enhancement feature matrix to obtain the temperature time series attention query oil pipe surface state feature matrix.

6. The oil pipe cleaning method combining medium frequency heating and mechanical cleaning according to claim 5, It is characterized in that Based on the temperature-series timing modulation oil pipe surface state optimization expression characteristic graph, determine the control result of the power supply voltage frequency, the control result is used to indicate whether the power supply voltage frequency at the current time point should increase, decrease or remain unchanged, including: passing the temperature-series timing modulation oil pipe surface state optimization expression characteristic graph through a classifier-based power supply voltage frequency controller to obtain a control result, the control result is used to indicate whether the power supply voltage frequency at the current time point should increase, decrease or remain unchanged.

Citation Information

Patent Citations

  • Method for enhancing defect features of pressure vessel based on eddy current pulse thermal imaging technology

    CN108682002A

  • Scale cleaning device for electromagnetic flowmeter

    CN117531786A