Intelligent ball throwing and receiving device and control system thereof

Abstract

The present application discloses an intelligent ball throwing and receiving device and its control system, which relates to the technical field of crude oil pipe wall cleaning. The intelligent ball throwing and receiving device control system includes: an acquisition module, which is used to obtain the operation data of the target pipeline through a sensor, wherein the operation data includes: temperature data, pressure data, flow data, and vibration data; a processing module, which is used to predict the wax deposition of the target pipeline based on the operation data of the target pipeline and generate control instructions using a preset machine learning model; wherein the preset machine learning model is constructed based on the historical operation data and historical wax removal data of the target pipeline; the preset machine learning model is used to generate pipeline wax deposition prediction information and ball throwing and receiving parameter information based on the input target features. The intelligent ball throwing and receiving device throws and receives the ball based on the control instructions to achieve the cleaning of the target pipeline.

Description

Technical Field

[0001] The present application relates to the technical field of crude oil pipe wall cleaning, and in particular to an intelligent ball throwing and receiving device and a control system thereof. Background Art

[0002] Waxing refers to the phenomenon that wax components in oil products are separated and deposited on the inner wall of the pipeline due to changes in temperature, pressure and other conditions, forming a layer of wax accumulation. This accumulation will cause the inner diameter of the pipeline to become smaller, increase the resistance to fluid flow, and may even cause pipeline blockage, seriously affecting the normal transportation of crude oil. Against this background, wax ball technology came into being. Wax balls can effectively remove wax accumulations on the inner wall of the pipeline through physical friction and impact in the pipeline. However, the traditional wax ball placement and recovery process mainly relies on manual experience and simple mechanical devices, lacking intelligent and automated means, resulting in low wax removal efficiency and difficulty in adapting to complex pipeline operating conditions.

[0003] For example, the Chinese patent with the authorization announcement number CN111974772B discloses a screw-pushing type wax-removing ball automatic delivery device, system and method. It includes: a motor, a connecting piece, a ball-dropping screw, a ball storage barrel, a wax-removing ball and a gate valve; the gate valve, the ball storage barrel, the connecting piece and the motor are connected in sequence from bottom to top; one end of the ball-dropping screw is connected to the output end of the motor, the ball-dropping screw is placed in the ball storage barrel, and the ball-dropping screw is threadedly connected with multiple wax-removing balls in sequence from top to bottom. The screw-pushing type wax-removing ball automatic delivery device, the crude oil pipeline connecting pipe and the equipment support are connected in sequence from top to bottom to form a screw-pushing type wax-removing ball automatic delivery system. This application can efficiently and reliably deliver wax-removing balls automatically, and can also effectively reduce equipment costs by reducing transmission links, which is of great significance for realizing unmanned stations in oil fields.

[0004] The above method has the problems raised by the background technology. The existing technology can realize the automatic placement of wax cleaning balls, but it is unable to make intelligent judgments and predict automatic ball placement parameters according to the wax deposition conditions in the pipeline. This makes the wax removal effect of a single ball placement parameter greatly reduced when facing various wax deposition conditions in the pipeline. Summary of the invention

[0005] In response to the above problems, the present application discloses an intelligent ball throwing and receiving device and its control system, which aims to optimize the data collection and analysis process, combine more sophisticated feature engineering and machine learning models, detect pipeline operation data, and dynamically adjust the analysis threshold based on the pipeline operation data to achieve intelligent cleaning of the pipeline system.

[0006] In a first aspect, the present application provides a control system for an intelligent ball throwing and receiving device, comprising:

[0007] A collection module, used to obtain the operation data of the target pipeline through a sensor, wherein the operation data includes: temperature data, pressure data, flow data, wax deposition data, and vibration data;

[0008] A processing module, configured to predict wax deposition on the target pipeline based on the operating data of the target pipeline and to generate a control instruction by using a preset machine learning model;

[0009] The preset machine learning model is constructed based on the historical operation data and historical wax removal data of the target pipeline; the preset machine learning model is used to generate pipeline wax deposition prediction information and ball casting and receiving parameter information based on the input target features;

[0010] The control instructions are used to control the intelligent pitching and receiving device to perform pitching and receiving based on the pitching and receiving parameter information.

[0011] As an optional implementation, based on the operation data of the target pipeline, using a preset machine learning model, performing pipeline wax deposition prediction on the target pipeline and generating control instructions include:

[0012] In response to the vibration data reaching a vibration data analysis threshold, feature extraction is performed on the vibration data to generate a vibration feature.

[0013] As an optional implementation, the vibration data analysis threshold is set based on the wax deposition data and the vibration data in the historical operation data of the target pipeline;

[0014] The vibration characteristics include: vibration time domain characteristics, vibration frequency domain characteristics, and vibration time and frequency domain characteristics;

[0015] Extracting features from the vibration data to generate vibration features includes: extracting statistical features such as peak value, mean value, standard deviation and skewness of the vibration data to generate the vibration time domain features;

[0016] Performing Fourier transform on the vibration data to generate the vibration frequency domain features;

[0017] The vibration data is subjected to time-frequency analysis using wavelet transform to generate the vibration time-frequency domain features.

[0018] As an optional implementation, the wax deposition data is used to characterize the wax layer thickness at each location in the target pipeline;

[0019] Marking a position in the target pipeline where the historical average value of the wax layer thickness is greater than the target thickness average value as a first wax deposition area;

[0020] The acquisition module further includes: a placement unit;

[0021] The arrangement unit is used to arrange the sensor at a target section position of the target pipeline, wherein the target section position includes: the first wax deposition area of ​​the target pipeline, the starting point of the target pipeline, and the bend of the target pipeline;

[0022] The acquisition module is also used to set the vibration data analysis threshold for each target section position based on the historical operation data collected by the sensor arranged at each target section position.

[0023] As an optional implementation, it also includes:

[0024] In response to the temperature data being greater than a preset reference temperature, increasing the vibration data analysis threshold based on a preset temperature influence coefficient;

[0025] In response to the pressure data being greater than a preset reference pressure, increasing the vibration data analysis threshold based on a preset pressure influence coefficient;

[0026] The preset temperature influence coefficient and the preset pressure influence coefficient are determined through linear regression analysis based on historical operation data of the target pipeline.

[0027] As an optional implementation, it also includes:

[0028] In response to the temperature data being greater than a preset reference temperature and the pressure data being greater than a preset reference pressure, the vibration data analysis threshold is increased by a weighted fusion method based on the preset temperature influence coefficient and the preset pressure influence coefficient.

[0029] As an optional implementation, it also includes:

[0030] In response to the flow data reaching a first change threshold within a first time period, extracting instantaneous frequency characteristics and energy characteristics of the vibration data within the first time period;

[0031] In response to the pressure data reaching a second change threshold within a second time period, time-frequency features of the vibration data within the second time period are extracted.

[0032] As an optional implementation, the historical wax cleaning data includes: wax cleaning ball data, ball throwing data;

[0033] Based on the wax cleaning ball data, the pitching data, the pressure data, and the flow data, wax cleaning effect characteristics are generated, and the wax cleaning effect characteristics include: flow recovery rate characteristics and pressure drop speed characteristics.

[0034] As an optional implementation, the acquisition module further includes: a synchronization unit; the synchronization unit is used to synchronize the operation data acquired by the sensors arranged at different positions of the target pipeline to the same timestamp.

[0035] In a second aspect, the present application also provides an intelligent ball throwing and receiving device, comprising:

[0036] A ball throwing unit is provided with a motor for adjusting and executing a ball throwing operation, wherein the ball throwing operation comprises: throwing a wax cleaning ball into the target pipe based on the ball throwing and receiving parameter information;

[0037] A ball collecting unit, equipped with a mechanical capture device, for collecting the dropped wax cleaning balls from the target pipeline;

[0038] A control unit, configured to receive a control instruction from the processing module and control the pitching unit and the receiving unit to operate based on the control instruction;

[0039] Wherein, the control unit adjusts the pitching frequency, ball size, pitching material and pitching intensity of the pitching unit based on the control instruction.

[0040] Compared with the prior art, the beneficial effects of the present application are: the control system can perform real-time analysis based on the operating data of the target pipeline, use machine learning models to accurately predict the wax deposition in the pipeline, and generate corresponding ball throwing and receiving strategies for use by the intelligent ball throwing and receiving device. This makes the wax removal operation more accurate and efficient, and reduces the impact of wax accumulation in the pipeline on crude oil transportation. The system can dynamically adjust the analysis threshold to adapt to changes in temperature, pressure and other conditions in the pipeline, ensuring efficient detection of abnormal conditions in complex and changeable operating environments. Through this dynamic adjustment mechanism, the system can maintain sensitivity to the pipeline status under different working conditions, thereby improving the reliability of the wax removal operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic diagram of the structure of a control system for an intelligent ball throwing and receiving device provided in an embodiment of the present application;

[0042] Figure 2 A schematic diagram of the structure of a collection module provided in an embodiment of the present application;

[0043] Figure 3 A schematic diagram of the structure of an intelligent ball throwing and receiving device provided in an embodiment of the present application;

[0044] Figure 4 A schematic diagram of the structure of another intelligent ball throwing and receiving device provided in an embodiment of the present application.

[0045] The corresponding relationship between the illustrations and component names in the figure is as follows:

[0046] 10. Collection module; 20. Processing module; 11. Arrangement unit; 12. Synchronization unit; 200. Intelligent ball throwing and receiving device; 210. Ball throwing unit; 220. Ball receiving unit; 230. Control unit; 211. Wax cleaning ball storage bin. DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments.

[0048] See also Figure 1 , Figure 1 A schematic diagram of a control system for an intelligent ball throwing and receiving device provided in an embodiment of the present application includes: a collection module 10 and a processing module 20;

[0049] The acquisition module 10 is used to obtain the operation data of the target pipeline through a sensor, wherein the operation data includes: temperature data, pressure data, flow data, wax deposition data, and vibration data;

[0050] The processing module 20 is used to predict the wax deposition of the target pipeline based on the operation data of the target pipeline and generate a control instruction by using a preset machine learning model;

[0051] The preset machine learning model is constructed based on the historical operation data and historical wax removal data of the target pipeline; the preset machine learning model is used to generate pipeline wax deposition prediction information and ball casting and receiving parameter information based on the input target features;

[0052] The control instructions are used to control the intelligent pitching and receiving device to perform pitching and receiving based on the pitching and receiving parameter information.

[0053] The present application obtains the operation data of the target pipeline through the acquisition module 10, and these operation data are obtained through sensors arranged in the target pipeline. The processing module 20 uses a preset machine learning model to predict the wax deposition in the target pipeline based on the operation data of the target pipeline obtained by the acquisition module 10, and generates corresponding control instructions to control the intelligent ball throwing and receiving device to throw pipeline wax cleaning balls (hereinafter referred to as "wax cleaning balls") to clean the wax in the target pipeline according to the target ball throwing strategy. Among them, the preset machine learning model is constructed based on the historical operation data and historical wax cleaning data of the target pipeline.

[0054] In a specific implementation, the collection of historical operation data and historical wax cleaning data can be carried out in the data collection stage before the implementation of this application.

[0055] For the above acquisition module 10:

[0056] In a specific implementation, the acquisition module 10 includes various types of sensors, which are arranged at target positions of the target pipeline and are used to monitor the operating status of the pipeline in real time.

[0057] See also Figure 2 , Figure 2 A schematic diagram of the structure of a collection module provided in an embodiment of the present application;

[0058] As an optional implementation, the acquisition module 10 further includes: an arrangement unit 11;

[0059] The arrangement unit 11 is used to arrange the sensor at a target section position of the target pipeline, wherein the target section position includes: a first wax deposition area of ​​the target pipeline, a starting point of the target pipeline, and a bend of the target pipeline;

[0060] For example, the sensors may be distributed at the starting point, the middle section, the bends of the target pipeline, and the wax deposition-prone areas determined based on the historical operation data of the pipeline.

[0061] It is understandable that, due to the different actual structures of different pipelines, the sensor arrangement schemes of different target pipelines may be different, but in order to effectively implement the present disclosure, the arrangement of sensors can be as described above, with emphasis on the starting point, middle section, bend, and wax-prone areas of the target pipeline. The wax-prone areas can be determined based on the historical wax deposition conditions of the target pipeline.

[0062] In a specific implementation, the position where the historical average value of the wax layer thickness is greater than the target thickness average value can be marked as the first wax deposition area, which is a wax deposition-prone area.

[0063] It is understandable that, since the operating objectives of different pipelines are different, different pipelines have different tolerances for wax layer thickness during operation, so the target thickness mean can be set based on actual needs.

[0064] In the embodiment of the present disclosure, the data collected by the sensor includes: temperature data, pressure data, flow data, wax deposition data, and vibration data.

[0065] Among them, temperature is an important factor affecting wax precipitation and wax formation. Temperature data can help determine whether the fluid in the pipeline is in a temperature range that is prone to wax formation. When the temperature drops to a certain critical point, wax may precipitate from the fluid and deposit on the pipe wall. By monitoring the temperature in real time, the system can predict the possibility of wax formation in advance and take cleaning measures in time.

[0066] Pressure data reflects the flow state of the fluid in the pipeline. By monitoring the pressure distribution and fluctuations, abnormalities in fluid flow, such as a drop in flow rate, blockage, or waxing, can be identified. A sudden drop in pressure may indicate a partial blockage in the pipeline, preventing the fluid from passing normally.

[0067] Flow data provides direct information about the efficiency of fluid flow through the pipeline. By monitoring the difference in flow at the inlet and outlet of the pipeline, the system can evaluate whether the fluid is flowing smoothly in the pipeline. If the flow rate is significantly reduced, this may indicate that wax deposition or blockage has occurred in the pipeline, affecting the normal flow of the fluid.

[0068] Vibration data reflects the interaction between the fluid and the pipe wall in the pipe, especially when flow rate changes, pressure fluctuations or wax deposition occurs. These interactions are usually reflected in the vibration data. Vibration data can help the system identify abnormal dynamic behavior, such as resonance, flow fluctuations or increased resistance caused by wax deposition.

[0069] By analyzing the frequency and energy characteristics of the vibration data, it is possible to locate where wax deposition may occur. This provides guidance for the subsequent operation of the ball delivery device to ensure that the cleaning ball is delivered to the required position.

[0070] In specific implementations, temperature sensors can be arranged at different locations on the inner and outer walls of the target pipeline to monitor the temperature changes of the fluid inside the pipeline and the influence of the external ambient temperature. Pressure sensors can be arranged at the inlet and outlet of the pipeline and other target nodes mentioned above to monitor the pressure distribution and fluctuation of the fluid in the pipeline. Flow sensors should be installed at the inlet and outlet of the pipeline to monitor the fluid flow in the pipeline in real time. Vibration sensors should be arranged at key parts of the pipeline, especially areas prone to wax deposition or blockage, to capture vibration data generated by the interaction between the fluid in the pipeline and the pipe wall.

[0071] As an optional implementation, the acquisition module 10 further includes a synchronization unit 12, and the synchronization unit 12 is used to synchronize the operation data acquired by sensors arranged at different positions of the target pipeline to the same timestamp.

[0072] In this way, by collecting various types of operating data of the target pipeline, the operating status of the target pipeline can be analyzed and judged more comprehensively, providing a basis for the subsequent prediction of pipeline wax deposition and the formulation of wax removal ball throwing strategies.

[0073] For the above processing module 20:

[0074] The embodiments of the present disclosure are not limited to a specific machine learning model, for example, decision trees, random forests, support vector machines, and neural networks can all be used.

[0075] In the specific implementation, feature extraction is performed on the operating data of the target pipeline to generate target features; the target features are input into the preset machine learning model to generate pipeline wax deposition prediction information and ball throwing and receiving parameter information. Although temperature, pressure and other data can be monitored in the operating data, and alarms are issued when such data are abnormal, this delayed response method has many disadvantages for pipeline wax deposition cleaning. For example, responding to abnormalities may have a greater impact on the working efficiency of the pipeline and subsequent wax removal work. After wax deposition to a certain extent, cleaning requires more resources, and the wax removal effect cannot be guaranteed. Only tentative throwing of wax removal balls can be performed to continuously optimize.

[0076] The present disclosure extracts target features from operating data to capture more meaningful and representative information rather than directly processing raw data. These features can often better describe the state of the pipeline and the possibility of wax deposition. For example, the frequency and energy features extracted from vibration data and the rate of change extracted from flow data can provide more detailed inputs for the model, enabling the model to more accurately predict wax deposition.

[0077] Generating ball throwing and receiving parameter information at the same time as generating pipeline wax deposition prediction information means that the system not only stays at the prediction level, but can also directly generate corresponding cleaning strategies based on the prediction results. This process allows the system to automatically generate the optimal ball throwing parameters, including ball throwing frequency, force, type and size of cleaning balls, etc., according to different wax deposition levels, locations and dynamic behaviors. In this way, the system can immediately take targeted actions to improve cleaning efficiency and reduce the time and cost of pipeline maintenance.

[0078] As an optional implementation, performing pipeline wax deposition prediction on a target pipeline includes:

[0079] In response to vibration data in the operation data reaching a vibration data threshold, feature extraction is performed on the vibration data to generate a vibration feature.

[0080] Among them, changes in vibration data are often a direct reflection of the interaction between the fluid in the pipeline and the pipe wall. When wax deposition occurs in the pipeline, fluid flow is hindered, resulting in abnormal changes in the frequency or amplitude of the vibration data. By setting a vibration data threshold and performing feature extraction when the threshold is exceeded, the system can capture abnormal dynamic behaviors related to wax deposition, thereby improving the accuracy of the prediction.

[0081] In the present disclosure, a threshold is set instead of directly performing feature extraction on all vibration data because vibration data is usually collected at a high frequency and the amount of data is very large. If feature extraction and analysis are performed on all vibration data, the consumption of computing resources will be very high. The flow of fluid in the pipeline will cause normal vibration fluctuations, which do not represent a problem in most cases. If the threshold is not set, the system may be disturbed by these normal fluctuations, resulting in misjudgment or increased false alarm rate. Threshold setting can filter out these normal vibration data fluctuations and only conduct in-depth analysis of signals that may represent abnormalities. By setting a threshold, feature extraction is only triggered when the vibration data exceeds a certain range, which can significantly reduce the amount of calculation and improve the response speed and real-time performance of the system.

[0082] In a specific implementation, historical operation data can be used to set one or more vibration data analysis thresholds. These thresholds can be determined by analyzing the vibration characteristics of wax deposition events in past data to ensure that they can effectively distinguish between normal operation and potential wax deposition.

[0083] When the vibration data reaches or exceeds the threshold, time domain features can be extracted, such as the peak value, mean, standard deviation, skewness, etc. of the vibration signal. These features can help identify the overall change trend of the vibration data and preliminarily determine whether there is an anomaly. By converting the vibration data from the time domain to the frequency domain through Fourier transform, the main frequency components can be extracted and the intensity and distribution of these frequencies can be analyzed. If certain frequency components are abnormally enhanced or weakened, it may indicate the presence of wax deposition or other obstructions in the pipeline. Using time-frequency analysis methods such as wavelet transform, the joint changes of time and frequency in the data can be captured, especially transient features. This method is very effective in identifying short-lived but significant vibration anomalies, such as flow changes caused by sudden wax deposition.

[0084] In specific implementations, the extracted time domain, frequency domain, and time-frequency domain features can be combined to form a multidimensional feature vector. This fused feature set can more comprehensively describe the changes in vibration data and improve the sensitivity and accuracy of the model to wax deposition.

[0085] Exemplarily, by analyzing the historical operation data of the target pipeline, especially the vibration signal characteristics when the wax deposition event occurs, one or more threshold values ​​of vibration data are set. For example, based on historical data, it can be found that when wax deposition occurs in the pipeline, the amplitude and frequency of the vibration data will change abnormally. By statistically analyzing the peak value, frequency and other related parameters of the vibration data corresponding to the wax deposition event, a threshold range of vibration data can be set.

[0086] The wax deposition event can be determined by judging whether the wax layer thickness in the target pipeline is greater than the target thickness value.

[0087] For example, under normal operating conditions, the peak value of the vibration signal is usually between 0.1g and 0.3g (gravitational acceleration), but when a waxing event occurs, the peak value may rise to above 0.5g. Therefore, a threshold of 0.4g can be set. When the peak value of the vibration signal monitored in real time exceeds this threshold, the system considers that there may be a waxing risk and triggers further feature extraction and analysis. Among them, g is the unit of gravity acceleration, which is one of the commonly used units for measuring vibration amplitude or acceleration. 1g is approximately 9.81m / s 2 .

[0088] When the vibration signal reaches or exceeds the set threshold, the relevant time domain, frequency domain and time-frequency domain features are immediately extracted.

[0089] Extract statistical features of the vibration signal such as peak value, mean, standard deviation and skewness. For example, in a wax deposition event, the peak value of the vibration signal increases from the normal 0.2g to 0.6g, and the standard deviation increases from 0.05g to 0.15g. These changes may indicate that the fluid flow in the pipeline is hindered, possibly due to wax deposition.

[0090] Perform Fourier transform on the vibration signal to extract the main frequency components. For example, in normal operation, the main frequency of the vibration signal is concentrated below 10 Hz, but when wax deposition occurs, high-frequency components (such as frequencies above 20 Hz) may be abnormally enhanced. This change in frequency components can indicate turbulence or blockage of fluid flow.

[0091] Use wavelet transform to perform time-frequency analysis on vibration signals and capture transient changes. Wavelet transform can identify frequency changes in vibration signals at specific time points. For example, at the moment of sudden flow changes or pressure fluctuations, the vibration signal may show a short-term enhancement of high frequency. This transient feature is a potential indicator of wax deposition.

[0092] The extracted time domain, frequency domain and time-frequency domain features are fused to construct a multi-dimensional feature vector. For example, for a vibration signal within a certain time period, the feature vector may include:

[0093] Time domain characteristics: peak (0.6g), standard deviation (0.15g), skewness (2.0), etc.

[0094] Frequency domain characteristics: main frequency (15Hz), high-frequency component intensity (the intensity of 20Hz is 0.3).

[0095] Time-frequency domain characteristics: transient high-frequency enhancement at a specific time point (such as t=30 seconds) (the transient signal intensity above 30Hz is 0.25g).

[0096] Please note that the above data are only exemplary and do not constitute any form of limitation.

[0097] By inputting the multi-dimensional feature vector into the preset machine learning model, these features can be comprehensively analyzed to predict whether there is wax deposition in the pipeline and the severity of wax deposition. For example, by analyzing the abnormal high-frequency enhancement of the vibration signal, the significant increase in the standard deviation, and the transient high-frequency phenomenon, the model predicts that the probability of wax deposition in the current pipeline is 80%, and recommends the immediate placement of cleaning balls.

[0098] In this way, by setting the threshold of the vibration signal and extracting the features in the time domain, frequency domain, and time-frequency domain when the threshold is exceeded, the possible wax deposition in the pipeline can be effectively identified. Combined with the construction of multidimensional feature vectors and the application of machine learning models, the system can accurately predict the risk of wax deposition and take cleaning measures in time to ensure the normal operation of the pipeline. This implementation method not only improves the accuracy of detection, but also optimizes the system's response speed and resource utilization efficiency.

[0099] As an optional implementation, the acquisition module 10 is further used to set a vibration data analysis threshold for each section position based on the historical operation data collected by the sensor arranged at each section position.

[0100] In practice, sensor data from different sections of the pipeline may be different, and different locations in the pipeline require different vibration data analysis thresholds. This difference stems from the changes in pipeline structure, materials, fluid behavior, and environmental conditions at different locations.

[0101] For example, pipelines may have different geometric shapes at different sections (such as straight sections, curved sections, joints, etc.), and these structural differences will affect the vibration characteristics. For example, curved pipeline sections may produce more complex vibration patterns than straight sections, and are more likely to accumulate impurities or wax, thus causing different vibration signals.

[0102] In some cases, different parts of the pipeline may be made of different materials or undergo different treatment processes. These differences in material properties will result in different thermal expansion coefficients and elastic moduli, which will affect the propagation and response of vibration signals.

[0103] The pipeline may pass through different environmental areas (such as underground, above ground, different temperature or pressure conditions), and these external environmental conditions will also affect the vibration behavior of the pipeline. For example, the pipeline section buried underground may show different vibration characteristics from the pipeline section exposed to the air due to different temperatures and pressures.

[0104] Due to these differences in local conditions, the sensor at each location may record different vibration signal patterns. This means that a uniform vibration signal threshold may not be applicable in all locations. To improve the accuracy of detection, a vibration data analysis threshold suitable for each location can be set based on the specific conditions of the location.

[0105] In a specific implementation, the analysis threshold of the vibration signal can be calculated and set separately for the historical operation data of each sensor location. For example, sensors in certain curved sections may require a higher threshold to filter out complex vibrations in normal operation, while sensors in straight sections may require a lower threshold to detect minor abnormal changes.

[0106] In addition, these location-specific thresholds can be dynamically adjusted based on real-time temperature and pressure data. By combining the physical model and historical data, the system can set an initial threshold for each sensor and dynamically adjust it based on real-time monitored conditions during actual operation.

[0107] In this way, by setting location-specific thresholds, the system can more accurately monitor the pipeline status and improve the accuracy and response speed of wax deposition detection. This approach ensures that the system can still operate stably and reliably in the face of complex and changing pipeline environments.

[0108] As an optional implementation, in response to temperature data being greater than a preset reference temperature, the vibration data analysis threshold is increased based on a preset temperature influence coefficient; in response to pressure data being greater than a preset reference pressure, the vibration data analysis threshold is increased based on a preset pressure influence coefficient; wherein the temperature influence coefficient and the preset pressure influence coefficient are determined by linear regression analysis based on historical operating data of the target pipeline.

[0109] In specific implementations, the vibration data analysis threshold can be dynamically adjusted based on the temperature data and pressure data of the target pipeline. By responding to changes in temperature and pressure data and adjusting the analysis threshold of the vibration signal accordingly, the system can better adapt to different operating conditions, reduce false alarms and missed alarms, and ensure that wax deposition can still be accurately identified and predicted under complex working conditions.

[0110] Temperature changes can affect the physical properties of pipe materials, such as thermal expansion or contraction, and these changes can directly affect the propagation characteristics of vibration signals. For example, increased temperature can cause pipe materials to become softer, thereby changing the frequency and amplitude of vibration. If the threshold of the vibration signal is not adjusted, the system may mistake normal changes caused by temperature as abnormal.

[0111] Similarly, pressure changes can also affect the fluid dynamics and vibration characteristics inside the pipe. For example, an increase in pressure may cause the fluid flow rate to increase, which in turn causes a change in vibration frequency. If the system maintains a fixed vibration signal threshold, it may not be able to accurately distinguish between normal vibration fluctuations caused by pressure changes and actual abnormal conditions (such as blockage caused by wax deposition).

[0112] The operating environment of pipelines is complex and changeable, and changes in temperature and pressure are common phenomena. Fixed vibration signal thresholds may not be applicable under different operating conditions. By responding to changes in temperature and pressure data and dynamically adjusting the threshold of vibration signals, the system can better adapt to different operating conditions and maintain sensitivity to abnormal situations.

[0113] During the operation of the pipeline, the acquisition module 10 continuously monitors the temperature, pressure and vibration data of the target pipeline.

[0114] Whenever the system detects that the temperature or pressure data exceeds the preset reference value, it automatically triggers the adjustment of the vibration data threshold. The adjustment method is as follows:

[0115] If the real-time temperature Greater than the reference temperature , the system is based on the preset temperature influence coefficient Dynamically increase the threshold of vibration signals:

[0116]

[0117] in, is the vibration data analysis threshold at time t after dynamic adjustment, is the initial vibration data analysis threshold, is the temperature at time t;

[0118] If the real-time pressure Greater than the base pressure , the system is based on the preset pressure influence coefficient Dynamically increase vibration data analysis thresholds:

[0119]

[0120] If changes in temperature and pressure are detected at the same time, the system will combine the effects of both and adjust the threshold by:

[0121]

[0122] Under the dynamically adjusted threshold, the system continues to monitor vibration data. When the vibration data exceeds the newly set threshold, the system extracts relevant features (such as time domain, frequency domain, and time-frequency domain features) and predicts wax deposition through a preset machine learning model.

[0123] In a specific implementation, the reference pressure and reference temperature can be set based on historical operation data, for example, by collecting and analyzing historical operation data of the pipeline, especially the changes in temperature and pressure under different operating conditions. By analyzing these data, the temperature and pressure range under normal operation of the pipeline can be identified, as well as the temperature and pressure change characteristics when wax deposition occurs.

[0124] Through statistical analysis, the mean value, standard deviation and other statistical characteristics of the temperature and pressure of the pipeline under normal operating conditions are determined. These statistical characteristics can help identify the common temperature and pressure fluctuation range and determine a benchmark point.

[0125] In order to use the weighted fusion method to comprehensively adjust the vibration data analysis threshold when changes in temperature and pressure are detected simultaneously, a new threshold may be calculated by taking a weighted average of the temperature influence coefficient and the pressure influence coefficient.

[0126] In the specific implementation, when the changes of temperature and pressure are detected at the same time, the weighted fusion method is used to comprehensively adjust the influence of temperature and pressure and calculate the new vibration signal threshold. :

[0127]

[0128] in: is the vibration data analysis threshold at time t, is the reference temperature and base pressure The initial vibration signal threshold under and are the weight coefficients of temperature and pressure respectively, satisfying .

[0129] Temperature influence weight coefficient Reflects the relative importance of temperature changes to vibration signals. This weight can be determined through experimental data or actual operation data. For example, if temperature has a greater impact on vibration signals, a higher weight is assigned. The actual temperature of pipelines with different materials, structures, and geographical locations affects the weight coefficient. May not be the same.

[0130] Reflects the relative importance of pressure changes to vibration signals. Similarly, the pressure weight is determined by experimental data or actual operation data. If the pressure change has a significant impact on the vibration signal, a higher weight is given. Similarly, the actual pressure of pipelines with different materials, structures, and geographical locations affects the weight coefficient. May not be the same.

[0131] In this way, by using the weighted fusion method, the system can adapt to complex operating environments more flexibly and ensure the sensitivity and accuracy of vibration signal analysis when facing temperature and pressure changes at the same time. This method can effectively reduce false alarms and missed alarms and improve the reliability of the system.

[0132] In specific implementation, in order to achieve a dynamic response to temperature and pressure changes, it is necessary to determine the degree of influence of temperature and pressure on the vibration signal. This process can be achieved through linear regression analysis. Regression analysis is a statistical method that can derive the relationship between variables based on historical experimental data. Specifically, linear regression analysis is used to estimate the impact of temperature changes and pressure changes on vibration signal characteristics (such as vibration amplitude or frequency), thereby determining the temperature influence coefficient and pressure influence coefficient.

[0133] First, experimental data are collected under different operating conditions. These data include: vibration data measurement results at different temperatures, vibration data measurement results under different pressure data, and vibration characteristics recorded under the above temperature and pressure conditions, such as vibration amplitude or frequency response.

[0134] Based on the collected data, a linear regression model is constructed to describe the relationship between temperature, pressure, and vibration signal characteristics. The regression model assumes that there is a linear relationship between the vibration signal characteristics and temperature and pressure. Specifically, the output of the model (vibration signal characteristics) can be expressed as the initial value under the reference condition, plus the influence of temperature change and the influence of pressure change.

[0135] The experimental data is input into the model, for example, statistical software, statistical libraries in programming languages ​​(such as Python's scikit-learn), data analysis tools in Excel, or MATLAB, and a fitting analysis is performed. The result of the regression analysis is to estimate the coefficients in the model, especially the temperature influence coefficient and the pressure influence coefficient. The temperature influence coefficient indicates the degree of influence of temperature changes on the characteristics of the vibration signal, while the pressure influence coefficient indicates the degree of influence of pressure changes on the characteristics of the vibration signal.

[0136] After determining the temperature and pressure influence coefficients, the accuracy of the model is ensured through a validation process, where the vibration data characteristics predicted by the model are compared with the actual measured values. If the model's predictions are highly consistent with the actual conditions, it means that the regression model is reliable and these coefficients can be used in actual operation to dynamically adjust the threshold of the vibration signal based on the real-time monitored temperature and pressure data.

[0137] For example, the linear regression model is as follows:

[0138]

[0139] Where: V(T,P) is the vibration amplitude at temperature T and pressure P. is the intercept of the model, representing the base temperature and base pressure The vibration amplitude below. is the temperature influence coefficient, which indicates the effect of temperature change on the vibration amplitude. is the pressure influence coefficient, which indicates the effect of pressure change on the vibration amplitude.

[0140] The collected experimental data (temperature, pressure, and vibration amplitude) are input into the regression analysis tool, which will fit the best-fitting linear model to the input data. This model will minimize the difference between the observed and predicted values ​​(e.g., by the least squares method).

[0141] The analysis results will provide , and The estimated value of .

[0142] For example, suppose the analysis results are as follows:

[0143] ; ;

[0144] These coefficients indicate that the vibration amplitude will increase by 0.01g for every 1°C increase in temperature and by 0.02g for every 1MPa increase in pressure.

[0145] Through this method, the system can effectively use linear regression analysis to determine the impact of temperature and pressure on vibration signals, thereby dynamically adjusting the threshold of vibration signals under different working conditions to ensure the accuracy and reliability of pipeline monitoring.

[0146] In this way, by dynamically adjusting the analysis threshold of vibration data, the real-time changes of temperature and pressure are adapted. By calculating the temperature influence coefficient and the pressure influence coefficient and combining the weighted fusion method, the system can accurately monitor the pipeline status under different environmental conditions and ensure efficient detection of abnormal conditions. This method not only improves the accuracy of wax deposition detection, but also enhances the system's adaptability under complex working conditions.

[0147] As an optional implementation, in response to the flow data reaching a first change threshold within a first time period, extracting instantaneous frequency characteristics and energy characteristics of the vibration data within the first time period;

[0148] In response to the pressure data reaching a second change threshold within a second time period, time-frequency features of the vibration data within the second time period are extracted.

[0149] In specific implementations, the extraction of vibration signal features can also be triggered based on changes in external operating conditions (such as flow or pressure) rather than relying directly on the threshold of the vibration data itself. Significant changes in flow and pressure may indicate changes in the physical state of the pipeline (such as blockage, leakage, wax deposition, etc.), which may be reflected in the vibration signal but may not directly exceed the vibration data analysis threshold. This method aims to identify potential pipeline problems in advance by monitoring changes in flow and pressure. When these changes occur, although the vibration signal may not significantly exceed the threshold, early signs of abnormality can be captured by analyzing its characteristics.

[0150] In a specific implementation, the present disclosure continuously monitors the flow data of the pipeline and sets a first change threshold (e.g., a percentage of flow change). When the change of the flow data within a certain period of time reaches or exceeds this threshold, it can be considered that a significant change may have occurred in the pipeline (e.g., a decrease in flow caused by wax deposition).

[0151] Once the flow rate changes reach the set threshold, the vibration data within this time period is analyzed for instantaneous frequency. The instantaneous frequency can reveal the frequency changes of the vibration signal over time and capture the vibration response caused by the change in fluid flow state. The extraction of instantaneous frequency features can help identify pipeline flow blockage or recovery caused by phenomena such as wax deposition.

[0152] It is also possible to extract the energy features of the vibration signal during this time period and analyze the intensity changes of the vibration signal. Significant changes in flow are usually accompanied by fluctuations in the energy of the vibration signal. Extracting these features can further confirm whether abnormal conditions have occurred in the pipeline, such as vibration responses caused by drastic changes in flow.

[0153] In this way, the simultaneous use of these two trigger mechanisms (based on vibration data analysis thresholds and based on flow and pressure changes) can improve the comprehensiveness and accuracy of anomaly detection. If the system needs to be simplified or when resources are limited, one of the mechanisms can be retained based on the specific monitoring objectives.

[0154] The present disclosure also monitors the pressure data and sets a second change threshold (eg, the amplitude of the pressure fluctuation). When the change of the pressure data within a certain period of time reaches or exceeds this threshold, the system will consider that there may be potential abnormalities in the pipeline.

[0155] Pressure changes usually cause complex changes in vibration signals, involving the joint action of multiple frequency components. To this end, time-frequency analysis (such as short-time Fourier transform or wavelet transform) can be performed on the vibration data within this time period to capture the joint change characteristics of time and frequency in the vibration signal. The extraction of time-frequency features helps to identify resonances or mutations caused by pressure fluctuations, and helps the system determine whether there is wax deposition, leakage or other structural problems.

[0156] In this way, the simultaneous use of these two trigger mechanisms (based on vibration data analysis thresholds and based on flow and pressure changes) can improve the comprehensiveness and accuracy of anomaly detection. If the system needs to be simplified or when resources are limited, one of the mechanisms can be retained based on the specific monitoring objectives.

[0157] As an optional implementation, the historical wax cleaning data includes: wax cleaning ball data, ball throwing data;

[0158] Based on the wax cleaning ball data, the pitching data, the pressure data, and the flow data, wax cleaning effect characteristics are generated, and the wax cleaning effect characteristics include: flow recovery rate characteristics and pressure drop speed characteristics.

[0159] In a specific implementation, the present disclosure generates pipeline wax deposition prediction information and ball casting and receiving parameter information simultaneously through a preset machine learning model. To this end, historical data related to pipeline wax removal needs to be introduced into the construction of the machine learning model, which includes:

[0160] Wax cleaning ball data: information on the type, size, material, quantity, etc. of the wax cleaning balls used in each wax cleaning operation;

[0161] Ball throwing data: the throwing parameters of wax cleaning balls in each wax cleaning operation, such as throwing time, throwing interval, throwing pressure, throwing speed, etc.

[0162] Pressure data: changes in pressure inside the pipeline before and after wax cleaning.

[0163] Flow data: changes in flow in the pipeline before and after wax cleaning, especially flow recovery.

[0164] Based on the data collected above, features related to wax removal effect are extracted through feature engineering. These features can be directly input into the machine learning model to help optimize the parameters of pitching and receiving. The main wax removal effect features extracted include:

[0165] Traffic recovery rate characteristics:

[0166] The flow recovery rate indicates the degree of recovery of the flow in the pipeline after the wax cleaning operation relative to that before the wax cleaning. The specific calculation method is as follows:

[0167]

[0168] By comparing the flow data before and after wax cleaning, the flow recovery rate after each wax cleaning operation is calculated. This feature can reflect the effectiveness of the wax cleaning operation - a higher recovery rate usually means a more successful wax cleaning operation.

[0169] Pressure drop rate characteristics:

[0170] The pressure drop rate indicates the speed at which the pressure in the pipeline returns to normal level after the wax removal operation. Usually, it is expressed by calculating the rate of change of pressure per unit time:

[0171]

[0172] By monitoring the changes in pressure data before and after wax cleaning, the time required for the pressure to drop from the peak value to a certain stable value is calculated. The faster the pressure drops, it usually indicates that the wax cleaning operation is more effective in restoring the normal pressure of the pipeline.

[0173] After the wax removal effect features are generated, they will be input into the preset machine learning model. The machine learning model will learn the relationship between the wax removal effect features and the pitching and receiving parameters through training with historical data.

[0174] The machine learning model is trained using historical operation data and historical wax cleaning data. The model will identify which ball-dropping parameters (such as drop pressure, drop interval, wax cleaning ball type, etc.) can produce the best wax cleaning effect under different pipeline conditions.

[0175] For example, the wax deposition data in the historical operation data can be used as labels (for example, by detecting the wax layer thickness, pipeline flow slowdown, etc.), and the training model can be used to identify possible wax deposition conditions under specific characteristic conditions. The wax removal effects under different pitching and receiving parameters (such as flow recovery rate, pressure drop rate) are labeled, and the training model can identify which pitching parameters can achieve the best wax removal effect under specific conditions.

[0176] There are many ways to detect the thickness of the wax layer, such as measuring the thickness of the wax layer by reflecting and transmitting ultrasonic waves between the inner wall of the pipe and the wax layer. When the ultrasonic sensor sends out ultrasonic signals, these signals will be partially reflected back when encountering the wax layer on the inner wall of the pipe, while part of them will continue to penetrate the wax layer and reflect between the wax layer and the fluid or pipe wall. By analyzing the reflected ultrasonic signals, especially the delay in reflection time and the change in signal intensity, the thickness of the wax layer can be calculated. This method usually uses sensors outside the pipeline and does not require damage to the pipeline.

[0177] Another example is the use of the reflection and absorption characteristics of electromagnetic waves in the wax layer. Similar to ultrasound, electromagnetic waves will reflect and refract when encountering different media, which can be used to infer the thickness of the wax layer. Using an electromagnetic wave transmitter and receiver device, electromagnetic waves are emitted into the pipe. When the electromagnetic waves pass through the wax layer, the reflection and attenuation characteristics of the signal will change. By analyzing these changes, the thickness of the wax layer can be estimated.

[0178] In addition, mechanical measurement, resistive sensor, X-ray and other measurement methods can also be selected according to actual conditions, which will not be described here.

[0179] In the specific implementation, by inputting the target features extracted from the operating data of the target pipeline into the trained machine learning model, the model will generate pipeline wax deposition prediction information and ball casting and receiving parameter information.

[0180] This information will include the recommended wax cleaning ball type, quantity, placement pressure, placement time, etc.

[0181] It should be noted that different ball throwing and receiving parameter information will produce different wax cleaning effects. The material of the wax cleaning ball (such as rubber, polyurethane or composite material) determines its flexibility and hardness in the pipeline. Soft materials can better adapt to the shape of the inner wall of the pipeline and clean the wax layer evenly; while materials with higher hardness are suitable for removing thicker or harder wax layers. The size of the wax cleaning ball directly affects its contact area in the pipeline. Larger wax cleaning balls can increase the contact area with the inner wall of the pipeline, thereby improving the wax cleaning efficiency, but it will also increase the difficulty of advancement. Wax cleaning balls that are too small may not be able to effectively remove the wax layer on the pipe wall.

[0182] The injection pressure determines the speed and friction of the wax cleaning ball in the pipeline. When the injection pressure is high, the wax cleaning ball moves faster in the pipeline and has greater friction on the wax layer, which can more effectively remove the wax layer on the inner wall of the pipeline. However, too high pressure may cause the wax cleaning ball to deform or damage, and even damage the pipeline.

[0183] Wax removal balls can generate greater kinetic energy at higher speeds, thereby increasing the impact and friction on the wax layer on the inner wall of the pipe. This is suitable for scenarios where thick wax layers need to be cleaned quickly. However, too high a speed may prevent the wax removal balls from fully contacting the inner wall of the pipe, especially at bends or joints in the pipe. When the delivery speed is lower, the wax removal balls move more smoothly in the pipe, and can more fully contact and remove the wax layer on the inner wall of the pipe. This is suitable for scenarios that require delicate cleaning, but too low a speed may reduce the wax removal efficiency.

[0184] Frequent placement of wax balls can keep the pipeline clean, especially when the wax is seriously deposited in the pipeline. Short intervals of wax balls can prevent the wax layer from accumulating again. However, too frequent placement may cause excessive pressure fluctuations in the pipeline, increasing the complexity of the wax removal operation. Longer placement intervals are suitable for situations where the pipeline is less waxed or the pipeline operating conditions are relatively stable. This method can reduce the frequency and cost of wax removal operations, but if the interval is too long, it may cause the wax layer to accumulate too thickly, increasing the difficulty of subsequent cleaning.

[0185] The number of wax cleaning balls placed directly affects the wax cleaning effect. Placing multiple wax cleaning balls at a time can increase the coverage of wax cleaning and improve the wax cleaning efficiency, but it will also increase the pressure load in the pipeline. If there are too many wax cleaning balls, it may cause mutual interference between the wax cleaning balls, affecting the overall wax cleaning effect.

[0186] The present invention generates pitch and receive parameter information by means of relevant data collection, data analysis, and feature engineering, using the historical operation data and wax cleaning data of the pipeline combined with a machine learning model. Based on this information, control instructions corresponding to the pitch and receive parameter information can be generated.

[0187] It should be noted that in the present disclosure, the ball collection operation will be uniformly performed at the end of the pipeline, and the timing of collecting the ball can be estimated based on the time it takes for the wax-clearing ball to pass through the pipeline. When the wax-clearing ball reaches the end, the system automatically triggers the ball collection operation. In the control instructions generated by the processing module 20, the specific ball collection related instructions can be set as: in response to the wax-clearing ball being thrown, the state of the device for collecting the ball is adjusted to the waiting state.

[0188] The present disclosure also provides an intelligent ball throwing and receiving device, see Figure 3 , Figure 3 The structure diagram of an intelligent ball throwing and receiving device provided in an embodiment of the present application is shown; the intelligent ball throwing and receiving device 200 is used to execute the control instructions generated by the above control system, and the device includes:

[0189] The ball throwing unit 210 is equipped with a motor for adjusting and executing the ball throwing operation, wherein the ball throwing operation includes: throwing the wax cleaning ball into the target pipe based on the ball throwing and receiving parameter information in the control instruction;

[0190] The ball collecting unit 220 is provided with a mechanical capture device for collecting the wax cleaning balls put into the target pipeline;

[0191] A control unit 230, configured to receive a control instruction from the processing module 20, and control the pitching unit 210 and the receiving unit 220 to operate based on the control instruction;

[0192] Wherein, the control unit 230 adjusts the pitching frequency, ball size, pitching material and pitching intensity of the pitching unit 210 based on the control instruction;

[0193] In a specific implementation, the pitching unit 210 is equipped with a high-precision motor, which controls the frequency, size and pitching intensity of the wax-clearing balls by adjusting the speed and torque. The motor is directly connected to the throwing mechanism of the pitching unit 210.

[0194] The ball throwing unit 210 is provided with an adjustable caliber mechanism, which can automatically adjust the ball throwing caliber according to the control command to adapt to wax cleaning balls of different sizes. The caliber adjustment mechanism adopts an electric adjustment ring and a retractable catheter structure, and the adjustment of the ball throwing caliber is achieved by driving the adjustment ring to rotate or the catheter to retract by the motor. This design allows the ball throwing unit 210 to flexibly handle wax cleaning balls of various specifications, ensuring the diversity and adaptability of the wax cleaning operation.

[0195] See also Figure 4 , Figure 4 This is a schematic diagram of the structure of another intelligent ball throwing and receiving device provided in an embodiment of the present application; an automatic switching mechanism is also provided in the ball throwing unit 210, which is used to automatically select wax cleaning balls of appropriate materials according to different wax cleaning tasks. The switching mechanism is implemented by multiple wax cleaning ball sub-storage bins in the wax cleaning ball storage bin 211, and each sub-storage bin stores wax cleaning balls of different materials (such as rubber balls, polyurethane balls, etc.). The system selects and throws appropriate wax cleaning ball materials through control instructions.

[0196] The control system of the motor integrates a frequency regulator and an intensity controller. The frequency regulator is used to set the frequency of releasing wax balls according to the wax deposition situation in the pipeline. The intensity controller controls the intensity of releasing the balls by adjusting the output power of the motor, so that the wax balls can enter the pipeline at a suitable speed and intensity.

[0197] The ball collecting unit 220 is disposed at the end of the pipeline and is used to collect the wax cleaning balls from the pipeline after the wax cleaning operation is completed, so as to ensure that the wax cleaning balls do not remain in the pipeline and affect subsequent operations. The ball collecting unit 220 is equipped with a mechanical capture device, which includes a retractable mechanical arm and a plurality of flexible capture claws. When the wax cleaning ball reaches the end of the pipeline, the mechanical arm extends and captures the wax cleaning ball.

[0198] The control unit 230 receives control instructions sent by the processing module 20 through an electrical connection with the processing module 20. These instructions include parameters such as pitching frequency, pitching size, pitching material, pitching intensity, etc. The control unit 230 parses these instructions and sends corresponding control signals to the pitching unit 210 and the ball receiving unit 220.

[0199] Based on the analyzed instructions, the control unit 230 adjusts the motor output and the caliber adjustment mechanism of the ball-dropping unit 210 to ensure that the wax-clearing balls are accurately dropped according to the specified frequency, size, material and strength. At the same time, the control unit 230 also monitors the real-time data of the ball-dropping operation and makes adjustments when necessary to adapt to the actual operation conditions in the pipeline.

[0200] When the wax cleaning ball is about to reach the end of the pipeline, the control unit 230 starts the mechanical capture device of the ball receiving unit 220, adjusts the extension and retraction of the mechanical arm and the strength of the capture claw to ensure that the wax cleaning ball is safely recovered.

[0201] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0202] It should be understood that determining B based on A does not mean determining B only based on A. B can also be determined based on A and / or other information.

[0203] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0204] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0205] The preferred embodiments of the present application disclosed above are only used to help explain the present application. The preferred embodiments do not describe all the details in detail, nor do they limit the present application to specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present application, so that those skilled in the art can understand and use the present application well. The present application is only limited by the claims and their full scope and equivalents.

Claims

1. A control system for an intelligent ball throwing and receiving device, used to control an intelligent ball throwing and receiving device, characterized in that: include: A collection module, used to obtain the operation data of the target pipeline through a sensor, wherein the operation data includes: temperature data, pressure data, flow data, wax deposition data, and vibration data; A processing module, configured to predict wax deposition on the target pipeline based on the operating data of the target pipeline and to generate a control instruction by using a preset machine learning model; The preset machine learning model is constructed based on the historical operation data and historical wax removal data of the target pipeline; the preset machine learning model is used to generate pipeline wax deposition prediction information and ball casting and receiving parameter information based on the input target features; The control instruction is used to control the intelligent pitching and receiving device to perform pitching and receiving based on the pitching and receiving parameter information; The method of predicting wax deposition on the target pipeline based on the operation data of the target pipeline and generating control instructions by using a preset machine learning model includes: In response to the vibration data reaching a vibration data analysis threshold, extracting features from the vibration data to generate vibration features; Wherein, the vibration data analysis threshold is set based on the wax deposition data and the vibration data in the historical operation data of the target pipeline; The vibration characteristics include: vibration time domain characteristics, vibration frequency domain characteristics, and vibration time and frequency domain characteristics; Extracting features from the vibration data to generate vibration features includes: extracting statistical features from the vibration data to generate the vibration time domain features; Performing Fourier transform on the vibration data to generate the vibration frequency domain features; The vibration data is subjected to time-frequency analysis using wavelet transform to generate the vibration time-frequency domain features.

2. The intelligent ball throwing and receiving device control system according to claim 1 is characterized in that: The wax deposition data is used to characterize the wax layer thickness at each location in the target pipeline; Marking a position in the target pipeline where the historical average value of the wax layer thickness is greater than the target thickness average value as a first wax deposition area; The acquisition module further includes: a placement unit; The arrangement unit is used to arrange the sensor at a target section position of the target pipeline, wherein the target section position includes: the first wax deposition area of ​​the target pipeline, the starting point of the target pipeline, and the bend of the target pipeline; The acquisition module is also used to set the vibration data analysis threshold for each target section position based on the historical operation data collected by the sensor arranged at each target section position.

3. The intelligent ball throwing and receiving device control system according to claim 2 is characterized in that: Also includes: In response to the temperature data being greater than a preset reference temperature, increasing the vibration data analysis threshold based on a preset temperature influence coefficient; In response to the pressure data being greater than a preset reference pressure, increasing the vibration data analysis threshold based on a preset pressure influence coefficient; The preset temperature influence coefficient and the preset pressure influence coefficient are determined through linear regression analysis based on historical operation data of the target pipeline.

4. The intelligent ball throwing and receiving device control system according to claim 3 is characterized in that: Also includes: In response to the temperature data being greater than a preset reference temperature and the pressure data being greater than a preset reference pressure, the vibration data analysis threshold is increased by a weighted fusion method based on the preset temperature influence coefficient and the preset pressure influence coefficient.

5. The intelligent ball throwing and receiving device control system according to claim 4 is characterized in that: Also includes: In response to the flow data reaching a first change threshold within a first time period, extracting instantaneous frequency characteristics and energy characteristics of the vibration data within the first time period; In response to the pressure data reaching a second change threshold within a second time period, time-frequency features of the vibration data within the second time period are extracted.

6. The intelligent ball throwing and receiving device control system according to claim 5 is characterized in that: The historical wax cleaning data includes: wax cleaning ball data and ball throwing data; Based on the wax cleaning ball data, the pitching data, the pressure data, and the flow data, wax cleaning effect characteristics are generated, and the wax cleaning effect characteristics include: flow recovery rate characteristics and pressure drop speed characteristics.

7. The intelligent ball throwing and receiving device control system according to claim 6 is characterized in that: The acquisition module further includes: a synchronization unit; the synchronization unit is used to synchronize the operation data acquired by the sensors arranged at different positions of the target pipeline to the same timestamp.

8. An intelligent ball throwing and receiving device, used to execute the control instructions generated by the control system of the intelligent ball throwing and receiving device according to any one of claims 1 to 7, characterized in that: include: A ball throwing unit is provided with a motor for adjusting and executing a ball throwing operation, wherein the ball throwing operation comprises: throwing a wax cleaning ball into the target pipe based on the ball throwing and receiving parameter information; A ball collecting unit, equipped with a mechanical capture device, for collecting the dropped wax cleaning balls from the target pipeline; A control unit, configured to receive a control instruction from the processing module and control the pitching unit and the receiving unit to operate based on the control instruction; Wherein, the control unit adjusts the pitching frequency, ball size, pitching material and pitching intensity of the pitching unit based on the control instruction.

Images