Control method of deep-well pump and deep-well pump

Abstract

The invention relates to a deep-well pump control method and a deep-well pump, and the method comprises the steps: continuously obtaining sound data collected by a vibration sensor of a deep-well pump body, and segmenting the sound data in a time dimension to obtain sound analysis data; performing filtering processing and noise reduction processing on the sound analysis data to obtain basic analysis data; converting the basic analysis data into image data; acquiring surrounding image data of the features and determining texture features of the surrounding image data; analyzing the texture features by using a comparison mode, and classifying the features by using an analysis result; and monitoring the occurrence time and / or occurrence frequency of the abnormal characteristics and adjusting the running state of the deep-well pump according to the monitoring result. According to the control method of the deep-well pump and the deep-well pump, the operation state of the deep-well pump in the operation process can be judged and the operation state of the deep-well pump can be adjusted at the same time in a localized comparison processing mode, so that safe operation of the deep-well pump is guaranteed.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a control method for a deep well pump and a deep well pump. Background Technology

[0002] The core of deep well pump control is automatic start-stop, variable frequency speed regulation, multiple protections and remote monitoring to achieve safety, energy saving and stable water supply. At present, the control of deep well pumps is shifting to intelligence. The core is to achieve unattended operation, safety and stability, high efficiency and energy saving through automatic sensing + intelligent control + comprehensive protection + data monitoring + remote operation and maintenance + network platform.

[0003] Taking fault diagnosis as an example, there are judgments based on electrical parameters (current, voltage, phase loss, three-phase imbalance), water level (water shortage, water pressure, abnormal fluctuation), and temperature (winding overheating, bearing overheating). These judgments can all be made by sensors and have clear indicator values.

[0004] Judgments based on non-indicative values ​​mainly focus on fault diagnosis, such as cavitation, bearing damage, impeller jamming and rubbing against the casing, rotor imbalance, loose resonance, and internal parts falling off. Currently, acoustic analysis is often used for this purpose, but the following problems exist: When the installation environment and working status of the deep well pump change, the sound samples will also change, and this change is uncontrollable, resulting in a lack of sufficient samples to train the recognition model and insufficient generalization ability of the recognition model. The accuracy of judgments is insufficient due to the simplification of parameters in locally deployed distillation models. Data transmission using cloud-edge collaboration is too costly. Summary of the Invention

[0005] This application provides a control method and a deep well pump for a deep well pump. The method can determine the operating status of the deep well pump during operation and adjust the operating status of the deep well pump through a localized comparison processing method to ensure the safe operation of the deep well pump.

[0006] The above-mentioned objective of this application is achieved through the following technical solution: Firstly, this application provides a method for controlling a deep well pump, including: Continuously acquire sound data from the vibration sensors at the deep well pump headquarters and segment the sound data in the time dimension to obtain sound analysis data; The sound analysis data is filtered and noise-reduced to obtain the basic analysis data; The basic analysis data is converted into image data, which is a grayscale image. The features included in the image data are obtained, including peak points and peak regions. Acquire the surrounding image data of the feature and determine the texture features of the surrounding image data; Texture features are analyzed using a comparative method, and the analysis results are used to classify the features, including normal features and abnormal features. Monitor the occurrence time and / or frequency of abnormal features and adjust the operating status of the deep well pump based on the monitoring results.

[0007] In one possible implementation of the first aspect, when analyzing texture features using a comparison method, a comparison sample matching the operating status parameters is selected based on the operating status parameters of the deep well pump. The operating parameters of a deep well pump include operating voltage, operating current, active power, bearing temperature, motor temperature, motor speed, inlet water pressure, outlet water pressure, and flow rate. Select at least one deep well pump's operating status parameters to choose a comparison sample.

[0008] In one possible implementation of the first aspect, converting the basic analysis data into image data includes: The basic analysis data is decomposed in the frequency domain to obtain multiple sound curves, and any two sound curves have different frequencies. Set data acquisition points at intervals along the time dimension and determine the amplitude points on the sound curve corresponding to the data acquisition points; The data acquisition points and their corresponding amplitude points are transferred into a coordinate system for representation to obtain image data.

[0009] In one possible implementation of the first aspect, the features included in the image data are: Generate multiple numerically continuous grayscale ranges; Image data is extracted using grayscale ranges to obtain the extracted content, with only one grayscale range used for each extraction. The extracted content is filtered to obtain the features included in the image data; The filtering process includes removing areas in the extracted content that are greater than or equal to a set area value.

[0010] In one possible implementation of the first aspect, the features included in the image data are: The image data is divided into multiple rectangular regions, each containing the same number of pixels; Calculate the average pixel value for each rectangular region; The average pixel values ​​of two adjacent rectangular regions are compared in a fixed direction, and a value is assigned based on the comparison result, including zero and one. The assignment results for each rectangular region are statistically analyzed, and the features included in the image data are obtained based on the assignment results. The fixed directions include up, down, left, right, upper left, upper right, lower left, and lower right.

[0011] In one possible implementation of the first aspect, acquiring the surrounding image data of the feature and determining the texture features of the surrounding image data includes: Determine the center point of the feature; A selection region is created based on the center point of the feature, and the pixels within the selection region are the surrounding image data of the feature. The selected area is divided into multiple rectangular regions, each containing the same number of pixels; Calculate the average pixel value for each rectangular region; The average pixel values ​​of two adjacent rectangular regions are compared in a fixed direction, and a value is assigned based on the comparison result, including zero and one. The assignment results for each rectangular region are statistically analyzed, and the assignment results are used as texture features of the surrounding image data. The fixed directions include up, down, left, right, upper left, upper right, lower left, and lower right.

[0012] In one possible implementation of the first aspect, analyzing texture features using a comparison method includes: Calculate the texture features of the standard samples; Calculate the similarity between the texture features of the standard sample and the texture features of the surrounding image data to obtain the similarity results, which include similar and dissimilar results; The similarity result of an image dataset is the ratio of dissimilar surrounding image data to all surrounding image data. Features are classified based on their ratios, and these features include normal features and abnormal features.

[0013] Secondly, this application provides a control device for a deep well pump, comprising: The data acquisition unit is used to continuously acquire sound data collected by the vibration sensor of the deep well pump and segment the sound data in the time dimension to obtain sound analysis data; The data processing unit is used to filter and reduce noise in the sound analysis data to obtain basic analysis data. The data conversion unit is used to convert basic analysis data into image data, which is a grayscale image. The first feature acquisition unit is used to obtain the features included in the image data, including peak points and peak regions. The second feature acquisition unit is used to acquire the surrounding image data of the feature and determine the texture features of the surrounding image data; The classification unit is used to analyze texture features using a comparative method and classify the features using the analysis results. The features include normal features and abnormal features. The control unit is used to monitor the occurrence time and / or frequency of abnormal features and adjust the operating status of the deep well pump based on the monitoring results.

[0014] Thirdly, this application provides a deep well pump, the deep well pump comprising: One or more memories for storing instructions; and One or more processors are configured to retrieve and execute the instructions from the memory to perform the methods described in the first aspect and any possible implementation thereof.

[0015] Fourthly, this application provides a computer-readable storage medium, the computer-readable storage medium comprising: The program, when run by a processor, is executed as described in the first aspect and any possible implementation thereof.

[0016] Fifthly, this application provides a computer program product, including program instructions that, when run by a computing device, execute the method described in the first aspect and any possible implementation thereof.

[0017] Sixthly, this application provides a chip system including a processor for implementing the functions involved in the foregoing aspects, such as generating, receiving, transmitting, or processing the data and / or information involved in the foregoing methods.

[0018] This chip system can consist of chips or include chips and other discrete components.

[0019] In one possible design, the chip system also includes a memory for storing necessary program instructions and data. The processor and the memory can be decoupled and located on different devices, connected via wired or wireless means, or the processor and the memory can be coupled to the same device. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the steps of a deep well pump control method provided in this application.

[0021] Figure 2 This is a diagram of sound data collected by a vibration sensor provided in this application.

[0022] Figure 3This is a schematic diagram of a process for converting basic analysis data into image data, as provided in this application.

[0023] Figure 4 This is a schematic diagram of a process for converting basic analysis data into image data, as provided in this application.

[0024] Figure 5 This is a schematic diagram of obtaining features included in image data, as provided in this application.

[0025] Figure 6 This is another schematic diagram of obtaining features included in image data provided in this application. Detailed Implementation

[0026] The technical solutions in this application will be further described in detail below with reference to the accompanying drawings.

[0027] This application discloses a control method for a deep well pump. Please refer to [link / reference]. Figure 1 In some examples, the deep well pump control method disclosed in this application includes the following steps: S101, continuously acquire sound data collected by the vibration sensor of the deep well pump and segment the sound data in the time dimension to obtain sound analysis data; S102 performs filtering and noise reduction on the sound analysis data to obtain basic analysis data; S103, convert the basic analysis data into image data, which is a grayscale image; S104, obtain the features included in the image data, including peak points and peak regions; S105, acquire the surrounding image data of the feature and determine the texture features of the surrounding image data; S106, use comparison to analyze texture features and use the analysis results to classify the features, including normal features and abnormal features; S107, monitor the occurrence time and / or frequency of abnormal features and adjust the operating status of the deep well pump based on the monitoring results.

[0028] First, we will further explain the deep well pump and its working conditions. The deep well pump works in a deep underground well, which is immersed in clean water or groundwater for a long time. The ambient temperature, water level and water quality fluctuate greatly. It contains a small amount of silt and sand, and the underwater pressure and humidity are high, making the overall working conditions complex.

[0029] The recognition model mentioned in the background technology can be deployed inside the local control cabinet of a deep well pump. Data generated by sensors inside the deep well pump is sent to the local control cabinet via data cable, and then the recognition model processes the data. The main problems are as follows: Insufficient computing power: Edge industrial control computers / edge gateways are mostly ARM or low-power x86, without independent GPUs or only have weak computing power AI chips, resulting in slow inference and high response latency, which cannot meet the requirements of real-time control; Insufficient memory: Even small quantized models (such as 1B to 3B) require 2GB to 8GB of memory, while edge devices typically only have 1GB to 4GB, making them prone to memory overflow and crashes; Harsh environment, low reliability; Strict power consumption limits mean that high power consumption in large-scale computing modules can easily lead to unstable power supply, tripping, and disruption of pump control.

[0030] One solution to these problems is to deploy a distillation model. However, distillation models suffer from insufficient model accuracy and generalization ability, limited edge hardware resources, difficulty in meeting industrial control requirements in terms of real-time performance, unreliable black-box decision-making, and high costs.

[0031] It should be noted that the above content is only a description of current technology and does not involve any restrictions on future technologies.

[0032] The solution proposed in this application is to compare normal sound samples with collected sound samples. If there is an anomaly in the collected sound sample, it is determined that the operating status of the deep well pump is abnormal during the time period corresponding to the collected sound sample.

[0033] Specifically, in step S101, sound data collected by the vibration sensor of the deep well pump will be continuously acquired. Figure 2 (As shown) The sound data is then segmented along the time dimension to obtain sound analysis data. This sound analysis data is arranged sequentially, with each analysis focusing on only one piece of sound analysis data. Next, in step S102, the sound analysis data undergoes filtering and noise reduction processing to obtain basic analysis data. The specific details of the filtering process are as follows: Mean filtering: Smooths continuous audio signals and removes glitches; Median filtering: removes impulse noise and spike interference; Band-stop filtering / notch filtering: filters out fixed-frequency interference (50Hz power frequency noise, inverter switching noise); The purpose of these three filtering methods is to remove the inherent noise and environmental noise of the deep well pump during operation, while retaining the noise that may be related to the malfunction of the deep well pump.

[0034] The specific steps of mean filtering are as follows: Determine the window size (e.g., 3, 5, 7, 9 points); The sliding window starts from the first point of the sound signal; Sum the sample points within each window; Divide the summation result by the window length to obtain the filtered value for the current point; The window slides from beginning to end, and the smoothed sound signal is obtained.

[0035] The specific steps of median filtering are as follows: 1. Determine the filter window length: Select an odd-numbered window length (commonly 3, 5, 7, or 9 points) to process the current sampling point and several adjacent points on its left and right sides together each time.

[0036] 2. Window sliding traversal of the signal: Starting from the starting point of the sound signal, slide the window from left to right to cover all sampling points one by one, and for each position, take out a set of signal values ​​within the window.

[0037] 3. Sort the data in the window: Arrange the sampled values ​​in the window in ascending order.

[0038] 4. Take the median as the current output: In the sorted data, take the value at the middle position and use it to replace the original center sampling point as the filtered value.

[0039] 5. Continue sliding until all processing is complete: keep moving the window forward and repeat steps 2 to 4 until the entire audio signal has been filtered.

[0040] Boundary handling: For points at the very beginning and end of the signal that cannot be filled in the window, repeat boundary values ​​or truncation methods are used to ensure complete filtering.

[0041] In step S103, the basic analysis data needs to be converted into image data, which is a grayscale image, such as... Figure 3 As shown, in step S104, the features included in the image data are obtained. The features include peak points and peak regions. The difference between peak points and peak regions is their area. For example, a feature consisting of 6 or fewer pixels is called a peak point, and a feature consisting of more than 6 pixels is called a peak region. The terms "peak point" and "peak region" are used here for explanation only.

[0042] It should be noted that conventional sound analysis methods involve transferring the basic analysis data into the frequency or time domain for processing, and then using methods such as comparison, fitting, and filtering to discover sounds of specific frequencies or combinations of specific frequencies in the basic analysis data.

[0043] This approach is based on substantial foundational data, which involves collecting and analyzing the operational data of deep well pumps under various environments. Then, through manual labeling, we determine which sounds are faulty, and this data is used as a database.

[0044] The training of the recognition model also adopts a similar approach, feeding a massive amount of manually labeled data to enable the recognition model to have high recognition accuracy. However, this method requires a long period of accumulation. As mentioned in the background technology, when the installation environment and working status of the deep well pump change, its sound samples will also change, which leads to insufficient sample size and difficulty in manual labeling. At present, it is difficult to use this data to train the recognition model.

[0045] It should also be noted that the recognition model trained using standard data has very limited application scenarios. When the application scenario changes, its prediction results become unreliable and cannot be directly used to determine the working status of deep well pumps.

[0046] In step S105, the surrounding image data of the feature is acquired and the texture features of the surrounding image data are determined. Then, in step S106, the texture features are analyzed using a comparison method and the analysis results are used to classify the features, which include normal features and abnormal features.

[0047] Finally, in step S107, the occurrence time and / or frequency of abnormal features are monitored, and the operating status of the deep well pump is adjusted according to the monitoring results.

[0048] The above steps omit traditional time-domain / frequency-domain analysis and instead directly use texture feature comparison to determine normal and abnormal features. After classifying the features, the operating status of the deep well pump is adjusted based on the occurrence time and / or frequency of abnormal features, as detailed below: First, the abnormalities are classified into occasional abnormalities (occasionally appearing and disappearing quickly), intermittent abnormalities (recurring and sometimes absent), and continuous abnormalities (constantly existing and becoming increasingly severe). For occasional anomalies, only record them and do not make any adjustments; in this case, the operating status of the deep well pump remains unchanged. For intermittent anomalies, record and adjust them. The adjustment method is to reduce the frequency / load and continue to observe. If the anomaly still occurs, it needs to be reported to maintenance. If the anomaly disappears, it means that the anomaly only occurs in a certain state. In subsequent work, it is necessary to avoid the occurrence of that state or quickly skip that state. This can be achieved by adjusting the operating current, operating voltage, head, etc. In the event of continuous anomalies, immediately shut down the system for protection.

[0049] In some cases, when analyzing texture features using a comparative approach, comparison samples that match the operating status parameters of the deep well pump are selected. The operating parameters of a deep well pump include operating voltage, operating current, active power, bearing temperature, motor temperature, motor speed, inlet water pressure, outlet water pressure, and flow rate. Select at least one deep well pump's operating status parameters to choose a comparison sample.

[0050] This section means that the comparison samples need to be selected based on the operating parameters of the deep well pump. This is because when the operating parameters of the deep well pump are different, the sound data collected by the vibration sensor will also change. If only a fixed comparison sample is used, false alarms are likely to occur.

[0051] The operating parameters of a deep well pump include operating voltage, operating current, active power, bearing temperature, motor temperature, motor speed, inlet water pressure, outlet water pressure, and flow rate. When selecting a comparison sample, at least one of these operating parameters must be selected. For example, operating voltage, operating current, inlet water pressure, outlet water pressure, and flow rate can be selected. Of course, this is just an example and is not intended to limit the application.

[0052] It should be noted that when there are many operating status parameters selected for deep well pumps, it is easy to encounter the problem of difficulty in selecting comparison samples. This is because the more parameters there are, the more detailed the operating status of the deep well pump needs to be. Therefore, it is generally recommended to select 2-3 operating status parameters.

[0053] In some cases, the specific steps for converting basic analytical data into image data are as follows: The basic analysis data is decomposed in the frequency domain to obtain multiple sound curves, and any two sound curves have different frequencies. Set data acquisition points at intervals along the time dimension and determine the amplitude points on the sound curve corresponding to the data acquisition points; The data acquisition points and their corresponding amplitude points are transferred into a coordinate system for representation to obtain image data.

[0054] Specifically, this involves using Fourier transform to decompose the basic analysis data, then determining the x-coordinate (using the x-coordinate of the data acquisition point as the x-coordinate) and y-coordinate (converting the frequency of the sound curve to the y-coordinate) based on the y-coordinate of the point on the sound curve that shares the x-coordinate with the data acquisition point. Finally, the y-coordinate of the point on the sound curve that shares the x-coordinate with the data acquisition point is used as the grayscale value. Figure 4 As shown in the figure, the transformation of two sound curves is used to replace the original.

[0055] This involves converting the data acquisition points and their corresponding amplitude points into a coordinate system for representation, which yields the image data.

[0056] From another perspective, the multiple sound curves obtained after the basic analysis data decomposition are first sorted according to frequency. Each frequency corresponds to a vertical axis, the horizontal axis is time, the horizontal axis is the time point, and the amplitude of the curve at the position of the horizontal and vertical axes is the pixel value of the pixel point.

[0057] In some examples, the specific ways to obtain the features included in image data are as follows: Generate multiple numerically continuous grayscale ranges; Image data is extracted using grayscale ranges to obtain the extracted content, with only one grayscale range used for each extraction. The extracted content is filtered to obtain the features included in the image data; The filtering process includes removing areas in the extracted content that are greater than or equal to a set area value.

[0058] The grayscale range here is formed by dividing the continuous values ​​from 0 to 255 into equal parts. Each grayscale range has a length of 5 to 10 consecutive values. Then, the image data is extracted using the grayscale ranges. At this point, the extracted content is obtained. The number of grayscale ranges is the same as the number of extracted content.

[0059] Then, the extracted content is filtered to obtain the features included in the image data. Figure 5 As shown, the filtering process includes removing areas in the extracted content that are greater than or equal to a set area value. This is because the technical solution in this application addresses early failures of deep well pumps, which are characterized by weak signals, intermittent occurrences, and high-frequency components. Therefore, it is necessary to remove areas in the extracted content that are greater than or equal to a set area value.

[0060] For a given area value, calculate it using the following method: The duration of minor noise is 0.1 to 0.5 seconds, and the duration of minor impact / friction is 0.3 to 1 second. After determining the duration, the number of horizontal pixels for the set area value is determined according to the sampling frequency. The number of vertical pixels for the set area value is generally less than or equal to the number of vertical pixels for the set area value. Of course, it can also be adjusted adaptively according to the specific scenario during use.

[0061] Another way to obtain the features included in image data is as follows: The image data is divided into multiple rectangular regions, each containing the same number of pixels; Calculate the average pixel value for each rectangular region; The average pixel values ​​of two adjacent rectangular regions are compared in a fixed direction, and a value is assigned based on the comparison result, including zero and one. The assignment results for each rectangular region are statistically analyzed, and the features included in the image data are obtained based on the assignment results. The fixed directions include up, down, left, right, upper left, upper right, lower left, and lower right.

[0062] Please see Figure 6 This method determines the value by comparing pixel values. The average pixel values ​​of two adjacent rectangular regions are compared, and the region with the higher average value is assigned a value of 1. Figure 6 (The black rectangle in the middle) assigns a value of 0 to a rectangular region where the comparison result is less than ( Figure 6 (a white rectangle).

[0063] If the number of times the value of 1 occurs in 8 comparisons (8 in a fixed direction, so 8 comparisons are required) is greater than the set number (e.g., 5 times), then this rectangular area is marked as a feature. Of course, the five times mentioned here is only used as an example and does not constitute a limitation of this application.

[0064] The core of this method is to determine the length (number of pixels) and width (number of pixels) of the rectangular area. Here, we still refer to the method of setting the area value. For the rectangular area corresponding to different situations, there are different lengths and widths. For example, if there are 10 dimensions (length and width) at this time, then the above method needs to be executed 10 times, using a different dimension each time, and the multiple results are combined and used.

[0065] In some examples, the way to obtain the surrounding image data of the feature and determine the texture features of the surrounding image data is as follows: Determine the center point of the feature; A selection region is created based on the center point of the feature, and the pixels within the selection region are the surrounding image data of the feature. The selected area is divided into multiple rectangular regions, each containing the same number of pixels; Calculate the average pixel value for each rectangular region; The average pixel values ​​of two adjacent rectangular regions are compared in a fixed direction, and a value is assigned based on the comparison result, including zero and one. The assignment results for each rectangular region are statistically analyzed, and the assignment results are used as texture features of the surrounding image data. The fixed directions include up, down, left, right, upper left, upper right, lower left, and lower right.

[0066] In the above method, the selected area is divided into multiple rectangular areas, where the size of the rectangular area is 3x3 to 8x8. The value refers to the number of pixels. The average pixel value of two adjacent rectangular areas is compared in a fixed direction. Taking the fixed direction as the top as an example, the average pixel value of the lower rectangular area is compared with the average pixel value of the upper rectangular area. If the comparison result is greater than the lower rectangular area, the lower rectangular area is assigned a value of one; otherwise, it is assigned a value of zero.

[0067] The comparison is performed eight times (in fixed directions including up, down, left, right, upper left, upper right, lower left, and lower right). Finally, the assignment results for each rectangular region are calculated. After the calculation is completed, the assignment results are used as the texture features of the surrounding image data. Specifically, the assignment results for each rectangular region are accumulated.

[0068] The advantages of obtaining texture features in this way are as follows: It has extremely fast calculation speed and strong real-time performance, which can meet the timeliness requirements of deep well pump control; Robust to noise and strong anti-interference ability; Texture features are intuitive and have clear physical meaning. The higher the cumulative value, the "larger" the neighborhood mean in more directions, and the more obvious the "bulging / brightness" trend of the texture. Multi-directional coverage, taking into account global texture trends; It has low feature dimensionality, making it easy to process later.

[0069] In some examples, the specific methods for analyzing texture features using comparison are as follows: Calculate the texture features of the standard samples; Calculate the similarity between the texture features of the standard sample and the texture features of the surrounding image data to obtain the similarity results, which include similar and dissimilar results; The similarity result of an image dataset is the ratio of dissimilar surrounding image data to all surrounding image data. Features are classified based on their ratios, and these features include normal features and abnormal features.

[0070] Here, texture features are obtained using gray-level co-occurrence matrix or gray-level difference statistics. Specifically, the results obtained by gray-level co-occurrence matrix include energy, contrast, correlation, and entropy, while the results obtained by gray-level difference statistics include mean, variance, entropy, energy, and contrast.

[0071] For gray-level co-occurrence matrix or gray-level difference statistics, one is usually chosen to be used, and any one or several of the corresponding results are selected for use (multiple results need to be weighted and calculated).

[0072] The similarity between the texture features of a standard sample and the texture features of surrounding image data can be calculated directly using numerical comparison. For example, an absolute difference of less than 10% indicates small differences in texture features and a similarity result, while an absolute difference of ≥10% indicates large differences in texture features and a dissimilarity result.

[0073] After completing the above steps, the similarity results of an image dataset are calculated as the ratio of dissimilar surrounding image data to all surrounding image data. Finally, the features are classified based on the ratio, including normal features and abnormal features.

[0074] The ratio here refers to a specific numerical value that needs to be compared with a set value. When the value is less than the set value, the feature is a normal feature; when the value is greater than the set value, the feature is an abnormal feature.

[0075] This application also provides a control device for a deep well pump, comprising: The data acquisition unit is used to continuously acquire sound data collected by the vibration sensor of the deep well pump and segment the sound data in the time dimension to obtain sound analysis data; The data processing unit is used to filter and reduce noise in the sound analysis data to obtain basic analysis data. The data conversion unit is used to convert basic analysis data into image data, which is a grayscale image. The first feature acquisition unit is used to obtain the features included in the image data, including peak points and peak regions. The second feature acquisition unit is used to acquire the surrounding image data of the feature and determine the texture features of the surrounding image data; The classification unit is used to analyze texture features using a comparative method and classify the features using the analysis results. The features include normal features and abnormal features. The control unit is used to monitor the occurrence time and / or frequency of abnormal features and adjust the operating status of the deep well pump based on the monitoring results.

[0076] Furthermore, when analyzing texture features using a comparison method, comparison samples that match the operating status parameters of the deep well pump are selected. The operating parameters of a deep well pump include operating voltage, operating current, active power, bearing temperature, motor temperature, motor speed, inlet water pressure, outlet water pressure, and flow rate. Select at least one deep well pump's operating status parameters to choose a comparison sample.

[0077] Furthermore, converting basic analytical data into image data includes: The basic analysis data is decomposed in the frequency domain to obtain multiple sound curves, and any two sound curves have different frequencies. Set data acquisition points at intervals along the time dimension and determine the amplitude points on the sound curve corresponding to the data acquisition points; The data acquisition points and their corresponding amplitude points are transferred into a coordinate system for representation to obtain image data.

[0078] Furthermore, the features included in the image data are as follows: Generate multiple numerically continuous grayscale ranges; Image data is extracted using grayscale ranges to obtain the extracted content, with only one grayscale range used for each extraction. The extracted content is filtered to obtain the features included in the image data; The filtering process includes removing areas in the extracted content that are greater than or equal to a set area value.

[0079] Furthermore, the features included in the image data are as follows: The image data is divided into multiple rectangular regions, each containing the same number of pixels; Calculate the average pixel value for each rectangular region; The average pixel values ​​of two adjacent rectangular regions are compared in a fixed direction, and a value is assigned based on the comparison result, including zero and one. The assignment results for each rectangular region are statistically analyzed, and the features included in the image data are obtained based on the assignment results. The fixed directions include up, down, left, right, upper left, upper right, lower left, and lower right.

[0080] Furthermore, acquiring the surrounding image data of the features and determining the texture features of the surrounding image data includes: Determine the center point of the feature; A selection region is created based on the center point of the feature, and the pixels within the selection region are the surrounding image data of the feature. The selected area is divided into multiple rectangular regions, each containing the same number of pixels; Calculate the average pixel value for each rectangular region; The average pixel values ​​of two adjacent rectangular regions are compared in a fixed direction, and a value is assigned based on the comparison result, including zero and one. The assignment results for each rectangular region are statistically analyzed, and the features included in the image data are obtained based on the assignment results. The fixed directions include up, down, left, right, upper left, upper right, lower left, and lower right.

[0081] Furthermore, the analysis of texture features using comparative methods includes: Calculate the texture features of the standard samples; Calculate the similarity between the texture features of the standard sample and the texture features of the surrounding image data to obtain the similarity results, which include similar and dissimilar results; The similarity result of an image dataset is the ratio of dissimilar surrounding image data to all surrounding image data. Features are classified based on their ratios, and these features include normal features and abnormal features.

[0082] In one example, the unit in any of the above devices may be one or more integrated circuits configured to implement the above methods, such as one or more application-specific integrated circuits (ASICs), or one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs), or a combination of at least two of these integrated circuit forms.

[0083] For example, when the units in the device can be implemented through a processing element scheduler, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor capable of calling programs. Alternatively, these units can be integrated together to form a system-on-a-chip (SOC).

[0084] In this application, various objects such as messages / information / devices / network elements / systems / apparatus / actions / operations / processes / concepts may be named. It is understood that these specific names do not constitute a limitation on the relevant objects. The names may be changed depending on the scenario, context, or usage habits. The understanding of the technical meaning of the technical terms in this application should be mainly determined from their functions and technical effects embodied / performed in the technical solution.

[0085] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0086] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0087] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0088] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0089] It should also be understood that in the various embodiments of this application, "first," "second," etc., are merely used to indicate that multiple objects are different. For example, a first time window and a second time window are only used to indicate different time windows. They should not have any effect on the time windows themselves, and the aforementioned "first," "second," etc., should not impose any limitations on the embodiments of this application.

[0090] It should also be understood that, in the various embodiments of this application, unless otherwise specified or in case of logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced by each other, and the technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationships.

[0091] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned computer-readable storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0092] This application also provides a deep well pump, the deep well pump comprising: One or more memories for storing instructions; and One or more processors are configured to retrieve and execute the instructions from the memory, performing the methods described above.

[0093] This application also provides a computer program product including instructions that, when executed, cause the terminal device and the network device to perform operations corresponding to the methods described above.

[0094] This application also provides a chip system including a processor for implementing the functions involved in the above description, such as generating, receiving, transmitting, or processing the data and / or information involved in the above methods.

[0095] This chip system can consist of chips or include chips and other discrete components.

[0096] The processor mentioned above can be a CPU, a microprocessor, an ASIC, or one or more integrated circuits that execute a program to control the method of transmitting the feedback information described above.

[0097] In one possible design, the chip system also includes a memory for storing necessary program instructions and data. The processor and the memory can be decoupled and located on different devices, connected via wired or wireless means to support the chip system in implementing the various functions described in the above embodiments. Alternatively, the processor and the memory can also be coupled to the same device.

[0098] Optionally, the computer instructions are stored in memory.

[0099] Optionally, the memory can be a storage unit within the chip, such as a register or cache. Alternatively, the memory can be a storage unit located outside the chip within the terminal, such as a ROM or other types of static storage devices that can store static information and instructions, such as RAM.

[0100] It is understood that the memory in this application may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory.

[0101] Non-volatile memory can be ROM, programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory.

[0102] Volatile memory can be RAM, which is used as an external cache. There are many different types of RAM, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus random access memory.

[0103] The embodiments described in this specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A control method for a deep well pump, characterized in that, include: Continuously acquire sound data from the vibration sensors at the deep well pump headquarters and segment the sound data in the time dimension to obtain sound analysis data; The sound analysis data is filtered and noise-reduced to obtain the basic analysis data; The basic analysis data is converted into image data, which is a grayscale image. The features included in the image data are obtained, including peak points and peak regions. Acquire the surrounding image data of the feature and determine the texture features of the surrounding image data; Texture features are analyzed using a comparative method, and the analysis results are used to classify the features, including normal features and abnormal features. Monitor the occurrence time and / or frequency of abnormal features and adjust the operating status of the deep well pump based on the monitoring results.

2. The control method for a deep well pump according to claim 1, characterized in that, When analyzing texture features using a comparison method, a comparison sample matching the operating status parameters of the deep well pump is selected. The operating parameters of a deep well pump include operating voltage, operating current, active power, bearing temperature, motor temperature, motor speed, inlet water pressure, outlet water pressure, and flow rate. Select at least one deep well pump's operating status parameters to choose a comparison sample.

3. The control method for a deep well pump according to claim 1, characterized in that, Converting basic analytical data into image data includes: The basic analysis data is decomposed in the frequency domain to obtain multiple sound curves, and any two sound curves have different frequencies. Set data acquisition points at intervals along the time dimension and determine the amplitude points on the sound curve corresponding to the data acquisition points; The data acquisition points and their corresponding amplitude points are transferred into a coordinate system for representation to obtain image data.

4. The control method for a deep well pump according to claim 1 or 3, characterized in that, The features included in the obtained image data include: Generate multiple numerically continuous grayscale ranges; Image data is extracted using grayscale ranges to obtain the extracted content, with only one grayscale range used for each extraction. The extracted content is filtered to obtain the features included in the image data; The filtering process includes removing areas in the extracted content that are greater than or equal to a set area value.

5. The control method for a deep well pump according to claim 1 or 3, characterized in that, The features included in the obtained image data include: The image data is divided into multiple rectangular regions, each containing the same number of pixels; Calculate the average pixel value for each rectangular region; The average pixel values ​​of two adjacent rectangular regions are compared in a fixed direction, and a value is assigned based on the comparison result, including zero and one. The assignment results for each rectangular region are statistically analyzed, and the features included in the image data are obtained based on the assignment results. The fixed directions include up, down, left, right, upper left, upper right, lower left, and lower right.

6. The control method for a deep well pump according to claim 1, characterized in that, Acquiring and determining the texture features of the surrounding image data includes: Determine the center point of the feature; A selection region is created based on the center point of the feature, and the pixels within the selection region are the surrounding image data of the feature. The selected area is divided into multiple rectangular regions, each containing the same number of pixels; Calculate the average pixel value for each rectangular region; The average pixel values ​​of two adjacent rectangular regions are compared in a fixed direction, and a value is assigned based on the comparison result, including zero and one. The assignment results for each rectangular region are statistically analyzed, and the assignment results are used as texture features of the surrounding image data. The fixed directions include up, down, left, right, upper left, upper right, lower left, and lower right.

7. The control method for a deep well pump according to claim 6, characterized in that, Analysis of texture features using comparative methods includes: Calculate the texture features of the standard samples; Calculate the similarity between the texture features of the standard sample and the texture features of the surrounding image data to obtain the similarity results, which include similar and dissimilar results; The similarity result of an image dataset is the ratio of dissimilar surrounding image data to all surrounding image data. Features are classified based on their ratios, and these features include normal features and abnormal features.

8. A control device for a deep well pump, characterized in that, include: The data acquisition unit is used to continuously acquire sound data collected by the vibration sensor of the deep well pump and segment the sound data in the time dimension to obtain sound analysis data; The data processing unit is used to filter and reduce noise in the sound analysis data to obtain basic analysis data. The data conversion unit is used to convert basic analysis data into image data, which is a grayscale image. The first feature acquisition unit is used to obtain the features included in the image data, including peak points and peak regions. The second feature acquisition unit is used to acquire the surrounding image data of the feature and determine the texture features of the surrounding image data; The classification unit is used to analyze texture features using a comparative method and classify the features using the analysis results. The features include normal features and abnormal features. The control unit is used to monitor the occurrence time and / or frequency of abnormal features and adjust the operating status of the deep well pump based on the monitoring results.

9. A deep well pump, characterized in that, The deep well pump includes: One or more memories for storing instructions; and One or more processors are configured to retrieve and execute the instructions from the memory to perform the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes: The program, when run by the processor, executes the method as described in any one of claims 1 to 7.

Images