Slurry pump control method and related equipment

Abstract

The invention discloses a slurry pump control method and related equipment, and relates to the technical field of seawater desalination, and the method comprises the following steps: obtaining sensor data of a slurry pump; data analysis is conducted on the sensor data, and the working condition abnormal index of the slurry pump is determined; determining a target control instruction corresponding to the working condition abnormal index based on the working condition abnormal index and a preset threshold interval; and based on the target control instruction, the slurry pump is controlled to execute corresponding target regulation and control operation. Real-time monitoring and intelligent judgment of the operation state of the slurry pump are achieved, early warning can be triggered at the initial stage of equipment abnormity, the operation parameters can be adjusted in a self-adaptive mode, the response delay problem of a traditional single control mode is effectively avoided, and the operation reliability and safety of the equipment are improved.

Description

Technical Field

[0001] The present application relates to the technical field of seawater desalination, and in particular to a control method for a mud pump and related equipment. Background Art

[0002] In the desalination process, mud pumps are critical equipment for treating waste from drainage ponds, and their operating status directly impacts production efficiency and system stability. Traditional control methods rely primarily on manual monitoring and fixed threshold adjustments, lacking dynamic response capabilities. This makes mud pumps susceptible to response lags or misjudgments under complex operating conditions. For example, when mud concentration suddenly changes or a pipeline becomes clogged, delayed manual intervention can cause equipment overload, abnormal wear, or even downtime, impacting production continuity while increasing safety risks and maintenance costs. Therefore, a mud pump control method is urgently needed to address the aforementioned technical issues. Summary of the Invention

[0003] The Summary of the Invention introduces a series of simplified concepts that will be further described in the Detailed Description of the Invention. The Summary of the Invention of this application is not intended to limit the key features and essential technical features of the claimed technical solution, nor is it intended to determine the scope of protection of the claimed technical solution.

[0004] In a first aspect, the present application provides a method for controlling a mud pump, comprising:

[0005] Get sensor data from the mud pump;

[0006] Analyze sensor data to determine abnormal operating index of mud pump;

[0007] Determining a target control instruction corresponding to the abnormal operating condition index based on the abnormal operating condition index and a preset threshold range;

[0008] Based on the target control instruction, the mud pump is controlled to perform the corresponding target control operation.

[0009] In some embodiments, the sensor data includes temperature data, current data, pressure data, liquid level data, and flow data. Acquiring the sensor data of the mud pump includes:

[0010] Obtain temperature data through the temperature sensor deployed on the mud pump body;

[0011] Obtain pressure data through a pressure sensor installed in the mud pump inlet pipe;

[0012] Obtain flow data through a flow meter installed on the mud pump outlet pipe;

[0013] The liquid level data is obtained through the liquid level sensor arranged inside the drainage tank;

[0014] The current data is obtained through the current transformer connected to the mud pump motor.

[0015] In some embodiments, performing data analysis on sensor data to determine an abnormal operating condition index of a mud pump includes:

[0016] Preprocessing the sensor data to generate standardized sensor data, wherein the preprocessing includes filtering and normalization operations;

[0017] Determine the operating condition abnormality index based on the preset fault database and standardized sensor data.

[0018] In some embodiments, the preset fault database includes normal state historical data and fault state characteristic data. Determining the abnormal operating condition index based on the preset fault database and the standardized sensor data includes:

[0019] Calculating the deviation between the standardized sensor data and the normal state historical data to obtain a first deviation value, wherein the normal state historical data includes a reference range of the sensor data when the mud pump is operating without fault;

[0020] Performing similarity matching on the standardized sensor data and the fault state characteristic data to obtain a second matching value, wherein the fault state characteristic data includes a sensor data characteristic pattern corresponding to a preset fault type;

[0021] Based on the dynamic weight distribution of the first deviation value and the second matching value, an operating condition abnormality index is determined.

[0022] In some embodiments, the preset threshold interval includes a first threshold interval, a second threshold interval, and a third threshold interval, the maximum value of the first threshold interval is less than or equal to the minimum value of the second threshold interval, the maximum value of the second threshold interval is less than or equal to the minimum value of the third threshold interval, the target control instruction includes a mud pump start instruction, a mud pump speed adjustment instruction, or a mud pump shutdown instruction, and determining the target control instruction corresponding to the abnormal operating condition index based on the abnormal operating condition index and the preset threshold interval includes:

[0023] When the abnormal operating condition index is within the first threshold range, a mud pump start instruction is generated; or,

[0024] When the abnormal operating condition index is within the second threshold range, a mud pump speed adjustment instruction is generated, the mud pump speed adjustment instruction including a speed adjustment magnitude and a corresponding action duration; or

[0025] When the abnormal operating condition index is in the third threshold range, a mud pump shutdown instruction is generated.

[0026] In some embodiments, based on the target control instruction, controlling the mud pump to perform the corresponding target control operation includes:

[0027] When the target control instruction is a mud pump start instruction, the starting power parameter of the mud pump is determined based on the liquid level data, and the mud pump is controlled to run at a speed corresponding to the starting power parameter; or

[0028] When the target control instruction is a mud pump speed adjustment instruction, the speed adjustment magnitude compensation coefficient is determined based on the current data and the pressure data; the frequency adjustment parameter of the frequency converter is generated according to the speed adjustment magnitude, the action duration and the compensation coefficient; the output speed of the mud pump motor is adjusted based on the frequency adjustment parameter of the frequency converter; or

[0029] When the target control instruction is a mud pump shutdown instruction, the shutdown cooling strategy is determined based on the temperature data; according to the shutdown cooling strategy, the valve closing sequence and motor deceleration curve are generated; based on the valve closing sequence, the mud conveying channel is cut off; and based on the motor deceleration curve, the shutdown operation is executed.

[0030] In some embodiments, before performing data analysis on the sensor data to determine the abnormal working condition index of the mud pump, the method further includes: acquiring flow state image data of the mud pump;

[0031] The sensor data is analyzed to determine the abnormal working condition index of the mud pump, including: performing multimodal fusion analysis on the flow image data and the sensor data to determine the abnormal working condition index of the mud pump.

[0032] In a second aspect, the present application proposes a control device for a mud pump, comprising:

[0033] A mud pump monitoring data acquisition unit, used to acquire sensor data of the mud pump;

[0034] An abnormal working condition index determination unit is used to analyze sensor data and determine the abnormal working condition index of the mud pump;

[0035] a target control instruction generating unit, which determines a target control instruction corresponding to the abnormal operating condition index based on the abnormal operating condition index and a preset threshold range;

[0036] The mud pump operation control unit controls the mud pump to perform corresponding target control operations based on the target control instructions.

[0037] In a third aspect, an electronic device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is configured to implement the steps of the mud pump control method of any one of the first aspects when executing the computer program stored in the memory.

[0038] In a fourth aspect, the present application proposes a computer-readable storage medium storing a computer program, which implements the control method of the mud pump according to any one of the first aspects when the computer program is executed by a processor.

[0039] In summary, this application obtains multi-dimensional sensor data of the mud pump, combines it with a preset fault database for dynamic data analysis, accurately calculates the abnormal working condition index, and generates hierarchical control instructions based on the threshold range. This application realizes real-time monitoring and intelligent judgment of the operating status of the mud pump, and can trigger early warnings and adaptively adjust operating parameters at the early stage of equipment abnormality. Through a hierarchical control strategy (startup, speed adjustment, shutdown), the response delay problem of the traditional single control mode is effectively avoided, and the reliability and safety of equipment operation are improved. At the same time, this application ensures that the mud pump is always in the best working state of high efficiency and low consumption, reduces energy waste and equipment wear, and extends service life, providing a high-reliability, low-cost mud pump control solution for industrial scenarios such as seawater desalination.

[0040] The control method of the mud pump proposed in this application, and other advantages, objectives and features of this application will be partially reflected in the following description, and will also be partially understood by those skilled in the art through research and practice of this application. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present description. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0042] Figure 1 A schematic flow chart of a control method for a mud pump provided in an embodiment of the present application;

[0043] Figure 2 A schematic structural diagram of a control device for a mud pump provided in an embodiment of the present application;

[0044] Figure 3 A schematic diagram of the control electronic device structure of a mud pump provided in an embodiment of the present application. DETAILED DESCRIPTION

[0045] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or that are inherent to these processes, methods, products or devices. 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 embodiments described are only part of the embodiments of the present application, not all of the embodiments.

[0046] See also Figure 1 , which is a flow chart of a control method for a mud pump provided in an embodiment of the present application, may specifically include:

[0047] S110, acquiring sensor data of a mud pump;

[0048] This step, illustratively, forms the foundation of the control method. Its goal is to collect multi-dimensional physical data reflecting the equipment's operating status in real time through a variety of sensors deployed at key locations on the mud pump. This data, including parameters such as temperature, current, pressure, liquid level, and flow rate, covers the mechanical, electrical, and fluid dynamic characteristics of the mud pump, providing the raw input for subsequent anomaly detection and control decisions. The coordinated collection of multi-source data comprehensively captures subtle changes in the equipment under varying operating conditions, avoiding the limitations of a single sensor and ensuring the comprehensiveness and accuracy of subsequent analysis.

[0049] Furthermore, the acquisition of sensor data relies on the deployment of sensors throughout the mud pump and its associated components. For example, temperature sensors monitor pump temperature changes, pressure and flow sensors collect fluid characteristics at the pipeline inlet and outlet, respectively, level sensors provide feedback on the liquid level in the drainage tank, and current transformers record the motor load status. This data is transmitted to the data processing module via a standardized interface, providing a unified input format for subsequent preprocessing and anomaly index calculation. The reliability and real-time performance of these steps directly impact the response efficiency and judgment accuracy of the entire control system and are a key prerequisite for ensuring the intelligent operation of the mud pump.

[0050] S120, analyzing the sensor data to determine an abnormal operating condition index of the mud pump;

[0051] Exemplarily, this step establishes a quantitative evaluation index for the mud pump's operating status by standardizing and dynamically analyzing multi-source sensor data. Specifically, after pre-processing operations such as filtering and normalization, the sensor data generates standardized sensor data to eliminate the influence of environmental noise and dimensional differences on the analysis results. Subsequently, based on the historical operating data in the preset fault database (including normal state benchmarks and typical fault mode characteristics), the real-time data is compared with the patterns in the database through deviation calculation and similarity matching algorithms. Through a dynamic weight allocation strategy, the degree of deviation of the real-time data from the normal state and the degree of matching with the fault characteristics are comprehensively evaluated, and finally an operating condition abnormality index characterizing the healthy operating status of the mud pump is generated.

[0052] Furthermore, the calculation process for the operating condition anomaly index incorporates correlation analysis of multi-dimensional data, ensuring comprehensive and robust anomaly detection. By dividing the index into preset threshold intervals, the continuous numerical value of the operating condition anomaly index is mapped to discrete control instruction trigger levels, providing a decision-making basis for subsequent hierarchical control. This process avoids the limitations of single-threshold determination, dynamically adapts to the operating characteristics of the mud pump under different operating conditions, and improves the sensitivity and accuracy of anomaly detection.

[0053] S130, determining a target control instruction corresponding to the abnormal operating condition index based on the abnormal operating condition index and a preset threshold range;

[0054] Exemplarily, the core logic of this step is to quantitatively analyze the real-time operating status of the mud pump, match the dynamically calculated operating condition abnormality index with the preset threshold interval, and thus trigger the hierarchical control instruction. Specifically, the operating condition abnormality index is a comprehensive indicator generated by preprocessing multi-source sensor data and comparing it with the fault database, reflecting the degree to which the mud pump deviates from the normal operating mode. The preset threshold interval is divided into three levels, and each interval corresponds to different equipment control requirements: when the abnormal index falls into the first threshold interval, it indicates that the mud pump is in a low-risk state or waiting to be started, and a start-up instruction is generated to activate the equipment; when entering the second threshold interval, it indicates that there is an adjustable deviation in the operating parameters, triggering a speed adjustment instruction to dynamically optimize the operating efficiency; if the abnormal index exceeds the second threshold interval and enters the third threshold interval, it is determined to be a high-risk fault, and a shutdown instruction is immediately generated to ensure equipment safety.

[0055] Furthermore, the above steps achieve a closed-loop response from state monitoring to control decision-making by mapping continuous abnormality indices to discrete control instruction levels, ensuring that abnormal conditions of different severities can trigger adaptive control actions. This not only avoids the limitations of single threshold judgment, but also provides clear instruction input for subsequent execution modules, thereby achieving an optimal balance between energy consumption and efficiency while ensuring the reliability of mud pump operation.

[0056] S140 : Based on the target control instruction, control the mud pump to perform a corresponding target control operation.

[0057] Exemplarily, this step generates an adaptive hierarchical control instruction by matching the dynamically calculated operating condition abnormality index with the preset threshold interval, and drives the mud pump to perform the corresponding control operation. Specifically, when the operating condition abnormality index falls into different threshold intervals, the start-up, speed adjustment or shutdown instruction is triggered according to the preset rules, and the specific control parameters are generated in combination with the real-time sensor data to achieve fine-grained adjustment of the mud pump operating state. For example, the start-up instruction dynamically sets the initial speed based on the liquid level data, the speed adjustment instruction corrects the adjustment level through current and pressure feedback, and the shutdown instruction formulates a step-by-step cooling strategy according to temperature changes. The above operations ensure the adaptability of the control instructions to the real-time operating conditions through a closed-loop feedback mechanism, avoiding the response lag problem of traditional single threshold control.

[0058] Furthermore, this step directly links the severity of abnormal conditions with the intensity of the executed action through a hierarchical control strategy. This not only optimizes operating parameters to maintain efficient drainage in the event of minor anomalies, but also quickly isolates the source of risk in the event of a serious fault, ensuring equipment safety. The execution logic of each control instruction strictly relies on the input data of the previous step, ensuring the continuity of the technical chain. Parametric control also enables quantification and traceability of actions, providing a technical foundation for intelligent operation and maintenance of mud pumps.

[0059] In summary, the embodiment of the present application monitors the operating status of the mud pump in real time and accurately calculates the abnormal working condition index through multi-dimensional sensor data collection and intelligent analysis, and implements a hierarchical control strategy in combination with preset threshold intervals to improve the reliability and safety of equipment operation. In the data collection stage, multi-source sensor data such as temperature, current, pressure, liquid level and flow are integrated to fully cover the mechanical, electrical and fluid dynamic characteristics of the mud pump, ensuring the comprehensiveness and accuracy of monitoring. In the data analysis stage, a dynamic weight distribution strategy is adopted to match the deviation and similarity of real-time data with the historical normal state and typical fault characteristics in the fault database to generate a comprehensive abnormal working condition index, effectively avoiding the limitations of single parameter judgment and improving the sensitivity and accuracy of abnormality detection. Based on the three-level threshold interval divided by the abnormal working condition index, the start-up, speed adjustment or shutdown instructions are adaptively generated to achieve hierarchical control: when the risk is low, the starting power is dynamically adjusted according to the liquid level; when the risk is medium, the speed adjustment magnitude is compensated by combining current and pressure feedback; when the risk is high, a step-by-step shutdown cooling strategy is formulated based on the temperature gradient. By dynamically optimizing operating parameters through a closed-loop feedback mechanism, the mud pump is always in the optimal operating condition with high efficiency and low energy consumption, reducing energy waste and equipment wear, and extending its service life. The embodiments of this application are particularly suitable for complex industrial scenarios such as seawater desalination. They can effectively address challenges such as the crystallization and corrosiveness of high-salinity mud, reduce the need for manual intervention, lower maintenance costs and safety risks, and provide a highly reliable, low-cost intelligent solution for industrial pump control systems.

[0060] In some examples, the sensor data includes temperature data, current data, pressure data, liquid level data, and flow data. Obtaining the sensor data of the mud pump includes:

[0061] Obtain temperature data through the temperature sensor deployed on the mud pump body;

[0062] Obtain pressure data through a pressure sensor installed in the mud pump inlet pipe;

[0063] Obtain flow data through a flow meter installed on the mud pump outlet pipe;

[0064] The liquid level data is obtained through the liquid level sensor arranged inside the drainage tank;

[0065] The current data is obtained through the current transformer connected to the mud pump motor.

[0066] For example, by deploying temperature sensors at key locations on the mud pump body, the temperature data of the pump body can be monitored in real time. The temperature sensor uses a contact measurement method and is directly installed on heat-sensitive areas such as the mechanical seal of the pump body casing, the bearing seat, and the motor connection flange to ensure accurate perception of the mechanical friction, lubrication status, and temperature rise of the motor winding inside the pump body. The temperature data is transmitted to the central processing unit in the form of an analog signal or a digital signal, and the sampling frequency is set to 10Hz to meet the needs of real-time monitoring. Under high-salinity conditions in seawater desalination, the sensor housing is made of corrosion-resistant alloy and is sealed to prevent electrolyte penetration, ensuring long-term stable operation. The acquired temperature data is used to judge the risk of failures such as pump overheating and motor overload, providing key input for the calculation of subsequent abnormal operating condition indexes.

[0067] Real-time motor current data is collected through a current transformer connected to the mud pump motor circuit. The current transformer uses a closed-loop Hall effect principle, with a rated current of 1.2 times the motor's nameplate current, a bandwidth of 0-5kHz, and an accuracy level of 0.2S. The current signal is converted into a 0-5V voltage signal by a signal conditioning circuit and then input into the data acquisition module. Three-phase current data is calculated using true RMS values to obtain a comprehensive current value, which is used to analyze the motor load status. Continuous current exceeding the limit indicates mechanical obstruction, while periodic fluctuations may be caused by impeller imbalance. The current data is matched with the current characteristic patterns in the preset fault database, providing key input for the dynamic weighting of the operating condition abnormality index.

[0068] A pressure sensor installed in the mud pump inlet pipe collects pressure data from the inlet in real time. The pressure sensor uses a piezoresistive or piezoelectric sensing element and is installed in the upstream straight pipe section of the inlet pipe 1.5 times the diameter of the pump body to avoid the influence of fluid turbulence on measurement accuracy. The sensor has a measuring range of 0-5MPa, an accuracy level of 0.5, and outputs a 4-20mA standard current signal. Real-time pressure data reflects changes in the flow resistance of the pumped medium and is used to detect abnormal operating conditions such as pipeline blockage and sudden changes in mud concentration. During the data preprocessing stage, the pressure signal is filtered through a sliding average filter to eliminate instantaneous fluctuation interference. After normalization to the range of 0-1, it is used to calculate the anomaly index to ensure the consistency of data of different dimensions.

[0069] Liquid level data is collected via a level sensor installed inside the drainage tank. The level sensor utilizes ultrasonic or radar ranging principles. Mounted on the tank wall 0.5 m from the bottom, it employs non-contact measurement with a range of 0-5 m and a resolution of 1 mm. Ultrasonic pulses transmit at a frequency of 20 Hz, calculating the liquid level based on the echo time and automatically compensating for the effects of temperature on sound velocity. This level data is used to trigger the start and stop of the slurry pump. The rate of change of the liquid level also contributes to the calculation of the abnormal operating condition index, which is used to identify any abnormalities in the drainage tank or pumping efficiency.

[0070] An electromagnetic flowmeter installed in the mud pump outlet pipe collects outlet flow data. Based on the Faraday principle of electromagnetic induction, the flowmeter is installed in a vertical section of the outlet pipe to ensure full flow. The measuring pipe diameter matches the pipe specifications. The lining material is polytetrafluoroethylene to resist seawater corrosion, and the electrode material is Hastelloy C276. The flow data collection frequency is 5Hz, and the range is 0-500m. 3 / h, with a linearity error of ≤±0.3%. Real-time flow values are analyzed collaboratively with inlet pressure data to calculate pump efficiency changes and identify fault modes such as impeller wear and cavitation. Flow data is de-noised using a Kalman filter and dynamically compared with historical normal flow curves in a pre-set fault database.

[0071] In summary, the above system, through the deployment of temperature sensors, current transformers, inlet pressure sensors, drainage tank level sensors, and outlet flow meters at key locations on the mud pump, acquires multi-dimensional sensor data in real time, enabling comprehensive monitoring of the mechanical, electrical, and fluid characteristics of the mud pump. The temperature sensor, made of corrosion-resistant alloy, accurately senses pump friction and motor temperature rise; the current transformer analyzes load anomalies through true RMS calculations; the pressure sensor, combined with a sliding average filter, eliminates interference and detects changes in pipeline resistance; the level sensor triggers start-stop control based on ultrasonic ranging; and the corrosion-resistant design of the electromagnetic flowmeter ensures high-precision flow monitoring.

[0072] In some examples, data analysis is performed on sensor data to determine abnormal operating condition indexes of mud pumps, including:

[0073] Preprocessing the sensor data to generate standardized sensor data, wherein the preprocessing includes filtering and normalization operations;

[0074] The preset fault database includes normal state historical data and fault state characteristic data. Based on the preset fault database and standardized sensor data, the abnormal operating condition index is determined, including:

[0075] Calculating the deviation between the standardized sensor data and the normal state historical data to obtain a first deviation value, wherein the normal state historical data includes a reference range of the sensor data when the mud pump is operating without fault;

[0076] Performing similarity matching on the standardized sensor data and the fault state characteristic data to obtain a second matching value, wherein the fault state characteristic data includes a sensor data characteristic pattern corresponding to a preset fault type;

[0077] Based on the dynamic weight distribution of the first deviation value and the second matching value, an operating condition abnormality index is determined.

[0078] For example, the sensor data of the mud pump is preprocessed to generate standardized sensor data. The preprocessing includes filtering and normalization operations. The filtering operation uses a sliding average filtering algorithm to eliminate transient noise caused by electromagnetic interference or mechanical vibration during the sensor acquisition process. Specifically, time windows are set for temperature, current, pressure, liquid level and flow data respectively, and the length of the time window is dynamically adjusted according to the sensor type: a 60-second window is used for temperature data, a 10-second window is used for pressure and flow data, a 5-second window is used for current data, and a 30-second window is used for liquid level data. The average value of the data in each time window is calculated by sliding average to retain the trend change characteristics. The normalization operation uses the minimum-maximum normalization method to map each sensor data to a numerical range of 0-1. Specifically, based on the maximum and minimum values of each sensor in the historical data, a normalization parameter table is established, and the real-time data is scaled proportionally to eliminate the interference of dimensional differences on subsequent analysis.

[0079] The deviation between the standardized sensor data and the normal state historical data in the preset fault database is calculated to obtain the first deviation value. The normal state historical data is the reference range of sensor data collected when the mud pump is operating without fault, including the statistical characteristic values (mean, standard deviation, extreme value) of temperature, current, pressure, liquid level and flow parameters. The deviation calculation is achieved in the following way, that is, for each sensor parameter, the absolute difference between its real-time data and the corresponding reference mean is calculated, and then divided by the reference standard deviation to obtain the standardized deviation component; the standardized deviation components of all sensors are weighted and summed to generate the first deviation value. The weight is preset according to the importance of the parameter, for example, the temperature weight is 0.3, the current weight is 0.25, the pressure weight is 0.2, the liquid level weight is 0.15, and the flow weight is 0.1. The larger the first deviation value, the greater the degree of deviation between the current operating state and the normal state.

[0080] The standardized sensor data is matched against the fault state feature data in the preset fault database for similarity to obtain a second matching value. The fault state feature data is the characteristic pattern of the sensor data corresponding to a preset fault type (such as motor overheating, pipe blockage, or impeller wear). For example, the characteristic pattern of motor overheating is a continuous temperature increase and current exceeding the limit, while the characteristic pattern of pipe blockage is a sudden drop in inlet pressure and a decrease in flow rate. The fault feature pattern with the highest correlation to the current sensor data is retrieved from the fault database. For example, if both the current temperature and current are abnormally elevated, the motor overheating pattern is prioritized. Using the cosine similarity algorithm, the real-time standardized data is compared to the fault feature pattern vector for similarity calculation. For example, the feature vector of the motor overheating pattern is [temperature 0.9, current 0.8, pressure 0.3, flow 0.5], while the real-time data vector is [temperature 0.85, current 0.75, pressure 0.4, flow 0.6]. The cosine value of the angle between the two is calculated as the similarity score to obtain the second matching value. A higher second matching value indicates that the current state is closer to a specific fault mode.

[0081] The final operating condition abnormality index is determined based on the dynamic weighting of the first deviation value and the second matching value. The dynamic weighting strategy adaptively adjusts according to the mud pump's operating stage. During stable operation, the weight of the first deviation value is set to 0.7, and the weight of the second matching value is set to 0.3. During the startup or shutdown transition phase, the weight of the first deviation value is adjusted to 0.5, and the weight of the second matching value is adjusted to 0.5. The operating condition abnormality index is calculated as follows: Abnormality Index = First Deviation Value × Weight A + Second Matching Value × Weight B, where the sum of Weight A and Weight B is 1.0. The resulting operating condition abnormality index is a continuous value ranging from 0 to 1, with larger values indicating a higher risk of abnormal mud pump operation. For example, when the first deviation value is 0.6, the second matching value is 0.8, and Weight A = 0.7, the abnormality index is 0.6 × 0.7 + 0.8 × 0.3 = 0.66. The continuous numerical value of the abnormal working condition index is mapped to a preset threshold range (e.g., 0-0.3 is the first threshold range, 0.3-0.7 is the second threshold range, and 0.7-1.0 is the third threshold range). The current state level is determined based on the preset threshold range and the corresponding control instruction is triggered. This provides a quantitative basis for the subsequent control instruction generation.

[0082] In summary, the above steps transform multi-source sensor data into an abnormality index representing the mud pump's operating status through data preprocessing, deviation calculation, similarity matching, and dynamic weight assignment. Preprocessing ensures data consistency and comparability; deviation calculation reflects the degree of deviation between real-time data and a normal baseline; similarity matching identifies potential failure modes; and dynamic weight assignment comprehensively assesses equipment health. Through interactive analysis of a pre-set fault database and real-time data, comprehensive and robust anomaly detection is achieved, providing a precise basis for subsequent hierarchical control strategies.

[0083] In some examples, the preset threshold interval includes a first threshold interval, a second threshold interval, and a third threshold interval, the maximum value of the first threshold interval is less than or equal to the minimum value of the second threshold interval, the maximum value of the second threshold interval is less than or equal to the minimum value of the third threshold interval, the target control instruction includes a mud pump start instruction, a mud pump speed adjustment instruction, or a mud pump shutdown instruction, and based on the abnormal operating condition index and the preset threshold interval, determining the target control instruction corresponding to the abnormal operating condition index includes:

[0084] When the abnormal operating condition index is within the first threshold range, a mud pump start instruction is generated; or,

[0085] When the abnormal operating condition index is within the second threshold range, a mud pump speed adjustment instruction is generated, the mud pump speed adjustment instruction including a speed adjustment magnitude and a corresponding action duration; or

[0086] When the abnormal operating condition index is in the third threshold range, a mud pump shutdown instruction is generated.

[0087] Exemplarily, the preset threshold intervals include a first threshold interval, a second threshold interval, and a third threshold interval, which are divided based on the health and risk level of the mud pump's operating status. The first threshold interval indicates that the mud pump is in a low-risk or waiting-to-start state, with a numerical range of 0 to 0.3 (inclusive) of the operating condition abnormality index; the second threshold interval indicates that the mud pump has an adjustable operating parameter deviation, with a numerical range of 0.3 (inclusive) to 0.7; and the third threshold interval indicates that the mud pump is in a high-risk fault state, with a numerical range of 0.7 (inclusive) to 1.0. The boundary values of the threshold intervals are determined through historical data analysis and fault simulation experiments. For example, under normal operating conditions, the operating condition abnormality index fluctuates between 0 and 0.2. When the index exceeds 0.3, it indicates a potential abnormality; when the index exceeds 0.7, the probability of failure increases and immediate intervention is required. The interval division follows the non-overlapping principle, that is, the maximum value of the first threshold interval is equal to the minimum value of the second threshold interval, and the maximum value of the second threshold interval is equal to the minimum value of the third threshold interval, ensuring that each value uniquely corresponds to a control instruction level.

[0088] When the abnormal working condition index is in the first threshold interval, it indicates that the mud pump is in a low-risk state or meets the starting conditions (for example, the liquid level in the drainage tank reaches the preset upper limit), and a mud pump start-up instruction is generated. The triggering logic of the start-up instruction is based on the matching result of the liquid level data and the preset safe liquid level range. If the liquid level is higher than the starting threshold and there are no other fault characteristics, it is determined to be in a startable state. When the abnormal working condition index falls into the second threshold interval, it indicates that there is a deviation in the operating parameters (such as pressure fluctuation or flow deviation), and a mud pump speed adjustment instruction is generated. The instruction includes the speed adjustment magnitude and the duration of action. The adjustment magnitude is dynamically corrected based on the real-time feedback of the current and pressure data. When the abnormal working condition index enters the third threshold interval, it indicates that there is a serious fault, and a mud pump shutdown instruction is generated. The triggering condition of the shutdown instruction is directly related to the real-time monitoring value of the temperature data. If the temperature exceeds the preset safety threshold or the continuous rise rate is abnormal, it is determined to be a high-risk state and the shutdown protection is triggered immediately.

[0089] In some examples, based on the target control instruction, controlling the mud pump to perform the corresponding target control operation includes:

[0090] When the target control instruction is a mud pump start instruction, the starting power parameter of the mud pump is determined based on the liquid level data, and the mud pump is controlled to run at a speed corresponding to the starting power parameter; or

[0091] When the target control instruction is a mud pump speed adjustment instruction, the speed adjustment magnitude compensation coefficient is determined based on the current data and pressure data; the frequency adjustment parameter of the frequency converter is generated according to the speed adjustment magnitude, action duration, and compensation coefficient; and the output speed of the mud pump motor is adjusted based on the frequency adjustment parameter of the frequency converter.

[0092] When the target control instruction is a mud pump shutdown instruction, the shutdown cooling strategy is determined based on the temperature data; according to the shutdown cooling strategy, the valve closing sequence and motor deceleration curve are generated; based on the valve closing sequence, the mud conveying channel is cut off; and based on the motor deceleration curve, the shutdown operation is executed.

[0093] For example, when the target control instruction is a mud pump start instruction, the starting power parameter of the mud pump is determined based on the real-time liquid level data obtained by the drainage tank liquid level sensor. Specifically, after filtering and normalization, the liquid level data is matched with the preset safe liquid level range. If the liquid level value is between the preset start threshold (for example, 80% of the liquid level upper limit) and the liquid level upper limit, the starting power parameter is dynamically calculated based on the liquid level height and the preset power mapping table. For example, when the liquid level is 80% of the upper limit, the starting power parameter is set to 60% of the rated power; when the liquid level is 95% of the upper limit, the starting power parameter is increased to 90% of the rated power. The starting power parameter is further converted into the target frequency value of the frequency converter (for example, 60% power corresponds to 30Hz). The frequency converter controls the mud pump motor to gradually increase to the set speed according to the target frequency, while monitoring the current and pressure data during the startup process to ensure that there is no instantaneous overload or pressure mutation. During the startup process, the liquid level drop rate is continuously monitored. If the rate is lower than the preset efficiency threshold (such as 0.2 meters per minute), the power parameter is automatically increased in steps of 5% until the liquid level drop rate meets the standard.

[0094] When the target control command is a mud pump speed adjustment command, a compensation coefficient for the speed adjustment level is determined based on real-time data from the current and pressure sensors. Current data reflects the motor load state. If the current value exceeds a preset safety range (e.g., 110% of the rated current), the compensation coefficient reduces the speed increment linearly (e.g., a compensation coefficient of 0.8 for a 10% overrun). Pressure data reflects pipeline resistance. If the inlet pressure falls below a preset lower limit (e.g., 80% of the normal pressure), the compensation coefficient increases the speed increment exponentially (e.g., a compensation coefficient of 1.2 for a 20% pressure drop). The frequency converter frequency adjustment parameters are generated based on the adjustment level (e.g., ±10%), duration (e.g., 30 seconds), and compensation coefficient in the speed adjustment command. For example, if the current frequency is 40 Hz, the adjustment level is +10%, and the compensation coefficient is 0.9, the target frequency is calculated as 40 × (1 + 10% × 0.9) = 43.6 Hz. The frequency converter adjusts the motor output speed according to the target frequency, using a closed-loop control algorithm to adjust the motor output speed in real time and monitor the adjusted current and pressure data. If the abnormal operating condition index does not fall back to the first threshold range after adjustment, the compensation coefficient is recalculated and the adjustment is iteratively performed until the operating parameters are stable.

[0095] When the target control command is a mud pump shutdown command, a shutdown cooling strategy is determined based on real-time pump temperature data monitored by a temperature sensor. If the temperature exceeds a first preset threshold (e.g., 90°C), a rapid cooling strategy is initiated, involving two-stage valve closure. First, the outlet valve is closed to 50% open within 5 seconds to prevent backflow. After 10 seconds, the inlet valve is fully closed to prevent water hammer. The pressure relief valve is simultaneously opened and maintained open for 15 seconds. Simultaneously, the motor speed is gradually reduced to shutdown at a rate of -5% / second. If the temperature remains within the specified threshold but continues to rise at a rate exceeding a second preset threshold (e.g., 2°C / second), a gentle cooling strategy is initiated, linearly closing the outlet valve to full close within 10 seconds. The motor deceleration curve is then decelerated at a rate of -2% / second to shut down the pump. The valve closure sequence is executed via a hydraulic actuator or electric valve controller to ensure reliable shutoff of the mud pumping channel. The motor deceleration curve is implemented via the inverter's ramp shutdown function to avoid mechanical shock. During shutdown, the temperature and current data are continuously monitored. If the temperature does not drop as expected or the current fluctuates abnormally, the emergency braking protection is triggered, the power is immediately cut off, and the fault log is uploaded.

[0096] In some embodiments, before performing data analysis on the sensor data to determine the abnormal working condition index of the mud pump, the method further includes: acquiring flow state image data of the mud pump;

[0097] The sensor data is analyzed to determine the abnormal working condition index of the mud pump, including: performing multimodal fusion analysis on the flow image data and the sensor data to determine the abnormal working condition index of the mud pump.

[0098] For example, during mud pump operation, in addition to acquiring sensor data such as temperature, current, pressure, liquid level, and flow rate, flow image data within the drainage tank is simultaneously collected. A multispectral imaging array deployed at the tank bottom captures raw images in both visible and near-infrared wavelengths. The visible light image is used to extract dynamic characteristics of the mud flow, while the near-infrared image is processed through noise suppression to enhance visualization of particle distribution. These two image types are fused at the pixel level to generate high-resolution flow image data, providing fluid dynamic information for subsequent multimodal analysis.

[0099] The standardized sensor data is fused with the flow direction parameters (such as plane discreteness and main flow vector angle) extracted from the flow image at the feature level. Through a preset fusion model (such as a cross-attention network), the sensor features and flow characteristics are dynamically weighted to generate a comprehensive operating condition anomaly index. At the same time, the suspended particle density parameter calculated based on the flow image is compared with the preset threshold to further correct the anomaly index. Finally, through the mapping relationship between the anomaly index and the preset threshold range, the adaptive control command (start, speed regulation or shutdown) is triggered to realize the intelligent regulation of the mud pump operating status.

[0100] In some examples, acquiring flow image data of a mud pump includes:

[0101] A multispectral imaging array based on the bottom of the drainage pond collects raw images in the visible light band and near-infrared band respectively;

[0102] Perform multi-frame difference processing on the original image in the visible light band to extract the dynamic characteristic areas of mud flow;

[0103] Perform background noise suppression on the original image in the near-infrared band to generate an enhanced grayscale image;

[0104] The dynamic feature region is fused with the enhanced grayscale image at pixel level to generate flow image data.

[0105] For example, during mud pump operation, a multispectral imaging array deployed at the bottom of the drainage tank simultaneously collects raw image data in the visible and near-infrared bands. The visible light camera uses an RGB sensor to capture visible light images in the wavelength range of 400-700nm, which is used to extract the macroscopic dynamic characteristics of mud flow. The near-infrared camera uses an InGaAs sensor to capture near-infrared images in the wavelength range of 900-1700nm, which is used to penetrate turbid media and visualize the microscopic distribution of suspended particles. The imaging array's sampling frequency is set to 10Hz, synchronized with the sensor data acquisition frequency to ensure temporal and spatial consistency. The resolution of both visible light and near-infrared images is 1920×1080 pixels, and a hardware synchronization trigger mechanism is used to avoid inter-frame offset, providing the raw data foundation for subsequent multimodal analysis.

[0106] Multi-frame differential processing is performed on the original image in the visible light band to extract the dynamic feature area of the mud flow. Specifically, three frames of visible light images (denoted as Frame1, Frame2, and Frame3) are continuously collected, and the pixel differences between adjacent frames are calculated respectively. First, the grayscale value difference operation of Frame1 and Frame2 is performed pixel by pixel to generate the first differential image; then the same operation is performed on Frame2 and Frame3 to generate the second differential image. The two differential images are logically ANDed to retain the pixel areas with significant changes at the same time, and the holes are filled and the edges are smoothed through morphological closing operations to finally generate a binary mask of the dynamic feature area. This mask marks the core dynamic area of the mud flow and effectively filters out static interference caused by fixed structures on the bottom of the pool or changes in illumination.

[0107] The original near-infrared image was processed for background noise suppression to generate an enhanced grayscale image. First, a median filter (window size 5×5) was used to remove impulse noise while preserving the continuity of the particle distribution. Second, the image contrast was enhanced using the contrast-limited adaptive histogram equalization (CLAHE) algorithm. The image was divided into 8×8 sub-blocks, and the clipping limit of the histogram of each sub-block was set to 2.0 to prevent local over-enhancement. Finally, the processed image was Gaussian smoothed (σ=1.0) to generate an enhanced grayscale image. The grayscale difference between the suspended particles and the background was significantly improved, and the clarity of the particle boundaries was enhanced by over 30%, providing high-quality input for quantitative analysis.

[0108] Dynamic feature regions extracted from visible light images are fused with near-infrared enhanced grayscale images at the pixel level to generate high-resolution flow image data. Specifically, the binary mask of the dynamic feature regions serves as a weight matrix, and a weighted average fusion of the visible light and near-infrared images is performed. Within the dynamic regions marked by the mask, the visible light image weight is set to 0.7, and the near-infrared image weight is set to 0.3 to highlight flow trajectories. Within the static background regions, the visible light weight is reduced to 0.3, and the near-infrared weight is increased to 0.7 to enhance deep structural information. The fused flow image data retains the dynamic details of the visible light and the particle distribution characteristics of the near-infrared, providing accurate input for 3D motion reconstruction and flow parameter extraction.

[0109] In some examples, multimodal fusion analysis is performed on flow image data and sensor data to determine the abnormal operating index of the mud pump, including:

[0110] Perform preset filtering algorithm operations and normalization processing on sensor data to generate standardized sensor features;

[0111] Based on the flow image data, the suspended particle density parameters and flow direction distribution parameters are extracted through the three-dimensional motion reconstruction algorithm, including:

[0112] Based on the flow image data, a three-dimensional motion vector field of the mud particles is generated by an optical flow estimation algorithm, wherein the three-dimensional motion vector field includes a first velocity component in the first axis direction, a second velocity component in the second axis direction, and a third velocity component in the third axis direction;

[0113] Determining flow direction distribution parameters based on a synthesis result of the first velocity component and the second velocity component, the flow direction distribution parameters including plane flow direction dispersion and main flow vector angle;

[0114] Based on the third velocity component, a suspended particle density parameter is determined, wherein the suspended particle density parameter includes suspended particle density and velocity deviation.

[0115] The standardized sensor features and flow direction distribution parameters are input into the multimodal feature fusion network model to generate the first abnormal feature vector, including:

[0116] Perform feature mapping on the standardized sensor features to generate sensor feature vectors;

[0117] Perform nonlinear transformation on the flow direction distribution parameters to generate the flow direction characteristic vector;

[0118] Based on the cross attention mechanism, feature interaction is performed on the sensor feature vector and the flow direction feature vector to generate the first abnormal feature vector.

[0119] Calculate the deviation between the suspended particle density parameter and the preset density threshold to generate a second abnormal feature vector;

[0120] An operating condition abnormality index is determined according to a weighted fusion result of the first abnormal feature vector and the second abnormal feature vector.

[0121] For example, a preset filtering algorithm and normalization process are performed on the sensor data of the mud pump (including temperature, current, pressure, liquid level, and flow) to generate standardized sensor features. The filtering algorithm uses Kalman filtering, and dynamic process noise and observation noise parameters are set according to different sensor characteristics: the temperature data filtering window is 60 seconds, the pressure and flow data window is 10 seconds, the current data window is 5 seconds, and the liquid level data window is 30 seconds to eliminate transient noise interference. The normalization process uses the minimum-maximum method to linearly map the real-time data to the range of 0-1 based on the maximum and minimum values of each sensor in the historical data, eliminating dimensional differences. For example, the temperature sensor range is 0-150°C, and the real-time value of 80°C is normalized to 0.53; the pressure sensor range is 0-5MPa, and the real-time value of 3MPa is normalized to 0.6. The standardized sensor features provide a unified input for multimodal fusion.

[0122] Based on the flow image data, the three-dimensional motion vector field of the mud particles is generated by the optical flow estimation algorithm. The optical flow estimation adopts the Lucas-Kanade algorithm, which takes the flow image after two consecutive frames are fused as input and calculates the velocity components of each pixel in the X, Y, and Z directions. The X axis is the horizontal direction, the Y axis is the vertical direction, and the Z axis is the depth direction (perpendicular to the imaging plane). For each mud particle, its velocity component (the first velocity component and the second velocity component) in the XY plane is calculated by the displacement between adjacent frames, and the Z axis velocity component (the third velocity component) is estimated in combination with the depth information of the near-infrared image. The three-dimensional motion vector field is stored in matrix form, covering the three-dimensional velocity information of each pixel.

[0123] Based on the velocity components (first velocity component and second velocity component) in the XY plane, the flow direction distribution parameters are calculated. The plane flow direction dispersion is obtained by counting the standard deviation of the direction angles of all particle motion, and the direction angle is calculated by the inverse tangent function of the XY velocity vector (with due north as 0° and calculated clockwise). For example, if the X velocity of a particle is 0.5m / s and the Y velocity is 0.3m / s, the direction angle is arctan(0.3 / 0.5)=30.96°. The main flow vector angle is the weighted average of the direction angles of all particles, and the weight is the modulus of the velocity vector of each particle. For example, if most particles flow in the northeast direction (average direction angle 45°), the main flow vector angle is 45°. This parameter is used to determine whether the overall flow direction of the mud deviates from the preset path.

[0124] Based on the Z-axis velocity component (the third velocity component), the suspended particle density parameter is calculated. The suspended particle density is defined as the percentage of particles whose Z-axis velocity exceeds a preset threshold (for example, 0.2 m / s) to the total number of particles. For example, if the total number of particles is 1000, and 200 particles have a Z-axis velocity greater than or equal to 0.2 m / s, the density is 20%. The velocity deviation is obtained by calculating the ratio of the standard deviation of the Z-axis velocity to the mean, reflecting the fluctuation amplitude of the vertical velocity. For example, if the mean Z-axis velocity is 0.15 m / s and the standard deviation is 0.05 m / s, the deviation is 0.05 / 0.15≈0.33. This parameter is used to evaluate the mud suspension stability.

[0125] Normalized sensor features are input into the feature mapping module to generate a sensor feature vector. The mapping module uses a fully connected neural network with a five-node input layer (corresponding to temperature, current, pressure, liquid level, and flow). The network contains two hidden layers (64 and 32 neurons, respectively). The activation function uses Reinforced Luminance (ReLU) and outputs a five-dimensional sensor feature vector. For example, if the normalized sensor data is [0.53, 0.6, 0.7, 0.8, 0.4], the resulting feature vector [0.2, -0.3, 0.7, 0.5, 0.1] captures the nonlinear relationships between the sensor data.

[0126] A nonlinear transformation of the flow direction distribution parameters (plane dispersion and main flow vector angle) is performed to generate a flow direction feature vector. The plane dispersion is normalized by dividing it by 90° (e.g., 5° / 90° ≈ 0.056); the main flow vector angle is normalized by dividing it by 360° (e.g., 36.87° / 360° ≈ 0.102). Subsequently, a nonlinear mapping using the hyperbolic tangent function (tanh) is performed to generate a two-dimensional flow direction feature vector, for example, [0.05, 0.10] → [0.049, 0.099], which characterizes the complex flow direction pattern.

[0127] The generated 5-dimensional sensor feature vector and 2-dimensional flow direction feature vector are input into the cross-attention mechanism module to realize the multimodal interactive fusion of the two types of features. Specifically, the sensor feature vector is used as the query, and the flow direction feature vector is used as the key and value. First, the sensor feature vector is mapped to the query space (dimension is 5→3) through linear transformation, and the flow direction feature vector is mapped to the key space (dimension is 2→3) and the value space (dimension is 2→3). Subsequently, the similarity matrix between the query and the key is calculated, and the attention weight distribution is generated by normalization through the Softmax function. For example, the pressure dimension (the third dimension) in the sensor feature vector and the plane discreteness dimension (the first dimension) in the flow direction feature have the highest attention weights, indicating that pressure anomalies are significantly correlated with flow direction disturbances. Finally, the attention weights are used to perform weighted summation on the value vectors to generate a 3D interaction feature vector. After concatenating it with the original sensor feature vector, the vector is reduced in dimension through a fully connected layer (8-dimensional input and 5-dimensional output) to obtain a 5-dimensional first anomaly feature vector (e.g., [0.25, -0.15, 0.68, 0.42, 0.09]), which comprehensively characterizes the association anomaly pattern between the sensor data and the flow direction.

[0128] Based on the suspended particle density parameters and velocity deviation parameters extracted from the flow image data, a second abnormal feature vector is generated. First, the deviation between the suspended particle density and the preset density threshold is calculated. The formula is density deviation = (real-time density - preset threshold) / preset threshold; if the real-time density is 20%, the density deviation is (20-15) / 15 = 0.33. Secondly, the velocity deviation is directly input as the second dimension. The density deviation and the velocity deviation are spliced into a 2-dimensional vector (for example, [0.33, 0.33]), and expanded into a 3-dimensional second abnormal feature vector (for example, [0.30, 0.35, -0.12]) through a linear transformation layer (input 2 dimensions, output 3 dimensions), which reflects the degree of abnormality of the suspended state and is used to correct the flow-specific risks not covered by the first abnormal feature vector.

[0129] The first anomaly feature vector (5-dimensional) and the second anomaly feature vector (3-dimensional) are dynamically weighted and fused to generate a comprehensive operating condition anomaly index. The weighting strategy is adaptively adjusted based on the mud pump's operating stage. During stable operation, the first anomaly feature vector is weighted 0.7, and the second anomaly feature vector is weighted 0.3, highlighting the steady-state correlation between the sensor and flow direction. During the transition phase (startup / shutdown), the weights are adjusted to 0.5:0.5 to balance the contributions of multimodal features. The fusion process is achieved through vector concatenation and a fully connected layer. The 5-dimensional first vector and the 3-dimensional second vector are concatenated into an 8-dimensional vector, which is then input into a fully connected network (with 4 hidden nodes and a ReLU activation function). The scalar output is then mapped to the 0-1 range using a sigmoid function. For example, if the first vector is [0.25, -0.15, 0.68, 0.42, 0.09] and the second vector is [0.30, 0.35, -0.12], the weighted fusion results in an anomaly index of 0.72. The index triggers hierarchical control instructions through preset threshold ranges (such as 0-0.3 for normal, 0.3-0.7 for warning, and 0.7-1.0 for fault), thereby achieving precise regulation of the mud pump's operating status.

[0130] In summary, the embodiment of the present application realizes comprehensive perception and precise control of the operating status of the mud pump through multi-spectral imaging array and multi-modal fusion technology. By utilizing dynamic feature extraction and pixel-level fusion of visible light and near-infrared images, combined with three-dimensional motion vector field reconstruction, the discreteness of mud flow direction, main flow vector angle and suspended particle density parameters are quantified; through the cross-attention mechanism, the sensor data and flow characteristics are deeply fused to dynamically generate a comprehensive working condition anomaly index, which effectively overcomes the limitations of the traditional single data source and improves the sensitivity and accuracy of anomaly detection. The hierarchical control strategy is based on the mapping relationship between the anomaly index and the preset threshold, and adaptively triggers the start, speed regulation or shutdown instructions, optimizes the energy efficiency and operational stability of the mud pump, and adapts to different operating stages through a dynamic weight distribution strategy to ensure the timeliness and safety of the control action. The embodiment of the present application reduces equipment failure rate and maintenance costs, extends service life, and provides a high-reliability, low-cost intelligent control solution for complex industrial scenarios such as seawater desalination.

[0131] See also Figure 2 , is a schematic structural diagram of a control device for a mud pump provided in an embodiment of the present application, comprising:

[0132] A mud pump monitoring data acquisition unit 21 is used to acquire sensor data of the mud pump;

[0133] The abnormal working condition index determining unit 22 is used to perform data analysis on the sensor data to determine the abnormal working condition index of the mud pump;

[0134] The target control instruction generating unit 23 determines a target control instruction corresponding to the abnormal operating condition index based on the abnormal operating condition index and a preset threshold interval;

[0135] The mud pump operation control unit 24 controls the mud pump to perform corresponding target control operations based on the target control instruction.

[0136] See also Figure 3 An embodiment of the present application also provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, the steps of any method for controlling a mud pump are implemented.

[0137] Since the electronic device introduced in this embodiment is a device used to implement a control device of a mud pump in the embodiment of this application, based on the method introduced in the embodiment of this application, technical personnel in this field can understand the specific implementation of the electronic device of this embodiment and its various variations. Therefore, how the electronic device implements the method in the embodiment of this application will not be introduced in detail here. As long as the equipment used by technical personnel in this field to implement the method in the embodiment of this application falls within the scope of protection to be protected by this application.

[0138] During the specific implementation process, when the computer program 311 is executed by the processor, any implementation method of the embodiments corresponding to the first aspect can be implemented.

[0139] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0140] Those skilled in the art will appreciate that the embodiments of the present application may provide methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-readable program code.

[0141] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0142] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0143] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0144] The present application also provides a computer program product, which includes computer software instructions. When the computer software instructions are executed on a processing device, the processing device executes Figure 1 The flowchart of a method for controlling a mud pump in the corresponding embodiment.

[0145] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via wired or wireless means. The computer-readable storage medium can be any available medium that a computer can store or a data storage device such as a server or data center that includes one or more available media integrated therein. The available medium can be a magnetic medium, an optical medium or a semiconductor medium, etc.

[0146] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0147] In the several embodiments provided in this application, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0148] Units described as separate components may or may not be physically separate, and 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 these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0149] In addition, the functional units in the various embodiments of the present application may be integrated into a processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The above-mentioned integrated units may be implemented in the form of hardware and / or software functional units.

[0150] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: various media that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory, a magnetic disk, or an optical disk.

[0151] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

[0152] Although the preferred embodiments of this specification have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of this specification.

[0153] Obviously, those skilled in the art may make various changes and modifications to this specification without departing from the spirit and scope of this specification. Thus, if such changes and modifications fall within the scope of the claims of this specification and their equivalents, this specification is intended to include such changes and modifications.

Claims

1. A method for controlling a mud pump, characterized in that: include: Get sensor data from the mud pump; Performing data analysis on the sensor data to determine an abnormal operating condition index of the mud pump; Determining a target control instruction corresponding to the abnormal operating condition index based on the abnormal operating condition index and a preset threshold range; Based on the target control instruction, the mud pump is controlled to perform a corresponding target control operation.

2. The method according to claim 1, characterized in that The sensor data includes temperature data, current data, pressure data, liquid level data and flow data. The sensor data of the mud pump is obtained, including: Acquiring the temperature data by using a temperature sensor deployed on a mud pump body; Obtaining the pressure data through a pressure sensor provided in the mud pump inlet pipe; Obtaining the flow data by means of a flow meter installed at the outlet pipe of the mud pump; The liquid level data is obtained by a liquid level sensor arranged inside the drainage tank; The current data is obtained through a current transformer connected to the mud pump motor.

3. The method according to claim 1, characterized in that The performing data analysis on the sensor data to determine the abnormal operating condition index of the mud pump includes: Preprocessing the sensor data to generate standardized sensor data, wherein the preprocessing includes filtering and normalization operations; The operating condition abnormality index is determined based on a preset fault database and the standardized sensor data.

4. The method according to claim 3, characterized in that The preset fault database includes normal state historical data and fault state characteristic data. The determining of the abnormal operating condition index based on the preset fault database and the standardized sensor data includes: Calculating a deviation between the standardized sensor data and the normal state historical data to obtain a first deviation value, wherein the normal state historical data includes a reference range of sensor data when the mud pump is operating without fault; Performing similarity matching on the standardized sensor data and the fault state characteristic data to obtain a second matching value, wherein the fault state characteristic data includes a sensor data characteristic pattern corresponding to a preset fault type; The operating condition abnormality index is determined based on dynamic weight allocation between the first deviation value and the second matching value.

5. The method according to claim 2, characterized in that The preset threshold interval includes a first threshold interval, a second threshold interval, and a third threshold interval, the maximum value of the first threshold interval is less than or equal to the minimum value of the second threshold interval, and the maximum value of the second threshold interval is less than or equal to the minimum value of the third threshold interval, the target control instruction includes a mud pump start instruction, a mud pump speed adjustment instruction, or a mud pump shutdown instruction, and determining the target control instruction corresponding to the abnormal operating condition index based on the abnormal operating condition index and the preset threshold interval includes: When the abnormal operating condition index is within the first threshold range, generating the mud pump start instruction; or, When the abnormal operating condition index is within the second threshold range, generating the mud pump speed adjustment instruction, the mud pump speed adjustment instruction including the speed adjustment magnitude and the corresponding action duration; or When the abnormal operating condition index is within the third threshold range, a mud pump shutdown instruction is generated.

6. The method according to claim 5, characterized in that The step of controlling the mud pump to perform a corresponding target control operation based on the target control instruction includes: When the target control instruction is the mud pump start instruction, determining the starting power parameter of the mud pump based on the liquid level data, and controlling the mud pump to operate at a speed corresponding to the starting power parameter; or When the target control instruction is the mud pump speed adjustment instruction, determining the compensation coefficient of the speed adjustment magnitude based on the current data and the pressure data; generating a frequency converter frequency adjustment parameter according to the speed adjustment magnitude, the action duration, and the compensation coefficient; and adjusting the output speed of the mud pump motor based on the frequency converter frequency adjustment parameter; or When the target control instruction is the mud pump shutdown instruction, a shutdown cooling strategy is determined based on the temperature data; a valve closing sequence and a motor deceleration curve are generated according to the shutdown cooling strategy; the mud delivery channel is cut off based on the valve closing sequence; and a shutdown operation is performed based on the motor deceleration curve.

7. The method according to claim 1, characterized in that Before analyzing the sensor data to determine the abnormal working condition index of the mud pump, the method further includes: acquiring flow image data of the mud pump; The performing data analysis on the sensor data to determine the abnormal operating condition index of the mud pump includes: performing multimodal fusion analysis on the flow image data and the sensor data to determine the abnormal operating condition index of the mud pump.

8. A control device for a mud pump, characterized in that: include: A mud pump monitoring data acquisition unit, used to acquire sensor data of the mud pump; an abnormal operating condition index determining unit, configured to perform data analysis on the sensor data to determine an abnormal operating condition index of the mud pump; a target control instruction generating unit, which determines a target control instruction corresponding to the abnormal operating condition index based on the abnormal operating condition index and a preset threshold interval; The mud pump operation control unit controls the mud pump to perform a corresponding target control operation based on the target control instruction.

9. An electronic device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is configured to implement the steps of the method for controlling a mud pump according to any one of claims 1 to 7 when executing the computer program stored in the memory.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the control method of the mud pump according to any one of claims 1 to 7 is implemented.

Images