High-silt-content working condition screw pump control system and method

Abstract

The invention discloses a high-silt-content working condition screw pump control system and method, and belongs to the technical field of soil layer drilling, and the system comprises a multi-mode sensing module, an intelligent decision and control module, a self-adaptive adjustment execution module and a self-cleaning and anti-blocking module. According to the high-mud-content working condition screw pump control system and method, the operation state and medium characteristic data of the screw pump are accurately collected through the multi-mode sensing module, and early prediction and accurate diagnosis of faults are achieved in combination with an advanced data processing algorithm and a deep learning model in the intelligent decision-making and control module; meanwhile, a self-adaptive adjustment execution module and a self-cleaning and anti-blocking module are utilized to dynamically optimize the rotating speed and the gap of the motor and intelligently clean and prevent blocking, so that full-process closed-loop control is formed; finally, multiple beneficial effects of improving the monitoring accuracy, enhancing the anti-interference capability of the system, reducing the maintenance cost, reducing the non-planned downtime, and improving the operation efficiency and stability are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of soil layer drilling, and in particular to a control system and method for a screw pump in a high-mud content working condition. Background Art

[0002] Screw pumps are widely used in oil extraction, mineral refining, wastewater treatment, and other fields due to their high conveying efficiency and strong adaptability. In high-mud conditions (such as muddy crude oil transportation and slurry processing), the medium contains high-content and high-viscosity mud particles, which easily adhere to and deposit on the surfaces of the screw pump's stator and rotor, resulting in reduced conveying efficiency, increased equipment wear, a high risk of blockage, and high maintenance costs. Existing screw pump control systems typically use a single type of sensor, which is unable to fully perceive the multi-dimensional parameters in high-mud conditions, resulting in a one-sided judgment of the operating status. Summary of the Invention

[0003] The purpose of the present invention is to provide a screw pump control system and method for high-mud conditions. Through a multimodal perception module, the operating status and medium characteristic data of the screw pump are accurately collected, and combined with the advanced data processing algorithms and deep learning models in the intelligent decision-making and control module, early prediction and accurate diagnosis of faults are achieved; at the same time, the adaptive adjustment execution module and the self-cleaning and anti-blocking module are used to dynamically optimize the motor speed and gap and intelligently clean and prevent blockage, forming a full-process closed-loop control, and ultimately achieving multiple beneficial effects of improving monitoring accuracy, enhancing the system's anti-interference ability, reducing maintenance costs, reducing unplanned downtime, and improving operating efficiency and stability.

[0004] To achieve the above objectives, the present invention provides a screw pump control system for high mud content working conditions, comprising the following modules:

[0005] Multimodal sensing module, used to fully perceive the operating status and medium characteristics of the screw pump;

[0006] Intelligent decision-making and control module, used to receive data collected by the multimodal perception module and evaluate the operating status of the screw pump;

[0007] Adaptive adjustment execution module, used to adjust the motor speed in real time to optimize the delivery flow and torque output of the screw pump;

[0008] The self-cleaning and anti-blocking module is used to clean the inside of the screw pump and flush away muddy deposits on the surface.

[0009] Preferably, the adaptive adjustment execution module, the multimodal perception module, and the self-cleaning and anti-blocking module are all electrically connected to the intelligent decision-making and control module.

[0010] Preferably, the multimodal sensing module includes a high-precision pressure sensor, a distributed strain sensor, a microwave mud concentration sensor, a vibration sensor and a temperature sensor. The high-precision pressure sensor is installed at the inlet and outlet of the screw pump, the distributed strain sensor is installed at the contact part between the stator and the rotor of the screw pump, the microwave mud concentration sensor is installed at the inlet straight pipe section of the screw pump, the vibration sensor is placed on the bearing seat and stator casing of the screw pump, and the temperature sensor is installed in the contact area between the stator and the rotor.

[0011] The present invention also provides a method for controlling a screw pump under high mud content working conditions, which adopts the above-mentioned high mud content working condition screw pump control system, comprising the following steps:

[0012] S1. Obtain raw data of pressure, strain, mud concentration, vibration, and temperature, and perform filtering, noise reduction, and normalization on the raw data;

[0013] S2. The processed data is input into the intelligent decision-making and control module. The intelligent decision-making and control module performs feature extraction and pattern recognition on the processed data, combines historical operation data, determines the operating status of the screw pump, and identifies the type and location of potential faults.

[0014] S3. Based on the evaluation results of the operating status, the reinforcement learning algorithm is used to generate the optimal control strategy.

[0015] Preferably, the specific operation of S1 is:

[0016] S11. Collect the original data of pressure, strain, mud concentration, vibration and temperature at the same moment through high-precision pressure sensors, distributed strain sensors, microwave mud concentration sensors, vibration sensors and temperature sensors;

[0017] S12, using median filtering to remove the pulse noise generated by mud flow and retain the pressure change trend;

[0018] S13, combining wavelet transform and bandpass filtering to separate mechanical vibration, flow pulse and electrical interference components, extract 10-1000Hz fault characteristics, and fuse historical temperature data and current measurement values through Kalman filtering algorithm to improve the smoothness and accuracy of temperature monitoring;

[0019] S14. Calculate the Z-score for each sensor's data, mark data points exceeding three times the standard as abnormal, and replace them with linear interpolation, sliding average, or historical data regression methods based on the distribution density and duration of the abnormal points;

[0020] S15. Comprehensively mine data features from the time domain, frequency domain, and time-frequency domain to extract mean, standard deviation, FFT dominant frequency, and wavelet packet energy entropy. Simultaneously, based on the requirements of high-mud working conditions, it enhances targeted features such as pressure gradient and strain distribution entropy to keenly capture potential fault signs such as equipment blockage and wear, providing rich information for subsequent analysis.

[0021] S16. Dynamically adjust sensor weights based on real-time working conditions, use DS evidence theory to fuse multi-sensor confidence, use sliding window dynamic normalization and quantile normalization to process data, and then use principal component analysis and random forest algorithm to perform feature dimensionality reduction and screening to retain the most discriminative key features and provide high-quality data for subsequent intelligent decision-making.

[0022] Preferably, the specific operation of S2 is:

[0023] S21. Using multi-domain signal processing technology, we mine hidden fault features from raw sensor data. We first calculate the dynamic pressure gradient of the pressure signal and perform time-domain synchronous averaging of the vibration signal to extract periodic fault features. Frequency-domain analysis uses the cepstrum to identify gear meshing frequency. The wavelet packet energy spectrum quantifies the energy distribution of eight sub-bands to capture early bearing wear characteristics. We also construct multi-sensor correlation features: the pressure-flow correlation coefficient is used to determine pump efficiency changes, and the temperature-concentration coupling model is used to predict friction heating risks.

[0024] S22. Convert pressure, vibration, and temperature data into 2D time-frequency graphs and input them into a 3-channel convolutional neural network model for feature extraction. The convolutional neural network model consists of 3 convolutional layers and 2 fully connected layers. A long short-term memory network is also used to predict key parameter trends. The model inputs historical 24-hour multi-sensor data and constructs a deep belief network consisting of a 3-layer restricted Boltzmann machine and a 1-layer BP neural network model to identify complex fault modes. The mode outputs are then adaptively weighted and fused to form the final diagnosis result.

[0025] S23. Based on the distributed strain sensor data, the inverse distance weighted interpolation method is used to calculate the stator surface stress distribution. At the same time, a five-level severity assessment system is established, with each level corresponding to a preset control strategy, forming a closed-loop process of "detection-location-classification-response";

[0026] S24. Combine historical wear data with current operating conditions and use a particle filter algorithm to predict the remaining life of key components. Then, based on a cost-risk model, consider the maintenance cost, downtime loss, and failure probability to generate the optimal maintenance plan.

[0027] Preferably, the specific operation of S3 is:

[0028] S31: When the system is assessed to be operating normally but the mud content is increasing, the adaptive adjustment execution module first slightly reduces the motor speed, while the self-cleaning and anti-blocking module starts the pre-flushing program to preventively clean the inside of the screw pump.

[0029] S32. When the assessment result indicates a moderate blockage risk, the adaptive adjustment execution module adjusts the motor speed to further reduce it. The hydraulic clearance adjustment device automatically increases the stator-rotor clearance. The self-cleaning and anti-blocking module activates the pulse cleaning mode, using high-frequency pulsed high-pressure water flow to forcefully flush away muddy attachments.

[0030] S33. When the evaluation result is a serious fault or blockage, the screw pump is stopped immediately and the emergency cleaning and fault alarm procedures of the self-cleaning and anti-blocking module are started.

[0031] Therefore, the present invention adopts the above-mentioned high-mud working condition screw pump control system and method, accurately collects the operating status and medium characteristic data of the screw pump through the multimodal perception module, and combines the advanced data processing algorithms and deep learning models in the intelligent decision-making and control module to achieve early prediction and accurate diagnosis of faults; at the same time, it uses the adaptive adjustment execution module and the self-cleaning and anti-blocking module to dynamically optimize the motor speed and gap and intelligently clean and prevent blockage, forming a full-process closed-loop control, and ultimately achieving multiple beneficial effects of improving monitoring accuracy, enhancing the system's anti-interference ability, reducing maintenance costs, reducing unplanned downtime, and improving operating efficiency and stability.

[0032] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a control system diagram of a screw pump in high mud content working condition according to the present invention;

[0034] Figure 2 This is a flow chart of a method for controlling a screw pump under high mud content working conditions according to the present invention. DETAILED DESCRIPTION

[0035] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0036] Unless otherwise defined, technical or scientific terms used in the present invention shall have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs.

[0037] Example 1

[0038] like Figure 1 As shown, the present invention provides a screw pump control system for high mud content working conditions, including the following modules:

[0039] The multimodal sensing module is used to comprehensively sense the operating status and medium characteristics of the screw pump. The multimodal sensing module includes a high-precision pressure sensor, a distributed strain sensor, a microwave mud concentration sensor, a vibration sensor, and a temperature sensor. The high-precision pressure sensor is installed at the inlet and outlet of the screw pump, the distributed strain sensor is installed at the contact point between the stator and rotor of the screw pump, the microwave mud concentration sensor is installed at the inlet straight pipe section of the screw pump, the vibration sensor is placed on the bearing seat and stator casing of the screw pump, and the temperature sensor is installed in the contact area between the stator and rotor.

[0040] The intelligent decision-making and control module receives data collected by the multimodal sensing module and evaluates the operating status of the screw pump. Based on a high-performance FPGA chip and a control model based on a reinforcement learning algorithm, the intelligent decision-making and control unit dynamically adjusts the screw pump's operating parameters based on preset rules and learned optimal strategies, and sends control instructions to the actuator. The unit also has a fault prediction function, predicting potential faults in advance and generating maintenance warnings based on historical data and current operating status.

[0041] The adaptive adjustment execution module is used to adjust the motor speed in real time to optimize the screw pump's delivery flow and torque output. It includes a variable frequency drive system and a hydraulic clearance adjustment device. The variable frequency drive system adjusts the motor speed in real time according to the instructions of the intelligent decision-making unit to optimize the screw pump's delivery flow and torque output. It reduces the speed to reduce wear when the mud content is high, and increases the speed to improve delivery efficiency when the mud content is low. The hydraulic clearance adjustment device uses a hydraulic servo system to automatically adjust the gap between the stator and rotor based on wear monitoring data and changes in mud content, ensuring the pump's volumetric efficiency while reducing abnormal wear caused by improper clearance.

[0042] The self-cleaning and anti-blocking module cleans the interior of the screw pump, flushing away mud deposits adhering to its surface. A spiral pulse cleaning channel, evenly distributed with micro-nozzles, is located on the inner wall of the screw pump's stator. When the intelligent decision-making and control module determines a blockage risk within the pump, the pulse cleaning system is activated, spraying high-pressure cleaning fluid through the micro-nozzles in pulses to flush away mud deposits from the stator and rotor surfaces. Furthermore, an ultrasonic pretreatment device is installed at the screw pump's inlet, utilizing the cavitation effect of ultrasound to disperse lumps in the mud and reduce the likelihood of blockage.

[0043] The adaptive adjustment execution module, the multimodal perception module, the self-cleaning and anti-blocking module are all electrically connected to the intelligent decision-making and control module.

[0044] like Figure 2 As shown, the present invention also provides a method for controlling a screw pump under high mud content working conditions, which adopts the above-mentioned screw pump control system under high mud content working conditions, including the following steps:

[0045] S1. Obtain the raw data of pressure, strain, mud concentration, vibration, and temperature, and perform filtering, noise reduction, and normalization on the raw data. The specific operations are as follows:

[0046] S11. Collect the original data of pressure, strain, mud concentration, vibration and temperature at the same moment through high-precision pressure sensors, distributed strain sensors, microwave mud concentration sensors, vibration sensors and temperature sensors;

[0047] S12, using median filtering to remove the pulse noise generated by mud flow and retain the pressure change trend;

[0048] S13, combining wavelet transform and bandpass filtering to separate mechanical vibration, flow pulse and electrical interference components, extract 10-1000Hz fault characteristics, and fuse historical temperature data and current measurement values through Kalman filtering algorithm to improve the smoothness and accuracy of temperature monitoring;

[0049] S14. Calculate the Z-score for each sensor's data, mark data points exceeding three times the standard as abnormal, and replace them with linear interpolation, sliding average, or historical data regression methods based on the distribution density and duration of the abnormal points;

[0050] S15. Comprehensively mine data features from the time domain, frequency domain, and time-frequency domain to extract mean, standard deviation, FFT dominant frequency, and wavelet packet energy entropy. Simultaneously, based on the requirements of high-mud working conditions, it enhances targeted features such as pressure gradient and strain distribution entropy to keenly capture potential fault signs such as equipment blockage and wear, providing rich information for subsequent analysis.

[0051] S16. Dynamically adjust sensor weights based on real-time working conditions, use DS evidence theory to fuse multi-sensor confidence, use sliding window dynamic normalization and quantile normalization to process data, and then use principal component analysis and random forest algorithm to perform feature dimensionality reduction and screening to retain the most discriminative key features and provide high-quality data for subsequent intelligent decision-making.

[0052] S2. The processed data is input into the intelligent decision-making and control module. The intelligent decision-making and control module performs feature extraction and pattern recognition on the processed data. Combined with historical operation data, it determines the operating status of the screw pump and identifies the potential fault type and location. The specific operations are as follows:

[0053] S21. Using multi-domain signal processing technology, we mine hidden fault features from raw sensor data. We first calculate the dynamic pressure gradient of the pressure signal and perform time-domain synchronous averaging of the vibration signal to extract periodic fault features. Frequency-domain analysis uses the cepstrum to identify gear meshing frequency. The wavelet packet energy spectrum quantifies the energy distribution of eight sub-bands to capture early bearing wear characteristics. We also construct multi-sensor correlation features: the pressure-flow correlation coefficient is used to determine pump efficiency changes, and the temperature-concentration coupling model is used to predict friction heating risks.

[0054] S22. Convert pressure, vibration, and temperature data into 2D time-frequency graphs and input them into a 3-channel convolutional neural network model for feature extraction. The convolutional neural network model consists of 3 convolutional layers and 2 fully connected layers. A long short-term memory network is also used to predict key parameter trends. The model inputs historical 24-hour multi-sensor data and constructs a deep belief network consisting of a 3-layer restricted Boltzmann machine and a 1-layer BP neural network model to identify complex fault modes. The mode outputs are then adaptively weighted and fused to form the final diagnosis result.

[0055] S23. Based on the distributed strain sensor data, the inverse distance weighted interpolation method is used to calculate the stator surface stress distribution. At the same time, a five-level severity assessment system is established, with each level corresponding to a preset control strategy, forming a closed-loop process of "detection-location-classification-response";

[0056] S24. Combine historical wear data with current operating conditions and use a particle filter algorithm to predict the remaining life of key components. Then, based on a cost-risk model, consider the maintenance cost, downtime loss, and failure probability to generate the optimal maintenance plan.

[0057] S3. Based on the evaluation results of the operating status, the reinforcement learning algorithm is used to generate the optimal control strategy. The specific operations are as follows:

[0058] S31: When the system is assessed to be operating normally but the mud content is increasing, the adaptive adjustment execution module first slightly reduces the motor speed, while the self-cleaning and anti-blocking module starts the pre-flushing program to preventively clean the inside of the screw pump.

[0059] S32. When the assessment result indicates a moderate blockage risk, the adaptive adjustment execution module adjusts the motor speed to further reduce it. The hydraulic clearance adjustment device automatically increases the stator-rotor clearance. The self-cleaning and anti-blocking module activates the pulse cleaning mode, using high-frequency pulsed high-pressure water flow to forcefully flush away muddy attachments.

[0060] S33. When the evaluation result is a serious fault or blockage, the screw pump is stopped immediately and the emergency cleaning and fault alarm procedures of the self-cleaning and anti-blocking module are started.

[0061] The convolutional neural network model and the BP neural network model are both existing technologies, but the input and output parameter limits are different, so they will not be elaborated in detail.

[0062] Applying this system to river dredging projects includes the following steps:

[0063] S1 Data Preprocessing: The mud concentration data collected by the microwave sensor exhibited pulse noise (fluctuation ±5%), and median filtering was used to retain the trend. The vibration signal was separated into mechanical vibration components (10-1000Hz) through wavelet transform, and it was found that the energy of the bearing characteristic frequency (108Hz) was enhanced.

[0064] S2 fault identification: A deep belief network (DBN) combined with the pressure-flow correlation coefficient (which decreased from 0.9 to 0.75) determined that the pump efficiency decline was caused by mud deposition. The particle filter algorithm predicted the remaining stator life was 600 hours and recommended scheduling maintenance within 72 hours.

[0065] S3 control strategy: After 30 seconds of cleaning, the mud concentration dropped to 32%, the pressure difference recovered to 0.7 MPa, and the system switched to low-speed stable operation mode (speed 1300 rpm), continuously monitoring vibration and pressure changes.

[0066] Therefore, the present invention adopts the above-mentioned high-mud working condition screw pump control system and method, accurately collects the operating status and medium characteristic data of the screw pump through the multimodal perception module, and combines the advanced data processing algorithms and deep learning models in the intelligent decision-making and control module to achieve early prediction and accurate diagnosis of faults; at the same time, it uses the adaptive adjustment execution module and the self-cleaning and anti-blocking module to dynamically optimize the motor speed and gap and intelligently clean and prevent blockage, forming a full-process closed-loop control, and ultimately achieving multiple beneficial effects of improving monitoring accuracy, enhancing the system's anti-interference ability, reducing maintenance costs, reducing unplanned downtime, and improving operating efficiency and stability.

[0067] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A screw pump control system for high mud content working conditions, characterized by: Includes the following modules: Multimodal sensing module, used to fully perceive the operating status and medium characteristics of the screw pump; Intelligent decision-making and control module, used to receive data collected by the multimodal perception module and evaluate the operating status of the screw pump; Adaptive adjustment execution module, used to adjust the motor speed in real time to optimize the delivery flow and torque output of the screw pump; The self-cleaning and anti-blocking module is used to clean the inside of the screw pump and flush away muddy deposits on the surface.

2. A screw pump control system for high mud content working conditions according to claim 1, characterized in that: The adaptive adjustment execution module, the multimodal perception module, the self-cleaning and anti-blocking module are all electrically connected to the intelligent decision-making and control module.

3. A screw pump control system for high mud content working conditions according to claim 1, characterized in that: The multimodal sensing module includes a high-precision pressure sensor, a distributed strain sensor, a microwave mud concentration sensor, a vibration sensor and a temperature sensor. The high-precision pressure sensor is installed at the inlet and outlet of the screw pump, the distributed strain sensor is installed at the contact point between the stator and rotor of the screw pump, the microwave mud concentration sensor is installed at the inlet straight pipe section of the screw pump, the vibration sensor is placed on the bearing seat and stator casing of the screw pump, and the temperature sensor is installed in the contact area between the stator and rotor.

4. A method for controlling a screw pump in a high mud content working condition, characterized by: A high-mud content working condition screw pump control system according to any one of claims 1 to 3 is adopted, comprising the following steps: S1. Obtain raw data of pressure, strain, mud concentration, vibration, and temperature, and perform filtering, noise reduction, and normalization on the raw data; S2. The processed data is input into the intelligent decision-making and control module. The intelligent decision-making and control module performs feature extraction and pattern recognition on the processed data, combines historical operation data, determines the operating status of the screw pump, and identifies the type and location of potential faults. S3. Based on the evaluation results of the operating status, the reinforcement learning algorithm is used to generate the optimal control strategy.

5. A method for controlling a screw pump in a high mud content working condition according to claim 4, characterized in that: The specific operations of S1 are: S11. Collect the original data of pressure, strain, mud concentration, vibration and temperature at the same moment through high-precision pressure sensors, distributed strain sensors, microwave mud concentration sensors, vibration sensors and temperature sensors; S12, using median filtering to remove the pulse noise generated by mud flow and retain the pressure change trend; S13, combining wavelet transform and bandpass filtering to separate mechanical vibration, flow pulse and electrical interference components, extract 10-1000Hz fault characteristics, and fuse historical temperature data and current measurement values through Kalman filtering algorithm to improve the smoothness and accuracy of temperature monitoring; S14. Calculate the Z-score for each sensor's data, mark data points exceeding three times the standard as abnormal, and replace them with linear interpolation, sliding average, or historical data regression methods based on the distribution density and duration of the abnormal points; S15. Comprehensively mine data features from the time domain, frequency domain, and time-frequency domain to extract mean, standard deviation, FFT dominant frequency, and wavelet packet energy entropy. Simultaneously, based on the requirements of high-mud working conditions, it enhances targeted features such as pressure gradient and strain distribution entropy to keenly capture potential fault signs such as equipment blockage and wear, providing rich information for subsequent analysis. S16. Dynamically adjust sensor weights based on real-time working conditions, use DS evidence theory to fuse multi-sensor confidence, use sliding window dynamic normalization and quantile normalization to process data, and then use principal component analysis and random forest algorithm to perform feature dimensionality reduction and screening to retain the most discriminative key features and provide high-quality data for subsequent intelligent decision-making.

6. A screw pump control system for high mud content working conditions according to claim 4, characterized in that: The specific operations of S2 are: S21. Using multi-domain signal processing technology, we mine hidden fault features from raw sensor data. We first calculate the dynamic pressure gradient of the pressure signal and perform time-domain synchronous averaging of the vibration signal to extract periodic fault features. Frequency-domain analysis uses the cepstrum to identify gear meshing frequency. The wavelet packet energy spectrum quantifies the energy distribution of eight sub-bands to capture early bearing wear characteristics. We also construct multi-sensor correlation features: the pressure-flow correlation coefficient is used to determine pump efficiency changes, and the temperature-concentration coupling model is used to predict friction heating risks. S22. Convert pressure, vibration, and temperature data into 2D time-frequency graphs and input them into a 3-channel convolutional neural network model for feature extraction. The convolutional neural network model consists of 3 convolutional layers and 2 fully connected layers. A long short-term memory network is also used to predict key parameter trends. The model inputs historical 24-hour multi-sensor data and constructs a deep belief network consisting of a 3-layer restricted Boltzmann machine and a 1-layer BP neural network model to identify complex fault modes. The mode outputs are then adaptively weighted and fused to form the final diagnosis result. S23. Based on distributed strain sensor data, the inverse distance weighted interpolation method is used to calculate the stator surface stress distribution. At the same time, a five-level severity assessment system is established, with each level corresponding to a preset control strategy, forming a closed-loop process of "detection-location-classification-response"; S24. Combine historical wear data with current operating conditions and use a particle filter algorithm to predict the remaining life of key components. Then, based on a cost-risk model, consider the maintenance cost, downtime loss, and failure probability to generate the optimal maintenance plan.

7. The high mud content screw pump control system according to claim 4 is characterized in that: The specific operations of S3 are: S31: When the system is assessed to be operating normally but the mud content is increasing, the adaptive adjustment execution module first slightly reduces the motor speed, while the self-cleaning and anti-blocking module starts the pre-flushing program to preventively clean the inside of the screw pump. S32. When the assessment result indicates a moderate blockage risk, the adaptive adjustment execution module adjusts the motor speed to further reduce it. The hydraulic clearance adjustment device automatically increases the stator-rotor clearance. The self-cleaning and anti-blocking module activates the pulse cleaning mode, using high-frequency pulsed high-pressure water flow to forcefully flush away muddy attachments. S33. When the evaluation result is a serious fault or blockage, the screw pump is stopped immediately and the emergency cleaning and fault alarm procedures of the self-cleaning and anti-blocking module are started.

Images