A digital prefabricated pumping station and digital control system

Abstract

This application relates to the field of digital control technology, and provides a digital prefabricated pump station and digital control system. The system acquires sensing data such as liquid level, pressure difference, pump flow rate, blockage displacement, and pipeline vibration through a sensing and judgment module, and comprehensively judges abnormal operating conditions, overcoming the limitations of single-parameter judgment and providing a reliable basis for subsequent control. A pump-valve coordination module dynamically adjusts valve opening through multiple parameters, optimizes the expansion and contraction of soft connections based on pipeline vibration and blockage displacement, and coordinates pump speed adjustment based on liquid level, pump flow rate, and blockage displacement to achieve linked control of the pump, valve, and pipeline. An energy efficiency optimization module uses a long short-term memory network to achieve accurate liquid level prediction, and combines it with a particle swarm optimization algorithm to optimize valve opening and pump speed with the goal of maximizing pump efficiency, dynamically generating an optimized pump shutdown liquid level. Simultaneously, it adjusts network parameters based on the deviation between predicted and optimized values, forming a continuously optimizing closed loop.

Description

Technical Field

[0001] This application relates to the field of digital control technology, and in particular to a digital prefabricated pumping station and digital control system. Background Technology

[0002] Digital prefabricated pumping stations, as key equipment for integrated sewage treatment and rainwater drainage, have gradually evolved from early concrete-cast structures to a new generation of modular prefabrication and intelligent integration products. In existing technologies, the control methods of prefabricated pumping stations have evolved from manual control to semi-automatic control and then to preliminary automation control. Manual control relies on manual inspection and operation, resulting in low efficiency and slow response. Semi-automatic control uses simple level sensors to start and stop the pump unit, but can only achieve coarse control of a single parameter. Preliminary automation control introduces sensors such as pressure and flow sensors, allowing for simple adjustments to pump speed and valve opening, but the control logic is mostly based on single-parameter feedback and lacks multi-parameter collaborative analysis.

[0003] In recent years, with the development of the Internet of Things (IoT) and industrial control technology, some prefabricated pumping stations have begun to introduce digital technologies. They collect operational data through sensor networks and combine this with PLCs to achieve basic automated control, which improves operational stability to some extent. Although the automation level of prefabricated pumping stations has improved, the following limitations still exist in practical applications: existing systems mostly only collect core parameters such as liquid level and flow rate, paying insufficient attention to parameters reflecting equipment health status, such as pipeline vibration and blockage displacement; abnormal condition judgment relies on single parameter threshold comparisons without considering the correlation between parameters, leading to a high rate of misjudgment and missed judgment.

[0004] The adjustment of pump speed and valve opening is mostly controlled independently, without forming a multi-parameter coordinated adjustment mechanism. The adjustment of flexible connection expansion and contraction depends only on the single parameter of vibration, without taking into account the cumulative effect of blockage displacement, which can easily lead to pipeline stress concentration or increased equipment wear. At the same time, the pump shutdown liquid level setting of the existing system is mostly a fixed value, without considering the dynamic characteristics of liquid level fluctuations and the influence of environmental factors such as rainfall, resulting in frequent pump start-ups and shutdowns or excessive energy consumption. Pump efficiency judgment is based only on theoretical head, without taking into account actual pipeline flow resistance and media properties for dynamic correction. Energy efficiency optimization lacks data support and algorithm-driven approaches.

[0005] To address the shortcomings of existing technologies, the technical problem this application aims to solve is: how to achieve multi-parameter collaborative control and dynamic optimization of digital prefabricated pumping stations through collaborative analysis of multi-sensor data in order to reduce energy consumption. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this application provides a digital prefabricated pumping station and a digital control system.

[0007] In a first aspect, this application provides a digital control system for a digital prefabricated pumping station, the system comprising: a sensing and judgment module, a pump-valve coordination module, and an energy efficiency optimization module;

[0008] The sensing and judgment module is used to acquire sensing data including liquid level, pressure difference, pump flow, blockage displacement and pipeline vibration through the digital prefabricated pump station, analyze the sensing data and judge abnormal operating conditions to generate staged start and stop pump commands, and output the liquid level fluctuation frequency and fluctuation amplitude.

[0009] The pump-valve coordination module is used to parse the staged start-stop pump commands to execute pump and valve control, including dynamically adjusting the valve opening based on liquid level, pressure difference and pump flow rate, adjusting the expansion and contraction of the flexible connection in combination with pipeline vibration and blockage displacement, dynamically adjusting the pump speed according to liquid level, pump flow rate and blockage displacement, and determining the pipeline flow resistance in combination with valve opening, pipeline parameters and medium properties.

[0010] The energy efficiency optimization module is used to determine real-time liquid level changes by combining the frequency and amplitude of liquid level fluctuations. It processes historical liquid level changes, real-time liquid level changes, and rainfall data through a long short-term memory network, outputs a predicted liquid level value, and determines the predicted pump stop level. It also judges the pump efficiency in real time by combining valve opening and pipeline flow resistance. The module optimizes the valve opening and pump speed through a particle swarm optimization algorithm to generate an optimized pump stop level. Based on the deviation between the predicted pump stop level and the optimized pump stop level, the parameters of the long short-term memory network are adjusted.

[0011] As an optional implementation, determining the predicted pump stop level includes:

[0012] Calculate the coupling coefficient between the frequency and amplitude of liquid level fluctuations, and combine the coupling coefficient with the boundary value of the normal liquid level range to generate real-time liquid level changes that characterize the dynamic trend of liquid level changes. The real-time liquid level changes include the rate of liquid level change and the direction of liquid level change.

[0013] Historical liquid level changes are divided into stable and fluctuating segments according to time series, and different time weights are assigned to each segment. At the same time, historical liquid level changes, real-time liquid level changes, and rainfall data are input into the long short-term memory network.

[0014] After processing by a long short-term memory network, the predicted liquid level is output. The time point when the liquid level first reaches the pump stop level-related threshold during the liquid level drop is extracted from the predicted liquid level value. The predicted liquid level value corresponding to the time point is determined as the predicted pump stop level.

[0015] As an optional implementation, generating the optimized pump stop level includes:

[0016] The head is dynamically corrected based on valve opening and pipeline flow resistance, and the pump efficiency is calculated in real time based on pump group flow rate, dynamically corrected head and motor power.

[0017] With the objective function of maximizing pump efficiency, valve opening and pump speed are used as optimization variables, and optimization constraints for valve opening and pump speed are set based on pipeline flow resistance and blockage displacement, respectively.

[0018] The particle swarm optimization algorithm outputs an optimized combination of valve opening and pump speed variables to simulate the liquid level change process under the variable combination. The cumulative energy consumption of the pump group when the liquid level drops to the pump stop level is recorded. The liquid level is selected based on the cumulative energy consumption of the pump group to generate the optimized pump stop level.

[0019] As an optional implementation, adjusting the parameters of the long short-term memory network includes:

[0020] Calculate the deviation between the predicted pump stop level and the optimized pump stop level, and determine the magnitude and direction of the deviation based on the deviation;

[0021] The parameter adjustment range for historical liquid level changes, real-time liquid level changes, and rainfall data in the long short-term memory network is determined based on the magnitude and direction of the deviation.

[0022] The parameters of the Long Short-Term Memory (LSTM) network are adjusted based on the parameter adjustment range. After the parameter adjustment, the adjustment effect is verified and the prediction bias rate is calculated to determine whether the parameters of the LSM network need to be readjusted.

[0023] As an optional implementation, the dynamically adjusting valve opening includes:

[0024] Analyze the valve adjustment in the staged start-stop pump command, and determine the valve adjustment direction based on the liquid level height;

[0025] During the process of adjusting the valve opening, the adjustment range of the valve opening direction is adjusted in combination with the pressure difference;

[0026] The difference between the pump flow rate and the pipeline transport demand is calculated, and the valve opening adjustment range after pressure difference adjustment is corrected again based on the flow difference to dynamically adjust the valve opening.

[0027] As an optional implementation, adjusting the extension / retraction amount of the flexible connection includes:

[0028] The pump start-up and shutdown commands are analyzed, and the adjustment trend of the expansion and contraction of the flexible connection is determined based on the obtained pipeline vibration.

[0029] Extract the obtained blockage displacement, and based on the cumulative displacement of the blockage displacement in the direction of approaching or moving away from the pump group, correct the adjustment trend of the expansion and contraction of the flexible connection.

[0030] By considering the rate of change of pipeline vibration and blockage displacement, the adjustment speed of the expansion and contraction of the flexible connection is determined, so as to adjust the expansion and contraction of the flexible connection.

[0031] As an optional implementation, the dynamically adjusted pump speed includes:

[0032] The speed adjustment and pump quantity adjustment in the staged start-up and shutdown pump commands are analyzed, and the liquid level height is compared with the pump stop liquid level to determine the direction of pump speed adjustment;

[0033] The difference between the obtained pump flow rate and the target flow rate is used to obtain the flow deviation, and the adjustment range of the pump speed is determined based on the flow deviation.

[0034] The adjustment range of pump speed is limited based on the blockage displacement. At the same time, if multiple submersible digital pumps have been started, the pump speed of the pump set identified as a high-load digital pump is limited first.

[0035] As an optional implementation, acquiring sensing data through a digital prefabricated pumping station includes:

[0036] The ultrasonic level gauge emits ultrasonic pulses and receives liquid level signals to calculate the liquid level height.

[0037] Real-time monitoring of the pressure difference between the valve inlet and outlet based on a zero-resistance bidirectional check valve;

[0038] The flow rate of the pump unit is calculated by measuring the flow velocity at the outlet using a digital device.

[0039] The amount of displacement of the basket grid on the guide rail is used to characterize the blockage displacement;

[0040] Vibration signals from the pipeline are monitored using digital devices and converted into pipeline vibration.

[0041] As an optional implementation, the generation of staged start-stop pump commands includes:

[0042] The real-time acquired sensing data is compared with the normal operating condition threshold to determine the degree of deviation of the sensing data from the normal operating condition threshold.

[0043] Based on the degree of deviation of the sensing data from the normal operating condition threshold and the correlation between the sensing data, abnormal operating conditions are comprehensively judged.

[0044] Based on the judgment of abnormal operating conditions, a tiered start-up and shutdown command for the pump is generated. The tiered start-up and shutdown command includes valve adjustment, speed adjustment, and pump unit quantity adjustment.

[0045] Secondly, this application provides a digital prefabricated pumping station, which includes: a submersible digital pump, a submersible digital pump set, a basket grid, a guide rail, a zero-resistance bidirectional check valve, an ultrasonic level gauge, a downward-bent outlet pipe, a flexible connection, a digital control system, digital devices, and a digital regulating device.

[0046] The basket grille intercepts debris at the inlet, and the guide rail provides displacement guidance for the basket grille, monitoring the displacement to convert it into sensing data of blockage displacement; the submersible digital pump set and the submersible digital pump control the pump flow rate by adjusting the pump set speed; the zero-resistance bidirectional check valve is installed in the outlet pipeline, monitoring the pressure difference before and after the valve in real time to reduce water flow back impact; the flexible connection is used to connect the submersible digital pump set and the outlet pipeline, compensating for pipeline vibration and displacement by adjusting the extension amount; the ultrasonic level gauge is used to transmit and receive ultrasonic pulses to obtain the liquid level height.

[0047] Compared with existing technologies, the beneficial effects of this application are as follows: The sensing and judgment module acquires multi-dimensional sensing data such as liquid level, pressure difference, pump flow rate, blockage displacement, and pipeline vibration, and comprehensively judges abnormal operating conditions, overcoming the limitations of single-parameter judgment in existing technologies, reducing misjudgments and omissions, and providing a reliable basis for subsequent control; the pump-valve coordination module dynamically adjusts valve opening through multiple parameters, optimizes the expansion and contraction of soft connections based on pipeline vibration and blockage displacement, and coordinates pump speed adjustment based on liquid level, pump flow rate, and blockage displacement to achieve linkage control of pumps, valves, and pipelines, improving system response speed and operational stability, and reducing equipment losses caused by stress concentration or overload; the energy efficiency optimization module integrates historical data, real-time liquid level changes, and rainfall data through a long short-term memory network to achieve accurate liquid level prediction, and optimizes valve opening and pump speed with the goal of maximizing pump efficiency, dynamically generating an optimized pump stop liquid level to solve the problem of excessive energy consumption caused by a fixed pump stop liquid level. Simultaneously, the deviation between predicted and optimized values ​​is used to adjust network parameters, forming a continuous optimization closed loop, further reducing operating energy consumption and improving the economic efficiency of the pumping station. This application upgrades traditional experience-based control to data-driven intelligent control through a full-process digital design that integrates perception, judgment, collaborative control, and intelligent optimization. This enhances the automation level and adaptability of digital prefabricated pumping stations, provides technical support for unattended operation and remote maintenance, and aligns with the development trend of smart water management. Attached Figure Description

[0048] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0049] Figure 1 A system flowchart of a digital control system for a digital prefabricated pumping station provided in an embodiment of this application;

[0050] Figure 2A logic flowchart for determining and predicting the pump stop liquid level in a digital control system of a digital prefabricated pumping station provided in an embodiment of this application;

[0051] Figure 3 A logic flowchart illustrating the generation and optimization of pump stop liquid level in a digital control system for a digital prefabricated pumping station, provided as an embodiment of this application.

[0052] Figure 4 This is a schematic diagram of the structure of a digital prefabricated pumping station provided in an embodiment of this application. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the embodiments of this application more apparent and understandable, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0054] Example 1

[0055] like Figure 1 The diagram shown is a system flowchart of a digital control system for a digital prefabricated pumping station provided in an embodiment of this application. The system includes a sensing and judgment module, a pump-valve coordination module, and an energy efficiency optimization module.

[0056] The sensing and judgment module is used to acquire sensing data including liquid level, pressure difference, pump flow rate, blockage displacement and pipeline vibration through the digital prefabricated pump station, analyze the sensing data and judge abnormal operating conditions to generate staged start and stop pump commands, and output the liquid level fluctuation frequency and fluctuation amplitude.

[0057] Furthermore, acquiring sensing data through digital prefabricated pumping stations includes:

[0058] The ultrasonic level gauge emits ultrasonic pulses and receives liquid level signals to calculate the liquid level height.

[0059] Real-time monitoring of the pressure difference between the valve inlet and outlet based on a zero-resistance bidirectional check valve;

[0060] The flow rate of the pump unit is calculated by measuring the flow velocity at the outlet using a digital device.

[0061] The amount of displacement of the basket grid on the guide rail is used to characterize the blockage displacement;

[0062] Vibration signals from the pipeline are monitored using digital devices and converted into pipeline vibration.

[0063] Liquid level is the core basis for determining whether a pumping station needs to start or stop its pumps. Precise measurement is essential to ensure the accuracy of subsequent operational condition assessments and command generation. An ultrasonic level gauge is installed in the collection tank of the digital prefabricated pumping station. This gauge is installed inside an vent cap at the top of the tank and is perpendicular to the liquid surface. During operation, the ultrasonic level gauge emits ultrasonic pulses at a preset cycle. These pulses are reflected by the liquid surface and received by the ultrasonic level gauge. By calculating the time difference between pulse emission and reception, and combining this with the speed of ultrasonic wave propagation in air, the liquid level is determined. The ultrasonic level gauge is positioned inside the vent cap to prevent water vapor and foam in the tank from affecting the measurement. The non-contact measurement method avoids direct contact with the liquid in the tank, reducing corrosion and wear, extending equipment lifespan. Anti-fog and anti-foam designs improve the stability and accuracy of the measurement, ensuring reliable liquid level data can be obtained even in complex tank environments.

[0064] Pressure difference reflects the resistance of fluid within a pipeline. Combined with liquid level and pump flow rate, it can determine the pump's operating load, providing a basis for valve adjustment. Pressure sensors are installed at appropriate pipeline locations before and after the zero-resistance bidirectional check valve. The selected pipeline locations must ensure smooth fluid flow, avoiding bends and changes in diameter that easily generate turbulence. The measurement point must be located on the pipeline centerline to obtain accurate pressure values. The pressure sensors collect pressure signals before and after the valve in real time, converting them into electrical signals. After amplification and filtering, the difference between the two electrical signals is calculated to determine the pressure difference. To ensure measurement stability, the pressure sensors are calibrated periodically. During calibration, different pressure values ​​are applied using a standard pressure source, and the pressure sensor output is recorded and corrected. The calibration cycle is consistent with the pump station maintenance cycle. Accurate pressure difference measurement can promptly reflect changes in fluid resistance within the pipeline, providing a reliable basis for judging the pump's operating status and adjusting valves, thus helping to improve the efficiency and stability of the pump station.

[0065] Pump flow rate directly reflects the pump's conveying capacity and is a crucial parameter for determining whether the pump meets pipeline conveying requirements and for adjusting pump speed. An electromagnetic flowmeter is installed as a digital device on the straight pipe section before the pump station outlet. The measuring section of the electromagnetic flowmeter is coaxially connected to the pipe to ensure full-pipe flow and avoid additional resistance to fluid flow. The electromagnetic flowmeter calculates the flow velocity by measuring the induced electromotive force generated by the fluid in a magnetic field, and then combines this with the pipe cross-sectional area to determine the pump flow rate. To eliminate the influence of uneven velocity distribution within the pipe, a rectifier is installed upstream of the electromagnetic flowmeter to ensure the fluid enters the measurement area at a stable velocity. Real-time and accurate pump flow rate provides a direct basis for determining whether the pump meets pipeline conveying requirements, facilitating timely adjustment of pump speed, making pump operation more aligned with actual needs, and improving energy efficiency.

[0066] The blockage displacement reflects the degree of blockage in the basket grille. Severe blockage increases the pump load, requiring timely adjustments to the pump's operating status to prevent equipment damage. Displacement sensors are installed on guide rails on both sides of the basket grille, arranged along the length of the guide rails and connected to the basket grille's frame. When the basket grille becomes blocked due to intercepting rainwater and debris, the debris accumulates on the grille, causing it to displace along the guide rails under the impact of water flow. The displacement sensors record the displacement in real time, representing the blockage displacement. They also identify the direction of the displacement, whether it's moving closer to or further away from the submersible digital pump unit. Limiting devices are installed at both ends of the guide rails to prevent the basket grille from detaching from the rails due to excessive displacement, ensuring safe equipment operation. The blockage displacement provides a direct and timely reflection of the basket grille's blockage degree, facilitating early intervention to reduce the additional load on the pump unit caused by blockage, extend the pump's service life, and ensure the normal operation of the pumping station.

[0067] Excessive pipeline vibration can affect the structural stability of pump stations and even cause loosening of pipeline connections. Combining this with blockage displacement data can help determine the cause of vibration and provide a basis for adjusting the expansion and contraction of flexible connections. Vibration sensors are installed at vibration-prone locations such as pipeline bends and valves. These sensors are magnetically fixed to the outer wall of the pipeline for easy installation and maintenance, ensuring accurate detection of pipeline vibration. The vibration sensors convert the mechanical vibration of the pipeline into an electrical signal. After filtering and amplification of the electrical signal, effective vibration information is extracted and converted into pipeline vibration data characterizing the intensity of the vibration. To comprehensively reflect the pipeline vibration situation, multiple vibration sensors can be installed at different locations for multi-point monitoring. Real-time monitoring of pipeline vibration can promptly detect abnormalities in pipeline operation. Combining this with data such as blockage displacement data can help determine the cause of vibration, facilitating targeted adjustments to the expansion and contraction of flexible connections. This reduces the impact of vibration on the pipeline and pump station structure, improving the safety and stability of pump station operation. Pipeline vibration is used in abnormal condition judgment and in the pump-valve coordination module to determine the adjustment trend of the expansion and contraction of flexible connections.

[0068] Specifically, generating staged start / stop pump commands includes:

[0069] The real-time acquired sensing data is compared with the normal operating condition threshold to determine the degree of deviation of the sensing data from the normal operating condition threshold.

[0070] Based on the degree of deviation of the sensing data from the normal operating condition threshold and the correlation between the sensing data, abnormal operating conditions are comprehensively judged.

[0071] Based on the judgment of abnormal operating conditions, a tiered start-up and shutdown command for the pump is generated. The tiered start-up and shutdown command includes valve adjustment, speed adjustment, and pump unit quantity adjustment.

[0072] Only by clearly defining the deviations of each sensing data point from the normal operating condition thresholds can a foundation be laid for subsequent comprehensive judgment of abnormal operating conditions, ensuring the accuracy of such judgments. A preset normal operating condition threshold is established for each sensing data point. This threshold is determined based on the design parameters and historical operating data of the digital prefabricated pump station. Real-time acquired sensing data such as liquid level, pressure difference, pump flow rate, blockage displacement, and pipeline vibration are compared with their corresponding normal operating condition thresholds. The position of each sensing data point within the normal operating condition threshold and the situations exceeding the threshold are analyzed. For those exceeding the normal operating condition threshold, the direction and degree of deviation are recorded. For those within the normal operating condition threshold, it is noted whether the sensing data point is close to the boundary of the normal operating condition threshold. Through this comparison method, the degree of deviation of each sensing data point from the normal operating condition threshold is clarified, providing clear foundational data for subsequent comprehensive judgments. Therefore, clarifying the degree of deviation of the sensing data provides accurate and specific evidence for the comprehensive judgment of abnormal operating conditions, avoiding judgment errors caused by data ambiguity, and making subsequent instruction generation more targeted.

[0073] Deviations in single sensing data are caused by accidental factors. Only by comprehensively considering the degree of deviation and correlation of multiple data points can abnormal operating conditions be more accurately identified, reducing misjudgments. This involves analyzing the degree of deviation of each sensing data point from the normal operating condition threshold and analyzing the correlation between different sensing data points. For example, when the liquid level deviates significantly from the normal range, is it accompanied by significant changes in pressure difference and abnormal pump flow? When the blockage displacement increases, is there simultaneous intensification of pipeline vibration? And is the increased pipeline vibration due to blockage of the basket grille? Based on preset correlation rules and judgment criteria, the degree of deviation and correlation of each sensing data point are comprehensively evaluated to determine the type and severity level of the abnormal operating condition. Severity levels include minor anomalies, moderate anomalies, severe anomalies, and emergency anomalies. Through comprehensive judgment, situations where a single data point anomaly is misjudged as an abnormal operating condition can be effectively avoided, improving the accuracy and reliability of abnormal operating condition judgment and providing a correct basis for subsequent generation of graded instructions.

[0074] Different control strategies are required for abnormal operating conditions of varying severity. Generating tiered start-up and shutdown pump commands enables the pump-valve coordination module to execute pump and valve control actions more precisely, ensuring the stable operation of the digital prefabricated pumping station. Based on the severity of the abnormal condition, corresponding tiered start-up and shutdown pump commands are generated. For example, for minor abnormalities, the command mainly adjusts the valve opening and pump speed slightly; for moderate abnormalities, the command includes starting or stopping one standby pump while adjusting the valve opening and pump speed accordingly; for severe abnormalities, it involves starting all standby pumps or stopping half of the operating pumps, increasing the adjustment range of valve opening and pump speed; and for emergency abnormalities, the command involves urgently stopping all pumps and closing the check valve. The tiered start-up and shutdown pump commands include specific requirements for valve adjustment, speed adjustment, and pump quantity adjustment. These commands make pump and valve control actions more targeted, avoiding over-control or under-control. While ensuring the stable operation of the pumping station, they also improve energy efficiency and equipment lifespan. The tiered start-up and shutdown pump commands are parsed by the pump-valve coordination module to execute corresponding pump and valve control actions, serving as the basis for the module's operation.

[0075] It should be explained that: the interval between the pump start-up level and the pump stop level is defined as the normal level interval. When the real-time measured level is not within the normal level interval, it is determined that a level fluctuation has occurred. When the real-time measured level is within the normal level interval but undergoes a sharp change, it is also determined that a level fluctuation has occurred. The fluctuation amplitude represents the maximum deviation between the real-time measured level and the boundary of the normal level interval, or the maximum difference of a sharp change within the normal level interval, during the occurrence of a level fluctuation. The level fluctuation frequency represents the number of times a level fluctuation occurs per unit time.

[0076] Liquid level fluctuations are a crucial characteristic reflecting the operating status of a pumping station. Accurately determining whether a fluctuation has occurred is essential for calculating its frequency and amplitude, providing fundamental data for the energy efficiency optimization module. The real-time liquid level height is compared to the normal range. If the liquid level exceeds this range, a fluctuation is identified. If the liquid level is within the normal range, continuous monitoring is performed. If the change in liquid level exceeds a preset threshold for abrupt changes over multiple consecutive sampling periods, a fluctuation is also identified. This comprehensive and accurate determination of liquid level fluctuations provides a reliable basis for subsequent calculations of fluctuation frequency and amplitude, ensuring that the data provided to the energy efficiency optimization module accurately reflects the dynamic changes in liquid level.

[0077] Fluctuation amplitude reflects the severity of liquid level fluctuations and is an important indicator for judging the operational stability of pumping stations. It also provides a basis for the energy efficiency optimization module to determine real-time liquid level changes. When the liquid level exceeds the normal range, the changes in liquid level are continuously tracked, and the maximum deviation between the real-time liquid level and the boundary of the normal range is calculated. This deviation is the fluctuation amplitude. When the liquid level is within the normal range but undergoes a sharp change, the maximum difference in liquid level during the sharp change is calculated as the fluctuation amplitude. At the same time, during the calculation process, the acquired liquid level is screened to remove obvious outliers to ensure the accuracy of the fluctuation amplitude. Thus, accurate calculation of the fluctuation amplitude provides the energy efficiency optimization module with key data reflecting the severity of liquid level fluctuations, which helps the energy efficiency optimization module to more accurately determine real-time liquid level changes and supports subsequent liquid level prediction and optimization control.

[0078] Liquid level fluctuation frequency reflects the frequency of liquid level fluctuations per unit time. Together with fluctuation amplitude, it reflects the dynamic change characteristics of the liquid level, providing a basis for liquid level prediction in the energy efficiency optimization module. A fixed statistical period is set, and the number of liquid level fluctuations is counted within this period. During the statistics, for consecutive liquid level fluctuations, if the interval between fluctuations is less than a preset interval threshold, it is considered as one fluctuation; if the interval is greater than the preset interval threshold, it is considered as multiple fluctuations. The number of liquid level fluctuations is divided by the duration of the statistical period to obtain the liquid level fluctuation frequency. The statistical period can be adjusted according to the pump station's operating conditions, for example, shortening the statistical period under special conditions such as rainfall to better reflect the actual fluctuation situation. The calculated liquid level fluctuation frequency reflects the frequency of liquid level fluctuations per unit time. Combined with fluctuation amplitude, it can comprehensively reflect the dynamic change characteristics of the liquid level, providing richer evidence for liquid level prediction in the energy efficiency optimization module and improving the accuracy of prediction. Liquid level fluctuation frequency and fluctuation amplitude will be used to determine real-time liquid level changes and are important components of the Long Short-Term Memory network for data processing.

[0079] The pump-valve coordination module is used to parse the staged start-stop pump commands to execute pump and valve control, including dynamically adjusting the valve opening based on liquid level, pressure difference and pump flow rate, adjusting the expansion and contraction of the flexible connection in combination with pipeline vibration and blockage displacement, dynamically adjusting the pump speed according to liquid level, pump flow rate and blockage displacement, and determining the pipeline flow resistance in combination with valve opening, pipeline parameters and medium properties.

[0080] Furthermore, dynamically adjusting the valve opening includes:

[0081] Analyze the valve adjustment in the staged start-stop pump command, and determine the valve adjustment direction based on the liquid level height;

[0082] During the process of adjusting the valve opening, the adjustment range of the valve opening direction is adjusted in combination with the pressure difference;

[0083] The difference between the pump flow rate and the pipeline transport demand is calculated, and the valve opening adjustment range after pressure difference adjustment is corrected again based on the flow difference to dynamically adjust the valve opening.

[0084] The tiered start / stop pump command includes the core requirement of valve adjustment. The adjustment direction must be clearly defined first to ensure subsequent actions meet the control objectives and avoid exacerbated liquid level fluctuations due to incorrect direction. After receiving the tiered start / stop pump command output from the sensing and judgment module, the system parses the command and extracts information related to valve adjustment, such as the trend indication of valve opening and closing. Simultaneously, it calls upon the real-time liquid level height transmitted by the sensing and judgment module and compares it with the normal liquid level range. If the liquid level height is higher than the upper limit of the normal range (i.e., the pump start level), combined with the implicit requirement of lowering the liquid level in the valve adjustment, the valve adjustment direction is determined to be increasing the opening. If the liquid level height is lower than the lower limit of the normal range (i.e., the pump stop level), combined with the requirement of raising the liquid level, the valve adjustment direction is determined to be decreasing the opening. An instruction verification mechanism is set up during the parsing process. When the tiered start / stop pump command conflicts with the liquid level status (e.g., the liquid level is too high but the command commands to close the valve), an alarm is triggered, and the direction is determined based on the liquid level status. This dual verification ensures the accuracy of the valve adjustment direction, avoiding misoperation caused by relying solely on the instruction, and laying a correct foundation for subsequent amplitude adjustments.

[0085] Pressure difference reflects the fluid resistance within the pipeline. Under the same liquid level deviation, different fluid resistances require different valve opening adjustment ranges. Determining the valve adjustment direction solely based on the liquid level height cannot meet the requirements for precise control. After determining the adjustment direction, the real-time monitored pressure difference is used and compared with the preset normal pressure difference range. If the pressure difference is greater than the normal pressure difference range, it indicates high pipeline resistance. The valve opening adjustment range is increased in the current adjustment direction to quickly reduce fluid resistance. If the pressure difference is less than the normal pressure difference range, it indicates low pipeline resistance. The valve opening adjustment range is decreased in the current adjustment direction to avoid a sudden drop in pressure due to over-adjustment. The adjustment range is achieved through a preset pressure range correction table. This pressure range correction table is pre-calibrated based on the characteristics of the pump station pipeline and covers the correction ratios corresponding to different pressure difference ranges. Thus, the pressure difference is introduced as the basis for correcting the opening adjustment range, making the valve adjustment more closely match the actual pipeline resistance state and avoiding liquid level oscillations or adjustment lag caused by a one-size-fits-all approach.

[0086] The matching degree between pump flow rate and pipeline transport demand directly affects the regulation effect. Considering only the pressure difference will ignore the flow supply and demand imbalance problem, and a secondary correction is required using the flow difference value. The calculated flow difference between the pump flow rate and the pipeline transport demand is used. If the flow difference is positive, it means that the pump flow rate is greater than the pipeline transport demand, indicating that the current transport capacity is excessive. Based on the opening adjustment range after the pressure difference adjustment, the opening adjustment range is further reduced. If the flow difference is negative, it means that the pump flow rate is less than the pipeline transport demand, indicating that the transport capacity is insufficient, and the adjustment range is further increased. During the correction process, an upper limit for the opening adjustment range is set to prevent water hammer effect caused by sudden changes in pump flow rate. Thus, through the secondary correction of the flow difference value, multi-parameter coordinated regulation of liquid level height, pressure difference, and pump flow rate is achieved, so that the valve opening can meet the liquid level control target and match the actual transport capacity of the pump, reducing energy waste. The final determined opening adjustment range will be directly used for the action of the valve actuator. Its accuracy determines the stability of liquid level control and also affects the basis for calculating pipeline flow resistance.

[0087] Furthermore, adjusting the scaling of the soft connection includes:

[0088] The pump start-up and shutdown commands are analyzed, and the adjustment trend of the expansion and contraction of the flexible connection is determined based on the obtained pipeline vibration.

[0089] Extract the obtained blockage displacement, and based on the cumulative displacement of the blockage displacement in the direction of approaching or moving away from the pump group, correct the adjustment trend of the expansion and contraction of the flexible connection.

[0090] By considering the rate of change of pipeline vibration and blockage displacement, the adjustment speed of the expansion and contraction of the flexible connection is determined, so as to adjust the expansion and contraction of the flexible connection.

[0091] The tiered start-stop pump command includes requirements for equipment operational stability. Pipeline vibration is the core triggering factor for flexible connection adjustment. The adjustment trend must first be determined based on pipeline vibration to quickly respond to equipment protection needs. The command analyzes the equipment protection-related content within the tiered start-stop pump command, such as the implicit requirements for vibration suppression and displacement compensation. Simultaneously, it retrieves real-time pipeline vibration data and compares it with a preset safe vibration threshold. If the pipeline vibration exceeds the safe threshold, the adjustment trend for the flexible connection's expansion / contraction is determined to be increasing, i.e., absorbing vibration energy by increasing the expansion / contraction. If the pipeline vibration is less than the safe threshold, the basic trend is determined to be decreasing, i.e., ensuring a rigid pipeline connection by reducing the expansion / contraction. During analysis, the frequency characteristics of the vibration are recorded simultaneously. If it is high-frequency vibration, it is based on pump group resonance, and a priority for rapid adjustment is added to the adjustment trend. Therefore, determining the adjustment trend based on pipeline vibration allows for direct measures to address equipment vibration risks. The high-frequency vibration marker ensures rapid response in emergencies, initially guaranteeing the safety of the pipeline structure.

[0092] Blockage displacement causes pipeline stress shift. Adjusting solely based on vibration will cause the flexible connection's expansion and contraction direction to conflict with displacement compensation requirements. Therefore, displacement direction correction is necessary to balance vibration suppression and displacement compensation. Real-time blockage displacement is extracted, and the cumulative displacement towards or away from the pump group is analyzed. If the blockage displacement accumulates towards the pump group, indicating a tendency for the pipeline to be pushed towards it, the expansion and contraction direction is shifted away from the pump group when the adjustment trend is to increase the expansion and contraction. Conversely, if the blockage displacement accumulates away from the pump group, the expansion and contraction direction is shifted towards the pump group when the adjustment trend is to increase the expansion and contraction. If the adjustment trend is to decrease the expansion and contraction, a portion of the expansion and contraction is appropriately retained based on the magnitude of the cumulative displacement to cope with potential displacement impacts. By combining the direction of blockage displacement with the adjustment trend correction, the flexible connection can both absorb vibration and compensate for pipeline displacement caused by blockage, avoiding stress concentration at the pipeline connection caused by single vibration adjustment.

[0093] The rate of change of pipeline vibration and blockage displacement reflects the speed of risk escalation. Adjusting too slowly leads to risk accumulation, while adjusting too quickly triggers new vibrations. Dynamic matching of the rate of change is necessary to balance safety and stability. The rate of change of pipeline vibration and blockage displacement are calculated. The rate of change of pipeline vibration ensures the increment of pipeline vibration per unit time, while the rate of change of blockage displacement represents the increment of cumulative displacement per unit time. If both rates of change exceed preset thresholds, the adjustment speed of the flexible connection's expansion / contraction is set to high speed. If only one exceeds the preset threshold, it is set to medium speed. If neither exceeds the preset threshold, the adjustment speed is set to low speed. During adjustment, the actual expansion / contraction of the flexible connection is monitored in real time by a displacement sensor. When approaching the target value, the speed is automatically reduced to avoid overshoot. Thus, the adjustment speed is dynamically determined based on the speed of risk escalation, enabling rapid risk control in emergencies and ensuring adjustment accuracy under stable operating conditions, reducing secondary vibrations caused by excessively rapid actions. The adjusted expansion / contraction of the flexible connection directly affects pipeline stability, and its status data is fed back to the sensing and judgment module as a reference for subsequent abnormal condition judgments.

[0094] Specifically, dynamically adjusting the pump set speed includes:

[0095] The speed adjustment and pump quantity adjustment in the staged start-up and shutdown pump commands are analyzed, and the liquid level height is compared with the pump stop liquid level to determine the direction of pump speed adjustment;

[0096] The difference between the obtained pump flow rate and the target flow rate is used to obtain the flow deviation, and the adjustment range of the pump speed is determined based on the flow deviation.

[0097] The adjustment range of pump speed is limited based on the blockage displacement. At the same time, if multiple submersible digital pumps have been started, the pump speed of the pump set identified as a high-load digital pump is limited first.

[0098] The speed adjustment and pump quantity adjustment in the staged start-stop pump command are directly related to the liquid level control target. The adjustment direction must be clarified by considering the relationship between the liquid level and the pump stop level to ensure that the speed change meets the liquid level control requirements. The analysis of the speed-related content and pump quantity adjustment requirements in the staged start-stop pump command reveals that the speed-related content includes acceleration and deceleration, while the pump quantity adjustment requirements include the implied insufficient current speed when starting the standby pump. Simultaneously, the liquid level height is detected by the sensing and judgment module and compared with the pump stop level. If the liquid level height is higher than the pump stop level, the liquid level needs to be lowered. In the tiered start-stop pump command, to enhance drainage demand, the direction of pump speed adjustment is determined to be increasing. If the liquid level is lower than the pump stop level, it is necessary to maintain or increase the liquid level. In conjunction with the drainage control requirements, the direction of pump speed adjustment is determined to be decreasing. When the number of pumps in the tiered start-stop pump command is adjusted to reduce the number of operating pumps, the speed adjustment priority of the remaining pumps is automatically increased. Thus, through the triple logic of command parsing, liquid level comparison, and pump quantity association, it is ensured that the direction of pump speed adjustment is consistent with the liquid level control target and equipment operating status, avoiding liquid level loss due to incorrect direction.

[0099] Flow deviation reflects the gap between the actual pumping capacity and the target. Simply determining the adjustment direction is insufficient for precise flow control; the adjustment range must be determined based on the magnitude of the deviation to quickly eliminate it. The flow deviation between the calculated pump flow rate and the preset target flow rate is calculated as follows: If the flow deviation is negative (i.e., the pump flow rate is less than the target flow rate), the adjustment range is determined based on the absolute value of the flow deviation in the determined adjustment direction, with a larger absolute value resulting in a larger adjustment range. If the flow deviation is positive (i.e., the pump flow rate is greater than the target flow rate), the adjustment range is determined based on the absolute value of the flow deviation in the adjustment direction. The adjustment range is determined based on the flow-speed adjustment curve, which is pre-calibrated based on the pump's flow and speed characteristics to ensure that the adjustment range matches the flow change requirements. A single-step adjustment upper limit is also set. Therefore, by determining the adjustment range based on the flow deviation, changes in pump speed directly serve the flow control target, avoiding flow fluctuations caused by blind adjustments and improving the economic efficiency of pump operation.

[0100] Excessive blockage displacement increases the operating load on the pump set, and excessively increasing the pump set speed can lead to overload damage. When multiple pumps are running, high-load equipment should be prioritized for protection to balance lifespan. The blockage displacement is detected by the sensing and judgment module. If the blockage displacement exceeds the warning threshold, the determined adjustment range is reduced. For example, if the blockage displacement exceeds the warning threshold by 50%, the adjustment range is halved. When the system starts multiple submersible digital pumps, the high-load pump set is identified by the current and vibration data of each pump set. When limiting the adjustment range, an additional reduction ratio is added to the adjustment range of the high-load pump set. The limited adjustment range must not be lower than the minimum speed threshold required to maintain the pump set's operation. Thus, pump set overload is avoided by limiting blockage displacement, and differentiated equipment protection is achieved by prioritizing the restriction of high-load pump sets, extending the overall service life of the pump set and balancing flow control and equipment safety. The final determined pump set speed will directly affect the pump set's operating status, and the data will be fed back to the energy efficiency optimization module as the basis for pump efficiency judgment and optimization.

[0101] Based on the current dynamically adjusted valve opening as the basic parameter, establish the correlation between valve opening and initial flow resistance;

[0102] Obtain pipe parameters including pipe diameter, pipe length, and inner wall roughness, and correct the initial flow resistance based on the pipe parameters;

[0103] By taking into account the properties of the medium, including its viscosity and impurity content, the initial flow resistance, after being corrected for the pipeline parameters, is readjusted to determine the pipeline flow resistance.

[0104] Valve opening is a core factor affecting the local resistance of a pipeline. Different openings correspond to different flow resistance benchmarks. The initial flow resistance must first be determined based on the opening to reflect the degree of obstruction the valve exerts on the fluid. The dynamically adjusted valve opening is then obtained, and a pre-defined opening-flow resistance correlation table is invoked. This table is pre-established based on the valve's structural parameters and experimental data. The valve's structural parameters include valve diameter and type, covering the initial flow resistance corresponding to all openings from fully closed to fully open. During the correlation process, the valve opening is filtered to remove outliers caused by instantaneous fluctuations, such as spurious opening changes due to vibration, ensuring the stability of the initial flow resistance. Therefore, determining the initial flow resistance based on the valve opening directly reflects the impact of valve adjustment on pipeline resistance, providing a true benchmark for subsequent corrections and avoiding discrepancies between theoretical calculations and actual results.

[0105] Pipeline parameters determine the friction loss. Considering only valve opening cannot reflect the resistance characteristics of the entire pipeline system. Pipeline parameter correction is necessary to comprehensively assess flow resistance. Preset pipeline parameters, including pipe diameter, length, and inner wall roughness, are used. Based on the actual pipeline layout, such as the distribution of straight and curved sections, the pipeline is divided into several segments. The friction loss correction factor for each segment is calculated. Segments with smaller diameters, longer lengths, and higher inner wall roughness have larger correction factors. The friction loss correction factors for all segments are summed and multiplied by the initial flow resistance to obtain the initial flow resistance after pipeline parameter correction. Additional corrections are applied to older pipelines to account for increased roughness due to aging. This pipeline parameter correction expands the initial flow resistance calculation from local valves to the entire pipeline, more accurately reflecting the fluid resistance throughout the system and providing a comprehensive basis for pump efficiency assessment.

[0106] The viscosity and impurity content of the medium increase the flow resistance of the fluid. Under the same pipe and valve conditions, different medium properties result in different actual flow resistances. It is necessary to adjust the medium properties to ensure the accuracy of the pipe flow resistance determination. The medium properties are obtained in real time. The medium viscosity is measured by an online viscometer, and the impurity content is monitored by a laser particle size analyzer. The medium properties are compared with the standard values ​​of clean water. If the medium viscosity is higher than the standard value, the initial flow resistance after pipe parameter correction is increased. If the impurity content exceeds the standard value, the initial flow resistance after pipe parameter correction is increased. The final pipe flow resistance is determined. The determined pipe flow resistance must meet the constraint that the pipe flow resistance does not exceed the design maximum value when the medium properties are extreme. By adjusting the medium properties, the pipe flow resistance calculation can be adapted to different water quality conditions, avoiding the distortion of flow resistance assessment caused by medium changes. It provides accurate resistance parameters for the energy efficiency optimization module to judge pump efficiency. The final determined pipe flow resistance is transmitted to the energy efficiency optimization module in real time as a key input for pump efficiency calculation and optimization of valve opening and pump speed. Its accuracy directly affects the optimization effect.

[0107] The energy efficiency optimization module is used to determine real-time liquid level changes by combining the frequency and amplitude of liquid level fluctuations. It processes historical liquid level changes, real-time liquid level changes, and rainfall data through a long short-term memory network, outputs a predicted liquid level value, and determines the predicted pump stop level. It also judges the pump efficiency in real time by combining valve opening and pipeline flow resistance. The module optimizes the valve opening and pump speed through a particle swarm optimization algorithm to generate an optimized pump stop level. Based on the deviation between the predicted pump stop level and the optimized pump stop level, the parameters of the long short-term memory network are adjusted.

[0108] Furthermore, such as Figure 2 As shown, determining the predicted pump stop level includes:

[0109] Calculate the coupling coefficient between the frequency and amplitude of liquid level fluctuations, and combine the coupling coefficient with the boundary value of the normal liquid level range to generate real-time liquid level changes that characterize the dynamic trend of liquid level changes. The real-time liquid level changes include the rate of liquid level change and the direction of liquid level change.

[0110] Historical liquid level changes are divided into stable and fluctuating segments according to time series, and different time weights are assigned to each segment. At the same time, historical liquid level changes, real-time liquid level changes, and rainfall data are input into the long short-term memory network.

[0111] After processing by a long short-term memory network, the predicted liquid level is output. The time point when the liquid level first reaches the pump stop level-related threshold during the liquid level drop is extracted from the predicted liquid level value. The predicted liquid level value corresponding to the time point is determined as the predicted pump stop level.

[0112] Liquid level fluctuation frequency only reflects the frequency of change, while fluctuation amplitude only reflects the severity of change. Using either alone is insufficient to fully reflect the dynamic characteristics of the liquid level. Only by combining both can the actual trend of liquid level changes be captured more accurately, providing a reliable basis for subsequent predictions. After obtaining the liquid level fluctuation frequency and fluctuation amplitude output by the sensing and judgment module, a coupling coefficient is calculated using the formula: weight of liquid level fluctuation frequency × liquid level fluctuation frequency + weight of fluctuation amplitude × fluctuation amplitude. The weights of liquid level fluctuation frequency and fluctuation amplitude are pre-set based on their influence on liquid level changes during the pump station's historical operation and can be dynamically adjusted according to actual operating conditions. The calculated coupling coefficient is then correlated with the boundary values ​​of the normal liquid level range. If the coupling... If the coupling coefficient is large and the liquid level moves towards the pump-stopped level, a real-time liquid level change with a fast rate of change and a direction pointing towards the pump-stopped level is generated. If the coupling coefficient is small and the liquid level fluctuates slightly within the normal liquid level range, a real-time liquid level change with a slow rate of change and no obvious direction is generated. During the generation process, abnormal coupling coefficients are filtered to ensure the authenticity of the real-time liquid level change. Abnormal coupling coefficients include abrupt changes caused by instantaneous sensor failures. Thus, the coupling coefficient organically combines the frequency and amplitude of liquid level fluctuations, overcoming the limitations of describing liquid level changes with a single parameter. This makes the generated real-time liquid level change more accurately reflect the dynamic trend of the liquid level and provides high-quality input data for the Long Short-Term Memory Network.

[0113] In historical liquid level changes, the characteristics of liquid level at different times have varying impacts on current predictions. The reference value of stable periods is relatively low, while fluctuating periods better reflect the liquid level change patterns under similar operating conditions. Differentiating these periods can improve the learning efficiency and prediction accuracy of the Long Short-Term Memory (LSTM) network. Historical liquid level changes are divided into time series, and stable and fluctuating periods are identified using a sliding window method. The size of the sliding window is set according to the periodic characteristics of the pump station's liquid level changes. Stable periods are assigned lower time weights, while fluctuating periods, especially recent fluctuations similar to the current operating conditions, are assigned higher time weights. The weight values ​​are determined through historical prediction error analysis. Simultaneously, rainfall data is classified and processed... The data is categorized into different levels, such as no rain, light rain, moderate rain, and heavy rain, with the start time and duration of rainfall clearly marked. The processed historical liquid level changes, real-time liquid level changes, and categorized rainfall data are standardized to obtain a unified data format and magnitude. This data is then aligned according to time series and input into a Long Short-Term Memory (LSTM) network. Before input, the data integrity is checked; if any data is missing, interpolation is used to supplement it. Through differentiated weighting of historical data and categorized processing of rainfall data, the LTM network can more effectively learn the patterns of liquid level changes, especially the characteristics of changes under complex operating conditions, thereby improving the learning efficiency and adaptability of the LTM network to different operating conditions.

[0114] The liquid level prediction value output by the Long Short-Term Memory (LSTM) network is a continuous time series. Key liquid level values ​​related to pump shutdown need to be extracted from it to provide specific reference for pump shutdown control in the digital prefabricated pumping station, giving the prediction results practical application value. After receiving the liquid level prediction values ​​for a future period from the LSM network, trend analysis is performed on the predicted values ​​to filter out portions showing a downward trend. Based on the design parameters and operating experience of the digital prefabricated pumping station, a pump shutdown liquid level correlation threshold is set. This threshold is typically slightly lower than the actual pump shutdown level to allow for a buffer time during shutdown operations. Then, from the filtered portions showing a downward trend, the time point when the pump shutdown liquid level correlation threshold is first reached is identified. The predicted pump stop level is the initially determined predicted pump stop level. The rationality of the initially determined predicted pump stop level is verified. If the deviation from the actual pump stop level under similar historical conditions is large, it is corrected by combining the pump station's drainage capacity and the current level change trend, and finally the predicted pump stop level is determined. By setting a pump stop level association threshold and extracting key time points from the predicted level, the predicted pump stop level can be directly combined with the pump station's pump stop control logic, providing a clear comparison benchmark for the subsequent generation of optimized pump stop levels, improving the practicality of the prediction results. The determined predicted pump stop level will be compared with the generated optimized pump stop level, and its accuracy will directly affect the calculation of the deviation between the two, and thus affect the rationality of the adjustment of the long short-term memory network parameters.

[0115] Furthermore, such as Figure 3 As shown, generating the optimized pump stop level includes:

[0116] The head is dynamically corrected based on valve opening and pipeline flow resistance, and the pump efficiency is calculated in real time based on pump group flow rate, dynamically corrected head and motor power.

[0117] With the objective function of maximizing pump efficiency, valve opening and pump speed are used as optimization variables, and optimization constraints for valve opening and pump speed are set based on pipeline flow resistance and blockage displacement, respectively.

[0118] The particle swarm optimization algorithm outputs an optimized combination of valve opening and pump speed variables to simulate the liquid level change process under the variable combination. The cumulative energy consumption of the pump group when the liquid level drops to the pump stop level is recorded. The liquid level is selected based on the cumulative energy consumption of the pump group to generate the optimized pump stop level.

[0119] Pump head is a crucial parameter for measuring the work capacity of a pump unit. Valve opening and pipeline flow resistance directly affect the actual head. Without correction, the pump efficiency calculated based on the theoretical head will deviate significantly from the actual situation, failing to provide a reliable basis for optimization. This method uses the current valve opening and pipeline flow resistance output from the pump-valve coordination module to dynamically determine the head correction coefficient. A smaller valve opening corresponds to a larger head correction coefficient, and vice versa. Multiplying the theoretical head by these two correction coefficients yields the dynamically corrected actual head. Based on the corrected head, pump flow rate, and motor power, the pump efficiency is calculated using the principle of (pump flow rate × corrected head × liquid density × gravitational acceleration) / (motor power × 3600). Abnormal motor power values ​​are smoothed during the calculation to ensure the stability of the pump efficiency calculation. Abnormal motor power values ​​include peak values ​​caused by instantaneous motor overload. By dynamically correcting the head, the calculated pump efficiency is made closer to the actual operating efficiency of the pump unit, providing accurate performance indicators for subsequent optimization processes and ensuring the correctness of the optimization direction.

[0120] Optimizing solely for pump efficiency can lead to extreme conditions in valve opening and pump speed, impacting the safe operation of pipelines and the pump unit. Therefore, reasonable constraints are needed to ensure equipment safety while maintaining efficiency. This study defines pump efficiency as the objective function of the particle swarm optimization algorithm, using valve opening and pump speed in the pump-valve coordination module as optimization variables. The valve opening ranges from 0 to its maximum, and the pump speed ranges from the minimum safe speed to the rated speed. Constraints on valve opening are set based on pipeline flow resistance; when the pipeline flow resistance exceeds a preset safety threshold, the valve opening is limited. The value must not be less than a certain minimum to avoid excessive pipeline resistance. Based on the blockage displacement obtained by the sensing and judgment module, constraints are set on the pump unit speed. When the blockage displacement exceeds the warning value, the pump unit speed is limited to a certain maximum value to prevent pump overload. The specific values ​​of the constraints are determined based on the pipeline's pressure resistance and the pump unit's load-bearing capacity, and can be dynamically updated according to the equipment status. Thus, by setting clear objective functions, optimization variables, and optimization constraints, the particle swarm optimization algorithm can search for the variable combination that maximizes pump efficiency while ensuring equipment safety, achieving a balance between efficiency and safety.

[0121] The optimized variable combinations output by the particle swarm optimization algorithm need to be verified through actual liquid level change simulations. Simultaneously, energy consumption factors must be considered for selection to determine the final optimized pump stop level, ensuring the optimization results are both efficient and economical. The set objective function, optimized variables, and optimization constraints are input into the particle swarm optimization algorithm. Through iterative search, the algorithm obtains multiple optimized combinations of valve opening and pump speed variables. A digital twin model of the pump station is used to simulate the liquid level change process under each variable combination, recording the entire process from the current liquid level to the pump stop level, including the time, speed, and fluctuations of the liquid level drop. During the simulation, the cumulative energy consumption of the pump unit is calculated in real time. The cumulative energy consumption of the pump unit is the simulated... The sum of the products of motor power and time within a time period is used to analyze the simulation results of multiple variable combinations. The liquid level height corresponding to the combination where the pump efficiency reaches the preset target and the pump group's cumulative energy consumption is the lowest is selected as the optimal pump stop liquid level. If multiple combinations have similar energy consumption and efficiency, the liquid level height corresponding to the variable combination with the most stable liquid level drop process is selected. Thus, through digital twin model simulation and energy consumption screening, the generated optimal pump stop liquid level not only ensures the efficient operation of the pump group but also minimizes energy consumption and improves the economic efficiency of the digital prefabricated pump station operation. The deviation between the generated optimal pump stop liquid level and the predicted pump stop liquid level is calculated, providing a specific basis for adjusting the parameters of the long short-term memory network.

[0122] Specifically, adjusting the parameters of the Long Short-Term Memory (LSTM) network includes:

[0123] Calculate the deviation between the predicted pump stop level and the optimized pump stop level, and determine the magnitude and direction of the deviation based on the deviation;

[0124] The parameter adjustment range for historical liquid level changes, real-time liquid level changes, and rainfall data in the long short-term memory network is determined based on the magnitude and direction of the deviation.

[0125] The parameters of the Long Short-Term Memory (LSTM) network are adjusted based on the parameter adjustment range. After the parameter adjustment, the adjustment effect is verified and the prediction bias rate is calculated to determine whether the parameters of the LSM network need to be readjusted.

[0126] The deviation between the predicted and optimized pump stop levels is a crucial indicator of the prediction accuracy of Long Short-Term Memory (LSTM) networks. Clearly defining the magnitude and direction of this deviation is essential to identifying problems in the LTM prediction and providing targeted guidance for parameter adjustments. The deviation is calculated as follows: a positive deviation indicates the predicted pump stop level is higher than the optimized level (positive deviation), while a negative deviation indicates the predicted level is lower than the optimized level (negative deviation). The magnitude of the deviation is determined by the ratio of the deviation to the pump stop level, categorized as small, medium, or large based on a threshold comparison. Before calculating the deviation, the predicted and optimized pump stop levels are time-synchronized to ensure they correspond to the same pump stop level under the same operating conditions. If a time discrepancy exists, it is corrected before further calculation. By accurately calculating the deviation and determining its magnitude and direction, the prediction bias of the LTM network can be clearly understood, providing a clear target for subsequent parameter adjustments.

[0127] Different deviations reflect different problems in the learning of Long Short-Term Memory (LSTM) networks with different input data. Different parameter adjustment ranges need to be determined based on the magnitude and direction of the deviation to effectively correct the prediction bias of the LTM network. Based on the magnitude and direction of the deviation, corresponding parameter adjustment rules are formulated. For large positive deviations, if analysis shows that the impact of rainfall data is underestimated, the time weight of rainfall in the LTM network is increased, with the adjustment range determined according to the magnitude of the deviation—the larger the deviation, the larger the adjustment—but a single adjustment cannot exceed the maximum proportion of the original weight. For small negative deviations, if it is determined that the time weight of the fluctuation period in historical liquid level changes is too high, the time weight of the fluctuation period is appropriately reduced. Thus, different parameter adjustment ranges are determined according to the deviation, making the parameter adjustment of the LTM network more targeted and effective, enabling rapid correction of the network's prediction bias and improving prediction accuracy.

[0128] After parameter adjustment, the predictive performance of the Long Short-Term Memory (LSTM) network will change. Verification is needed to determine the effectiveness of the adjustment. If the adjustment is ineffective, it needs to be readjusted to ensure the LTM network maintains high prediction accuracy. The relevant parameters of the LTM network are modified according to the determined adjustment range. These parameters include the weights of each input data point and the connection weights between network layers. After parameter modification, historical data from a recent period is selected as the test set and input into the adjusted LTM network for prediction. This yields new predicted liquid level values ​​and predicted pump stop levels. The prediction relationship between the new predicted pump stop levels and the corresponding optimized pump stop levels is then calculated. The deviation rate is calculated and compared with the prediction deviation rate before adjustment. If the prediction deviation rate decreases, the adjustment is effective. If the prediction deviation rate increases or remains basically unchanged, the parameter adjustment range is re-determined based on the new deviation situation, and the adjustment is repeated until the prediction deviation rate reaches the preset target range. By verifying and adjusting the adjusted Long Short-Term Memory Network (LSTM) multiple times, the effectiveness of the parameter adjustment is ensured, enabling the LTM to maintain high prediction accuracy and providing reliable liquid level prediction support for the optimized operation of the pump station. The verified LTM will be used for the next round of liquid level prediction, forming a continuous optimization closed loop and continuously improving the overall performance of the system.

[0129] Example 2

[0130] like Figure 4 The diagram shown is a structural schematic of a digital prefabricated pumping station provided in this application embodiment. The pumping station includes a submersible digital pump, a well shaft, a submersible digital pump set, an inlet flange, a basket grid, a cylinder, a guide rail, a zero-resistance bidirectional check valve, a multi-functional liquid level device, an exhaust cap, an ultrasonic liquid level gauge, a downward-bend outlet pipe, a flexible connection, a digital control system, digital devices, a digital regulating device, an intelligent digital control well, and an outlet flange.

[0131] The cylinder and well shaft serve as the supporting shell of the pump station, providing space for water flow storage and equipment installation, and separating the functional areas of the inlet, submersible digital pump set, and outlet. The basket screen intercepts debris at the inlet, and the guide rail provides displacement guidance for the basket screen, monitoring the displacement to convert it into sensing data of blockage displacement. The inlet flange connects to the external pipe network, and the outlet flange connects to the outlet pipe, realizing the introduction and discharge of water flow.

[0132] The cylinder and well shaft, as the core load-bearing shell of the digital prefabricated pump station, are made of high-strength, corrosion-resistant materials, possessing excellent structural stability. Their internal space is rationally divided into an inlet, a submersible digital pump unit working area, and an outlet functional area, providing temporary storage space for water flow while offering a stable installation foundation for each piece of equipment, ensuring orderly layout and coordinated operation. The inlet flange connects precisely to the external water supply network via a standard pipe connection interface, achieving stable water flow introduction. Similarly, the outlet flange connects to the outlet pipe via a standard interface, ensuring that the water treated by the pump station can be smoothly discharged to the external drainage network, playing a crucial connecting and guiding role in the entire water transmission process.

[0133] The basket grille is installed at the water inlet. With its mesh structure, it can effectively intercept debris flowing in from the external pipe network, such as tree branches and plastic waste, preventing debris from entering the pump station and causing equipment blockage or damage, thus ensuring the smooth flow of subsequent drainage. The guide rail provides a precise displacement guide for the basket grille. During the process of intercepting debris, the displacement caused by debris accumulation can be clearly shown along the guide rail. By monitoring the displacement in real time, it can be accurately converted into sensing data reflecting the blockage status of the pump station, providing key basis for subsequent abnormal operating condition judgment.

[0134] Submersible digital pump sets and submersible digital pumps are drainage equipment. The flow rate of the pump set is controlled by adjusting the pump set speed. A zero-resistance bidirectional check valve is installed in the outlet pipeline to monitor the pressure difference before and after the valve in real time to reduce the backflow impact of water. A flexible connector is used to connect the submersible digital pump set to the outlet pipeline, and the expansion and contraction amount is adjusted to compensate for pipeline vibration and displacement. A digital regulating device controls the valve opening, and the digital device measures the outlet flow velocity, monitors the pump set flow rate, and monitors pipeline vibration. An ultrasonic level gauge is used to transmit and receive ultrasonic pulses to obtain the liquid level height.

[0135] Submersible digital pump sets and submersible digital pumps are the core drainage equipment of pumping stations. Submersible digital pump sets integrate multiple submersible digital pumps and can flexibly control the flow rate of the pump set by precisely adjusting the pump speed according to the actual drainage needs. When the liquid level is high and rapid drainage is required, the speed is increased to increase the flow rate, and when the liquid level is stable, the speed is reduced to maintain a stable flow rate, thus adapting to different drainage conditions.

[0136] The zero-resistance bidirectional check valve is installed in the outlet pipeline. Utilizing its special valve disc structure, it can monitor the pressure difference between the upstream and downstream of the valve in real time. When the water flows in the forward direction, the valve disc opens to drain water smoothly. When the water flow shows a reverse flow trend, the valve disc closes quickly, effectively reducing the damage to the pump station equipment caused by the reverse water flow and ensuring the stable operation of the pump station. The flexible connector is used to connect the submersible digital pump set to the outlet pipeline. It has the physical characteristic of being expandable and contractible. During the operation of the pump station, if pipeline vibration or pipeline displacement occurs due to temperature changes, equipment displacement, etc., the flexible connector can effectively compensate for these changes by adjusting its own expansion and contraction, avoiding pipeline rupture and other failures due to stress concentration, and protecting the integrity of the pipeline system.

[0137] The digital control device can precisely control the valve opening according to the pump station's operating requirements and control commands. When it is necessary to increase the flow rate, the valve opening is increased; when it is necessary to reduce the flow rate to maintain stability, the valve opening is decreased. This achieves precise regulation of the water flow rate and works in conjunction with the adjustment of the pump unit speed to ensure the drainage effect. The digital device has multiple monitoring functions. On the one hand, it can accurately measure the flow velocity at the outlet and, combined with parameters such as the cross-sectional area of ​​the outlet pipe, further calculate the pump unit flow rate, providing data for monitoring the pump station's operating status. On the other hand, it can monitor the pipe vibration in real time and convert the vibration signal into an electrical signal and transmit it to the control system to detect pipe abnormalities in a timely manner.

[0138] Ultrasonic level gauges utilize the physical properties of ultrasound to emit ultrasonic pulses towards the liquid surface in a digital prefabricated pumping station. The ultrasonic pulses are received after being reflected by the liquid surface. By accurately calculating the time difference between the emitted and received pulses, and combining parameters such as the speed of ultrasonic wave propagation in air, the liquid level height data in the pumping station can be accurately obtained, providing basic and crucial liquid level information for pumping station start-up and shutdown control and flow regulation.

[0139] As the core of intelligent control for the pumping station, the digital control system integrates a sensing and judgment module, a pump-valve coordination module, and an energy efficiency optimization module. It can comprehensively receive various sensing data from ultrasonic level gauges, pressure sensors of zero-resistance bidirectional check valves, and digital devices, and accurately control the pump speed of the submersible digital pump set, the valve opening of the digital regulating device, and the extension and contraction of the flexible connection, so as to realize the automated and intelligent operation of the pumping station and ensure efficient and stable drainage.

[0140] The intelligent digital control well is the hardware carrier and installation space of the digital control system. It integrates various control circuits and communication modules. Through internal signal transmission lines and interfaces, it establishes a stable connection with the sensors and actuators in the digital prefabricated pump station. It is responsible for the centralized reception and processing of sensing data and the accurate transmission of control commands, playing a key physical support role in the entire digital control process of the pump station.

[0141] The water flow path includes external pipe network, inlet flange, basket screen filter, temporary storage in cylinder and well, submersible digital pump set pumping, downward bend outlet pipe, zero resistance bidirectional check valve, flexible connection, digital regulating device, outlet flange and external drainage pipe network.

[0142] The sensors include ultrasonic level gauges, pressure sensors for zero-resistance bidirectional check valves, digital devices, and displacement sensors for basket grids, and are controlled through intelligent digital control wells, submersible digital pump sets, digital regulating devices, and flexible connections.

[0143] The cylinder and well shaft are connected by welding and bolts to form a stable integrated pump station shell structure, providing a reliable installation foundation for the internal equipment. The basket grid is connected to the corresponding installation part of the cylinder or well shaft through guide rails and can slide smoothly along the guide rails. The inlet flange and outlet flange are firmly connected to the inlet and outlet openings of the cylinder by bolts and other means to ensure reliable connection with the external pipe network and outlet pipe.

[0144] The submersible digital pump unit is fixedly installed in a designated position inside the well shaft using a specialized mounting bracket or base to ensure stability during operation. The zero-resistance bidirectional check valve is installed in the outlet pipeline via a flange connection, achieving a sealed and secure connection with the upstream and downstream pipelines. The two ends of the flexible connection are connected to the outlet end of the submersible digital pump unit and the outlet pipeline via flanges or other suitable connection methods, ensuring the sealing and flexibility of the connection. The digital regulating device is also installed on the outlet pipeline via flange connections or other methods to achieve precise control of the water flow in the pipeline. The ultrasonic level gauge is installed at a suitable height inside the cylinder or well shaft via a bracket or mounting base to ensure accurate detection of the liquid level.

[0145] Ultrasonic level gauges, pressure sensors for zero-resistance bidirectional check valves, digital devices, and displacement sensors for basket grids transmit acquired sensing data such as liquid level height, pressure difference, pump flow rate, pipeline vibration, and blockage displacement to the digital control system within the intelligent digital control well via dedicated signal transmission cables. During transmission, anti-interference shielded cables or wireless communication modules are used to ensure stable and accurate signal transmission, providing data support for the analysis, judgment, and control of the digital control system.

[0146] The digital control system, located within the intelligent digital control well, generates tiered start-up and shutdown pump commands after analysis and computation. These commands include speed adjustment of the submersible digital pump set and valve opening adjustment of the digital regulating device. These commands are transmitted to the corresponding actuators via control cables. Upon receiving the commands, the actuators precisely execute actions such as adjusting the pump set speed and valve opening, thereby achieving automated control of the digital prefabricated pump station. During signal transmission, the system ensures accurate decoding and execution of commands, guaranteeing the real-time performance and precision of the control effect.

[0147] Since the principle of the pump station in this application embodiment is similar to that of the system described above in this application embodiment, the implementation of the pump station is the same as that of the system, and the repeated parts will not be described again.

Claims

1. A digital control system for a digital prefabricated pumping station, characterized in that, include: The module includes a perception and judgment module, a pump and valve coordination module, and an energy efficiency optimization module. The sensing and judgment module is used to acquire sensing data including liquid level, pressure difference, pump flow, blockage displacement and pipeline vibration through the digital prefabricated pump station, analyze the sensing data and judge abnormal operating conditions to generate staged start and stop pump commands, and output the liquid level fluctuation frequency and fluctuation amplitude. The pump-valve coordination module is used to parse the staged start-stop pump commands to execute pump and valve control, including dynamically adjusting the valve opening based on liquid level, pressure difference and pump flow rate, adjusting the expansion and contraction of the flexible connection in combination with pipeline vibration and blockage displacement, dynamically adjusting the pump speed according to liquid level, pump flow rate and blockage displacement, and determining the pipeline flow resistance in combination with valve opening, pipeline parameters and medium properties. The energy efficiency optimization module is used to determine real-time liquid level changes by combining the frequency and amplitude of liquid level fluctuations. It processes historical liquid level changes, real-time liquid level changes, and rainfall data through a long short-term memory network, outputs a predicted liquid level value, and determines the predicted pump stop level. It also judges the pump efficiency in real time by combining valve opening and pipeline flow resistance. The module optimizes the valve opening and pump speed through a particle swarm optimization algorithm to generate an optimized pump stop level. Based on the deviation between the predicted pump stop level and the optimized pump stop level, the parameters of the long short-term memory network are adjusted.

2. The digital control system for a digital prefabricated pumping station as described in claim 1, characterized in that, The determination of the predicted pump stop level includes: Calculate the coupling coefficient between the frequency and amplitude of liquid level fluctuations, and combine the coupling coefficient with the boundary value of the normal liquid level range to generate real-time liquid level changes that characterize the dynamic trend of liquid level changes. The real-time liquid level changes include the rate of liquid level change and the direction of liquid level change. Historical liquid level changes are divided into stable and fluctuating segments according to time series, and different time weights are assigned to each segment. At the same time, historical liquid level changes, real-time liquid level changes, and rainfall data are input into the long short-term memory network. After processing by a long short-term memory network, the predicted liquid level is output. The time point when the liquid level first reaches the pump stop level-related threshold during the liquid level drop is extracted from the predicted liquid level value. The predicted liquid level value corresponding to the time point is determined as the predicted pump stop level.

3. The digital control system for a digital prefabricated pumping station as described in claim 1, characterized in that, The process of generating the optimized pump stop level includes: The head is dynamically corrected based on valve opening and pipeline flow resistance, and the pump efficiency is calculated in real time based on pump group flow rate, dynamically corrected head and motor power. With the objective function of maximizing pump efficiency, valve opening and pump speed are used as optimization variables, and optimization constraints for valve opening and pump speed are set based on pipeline flow resistance and blockage displacement, respectively. The particle swarm optimization algorithm outputs an optimized combination of valve opening and pump speed variables to simulate the liquid level change process under the variable combination. The cumulative energy consumption of the pump group when the liquid level drops to the pump stop level is recorded. The liquid level is selected based on the cumulative energy consumption of the pump group to generate the optimized pump stop level.

4. The digital control system for a digital prefabricated pumping station as described in claim 1, characterized in that, The parameters for adjusting the Long Short-Term Memory (LSTM) network include: Calculate the deviation between the predicted pump stop level and the optimized pump stop level, and determine the magnitude and direction of the deviation based on the deviation; The parameter adjustment range for historical liquid level changes, real-time liquid level changes, and rainfall data in the long short-term memory network is determined based on the magnitude and direction of the deviation. The parameters of the Long Short-Term Memory (LSTM) network are adjusted based on the parameter adjustment range. After the parameter adjustment, the adjustment effect is verified and the prediction bias rate is calculated to determine whether the parameters of the LSM network need to be readjusted.

5. The digital control system for a digital prefabricated pumping station as described in claim 1, characterized in that, The dynamically adjustable valve opening includes: Analyze the valve adjustment in the staged start-stop pump command, and determine the valve adjustment direction based on the liquid level height; During the process of adjusting the valve opening, the adjustment range of the valve opening direction is adjusted in combination with the pressure difference; The difference between the pump flow rate and the pipeline transport demand is calculated, and the valve opening adjustment range after pressure difference adjustment is corrected again based on the flow difference to dynamically adjust the valve opening.

6. The digital control system for a digital prefabricated pumping station as described in claim 1, characterized in that, The adjustment of the flexible connection's extension / retraction includes: The pump start-up and shutdown commands are analyzed, and the adjustment trend of the expansion and contraction of the flexible connection is determined based on the obtained pipeline vibration. Extract the obtained blockage displacement, and based on the cumulative displacement of the blockage displacement in the direction of approaching or moving away from the pump group, correct the adjustment trend of the expansion and contraction of the flexible connection. By considering the rate of change of pipeline vibration and blockage displacement, the adjustment speed of the expansion and contraction of the flexible connection is determined, so as to adjust the expansion and contraction of the flexible connection.

7. The digital control system for a digital prefabricated pumping station as described in claim 1, characterized in that, The dynamically adjustable pump set speed includes: The speed adjustment and pump quantity adjustment in the staged start-up and shutdown pump commands are analyzed, and the liquid level height is compared with the pump stop liquid level to determine the direction of pump speed adjustment; The difference between the obtained pump flow rate and the target flow rate is used to obtain the flow deviation, and the adjustment range of the pump speed is determined based on the flow deviation. The adjustment range of pump speed is limited based on the blockage displacement. At the same time, if multiple submersible digital pumps have been started, the pump speed of the pump set identified as a high-load digital pump is limited first.

8. The digital control system for a digital prefabricated pumping station as described in claim 1, characterized in that, The acquisition of sensing data through digital prefabricated pumping stations includes: The ultrasonic level gauge emits ultrasonic pulses and receives liquid level signals to calculate the liquid level height. Real-time monitoring of the pressure difference between the valve inlet and outlet based on a zero-resistance bidirectional check valve; The flow rate of the pump unit is calculated by measuring the flow velocity at the outlet using a digital device. The amount of displacement of the basket grid on the guide rail is used to characterize the blockage displacement; Vibration signals from the pipeline are monitored using digital devices and converted into pipeline vibration.

9. The digital control system for a digital prefabricated pumping station as described in claim 1, characterized in that, The generation of graded start / stop pump commands includes: The real-time acquired sensing data is compared with the normal operating condition threshold to determine the degree of deviation of the sensing data from the normal operating condition threshold. Based on the degree of deviation of the sensing data from the normal operating condition threshold and the correlation between the sensing data, abnormal operating conditions are comprehensively judged. Based on the judgment of abnormal operating conditions, a tiered start-up and shutdown command for the pump is generated. The tiered start-up and shutdown command includes valve adjustment, speed adjustment, and pump unit quantity adjustment.

10. A digital prefabricated pumping station, implemented based on a digital control system for a digital prefabricated pumping station according to any one of claims 1-9, characterized in that, include: Submersible digital pumps, submersible digital pump sets, basket grids, guide rails, zero-resistance bidirectional check valves, ultrasonic level gauges, downward-bend outlet pipes, flexible connections, digital control systems, digital devices, and digital regulating devices. The basket grille intercepts debris at the inlet, and the guide rail provides displacement guidance for the basket grille, monitoring the displacement to convert it into sensing data of blockage displacement; the submersible digital pump set and the submersible digital pump control the pump flow rate by adjusting the pump set speed; the zero-resistance bidirectional check valve is installed in the outlet pipeline, monitoring the pressure difference before and after the valve in real time to reduce water flow back impact; the flexible connection is used to connect the submersible digital pump set and the outlet pipeline, compensating for pipeline vibration and displacement by adjusting the extension amount; the ultrasonic level gauge is used to transmit and receive ultrasonic pulses to obtain the liquid level height.

Images