Well electric dual-control intelligent metering and accounting system and method based on edge computing

Abstract

This invention discloses an intelligent metering and accounting system and method for well power dual control based on edge computing. The method identifies stages in the well pump operation process, dividing it into a startup stage, a stable pumping stage, and a shutdown stage. When identified as a stable pumping stage, the system statistically analyzes the water output and effective electricity generation during this stage, establishing a hydroelectric coupling model. Simultaneously, the pumping efficiency parameters are dynamically adjusted using a recursive update method, and the water consumption is calculated based on the updated efficiency parameters. The system deploys edge computing nodes at the well site to perform local analysis and accounting of the collected operational data. This invention can eliminate non-pumping energy consumption generated during the well pump startup stage, improving the accuracy of water consumption calculation and enhancing the operational reliability of the well power dual control system in environments with unstable communication conditions.

Description

Technical Field

[0001] This invention relates to the field of agricultural irrigation water intake management technology, and in particular to a well-electricity dual-control intelligent metering and accounting system and method based on edge computing. Background Technology

[0002] In agricultural irrigation, well pumping is a widely used method of groundwater extraction. To strengthen groundwater resource management and monitor well pumping, a dual-control system for wells and electricity has been gradually promoted in recent years. This dual-control system monitors and controls the electricity consumption and water output during well pumping, enabling metering, management, and quota control of well water usage.

[0003] In existing technologies, well-electricity dual-control systems typically collect electricity data during well pump operation via an electricity metering device and water pumping volume data via a flow metering device. The system establishes a water-electricity conversion relationship based on the collected electricity and water volume data to calculate the water consumption for pumping. In some application scenarios, when the flow metering equipment malfunctions or is difficult to maintain, the water consumption for pumping will also be estimated using the electricity data and a preset water-electricity conversion coefficient.

[0004] However, in actual well pumping operations, the pump's operation typically includes a startup phase, a stable pumping phase, and a shutdown phase. During the startup phase, the motor consumes a significant amount of electrical energy in a short period due to the large starting current surge, but the pump has not yet achieved a stable water output. Therefore, the electricity generated during this phase does not correspond to the actual pumped volume.

[0005] If the electricity generated during the startup phase is directly included in the water-electricity conversion calculation, non-pumping energy consumption will be introduced into the statistical data, thereby interfering with the water-electricity conversion relationship and affecting the accuracy of the water consumption calculation.

[0006] Furthermore, during long-term operation, factors such as changes in groundwater level, pump wear, changes in pipeline resistance, and fluctuations in power supply voltage can cause variations in well pump pumping efficiency, meaning that the water output per unit of electricity is not constant. Existing dual-control systems for wells and electricity typically use fixed hydroelectricity conversion factors for water consumption estimation in practical applications, making it difficult to adapt to the dynamic changes in pumping efficiency caused by variations in operating conditions.

[0007] Furthermore, in some well distribution areas, communication conditions may be unstable, and traditional well-electricity dual-control systems often rely on cloud platforms for centralized processing of collected data and water consumption calculation. When communication links malfunction, the system may be unable to complete water consumption calculations in a timely manner, thus affecting the continuity of system operation. Moreover, in actual well operating environments, flow metering devices are constantly exposed to damp, sandy water underground or semi-underground, making them prone to sensor blockage, impeller jamming, cable corrosion, or poor contact, leading to distorted or interrupted flow metering data. Under these circumstances, existing well-electricity dual-control systems typically use pre-set fixed water-electricity conversion factors to estimate pumping water consumption. However, these fixed conversion factors are often determined based on initial equipment installation calibration data and cannot reflect changes in pumping efficiency caused by factors such as groundwater level decline, pump wear, and pipeline resistance changes during long-term operation. Therefore, when the flow metering device malfunctions, the systematic deviation introduced by using fixed coefficients for estimation will continue to accumulate as the equipment operates for longer periods, leading to a significant deviation between the water consumption calculation results and the actual pumping volume, thus affecting the accuracy of groundwater resource management.

[0008] Therefore, how to accurately identify the operation process of well pumps at the well site, and establish a dynamic hydroelectric coupling relationship based on eliminating non-pumping energy consumption during the start-up phase, thereby improving the accuracy of water consumption calculation and reducing reliance on cloud computing, has become a technical problem that needs to be solved in the field of well-electric dual control systems. Summary of the Invention

[0009] The purpose of this invention is to provide an intelligent metering and accounting system for well power dual control based on edge computing, so as to solve the problems of existing well power dual control systems in achieving accurate accounting under conditions such as non-pumping energy consumption, unstable coupling relationship between power and water volume, and unstable communication conditions during the well pump start-up phase.

[0010] To achieve the above objectives, this invention provides a smart metering and accounting method for well and electricity dual control based on edge computing. This method sets up an edge computing node at the well to monitor the well pump operation process in real time, divides the pumping process into stages according to the well pump operation status, and establishes a water-electricity coupling relationship model after eliminating non-pumping energy consumption in the start-up stage, thereby realizing intelligent accounting of pumping water consumption.

[0011] Specifically, the method includes the following steps.

[0012] Step S1: Well pump operation data acquisition: An edge computing node is set up at the well, and the well pump operation data is periodically acquired through the power metering device and flow metering device connected to the edge computing node.

[0013] The operational data includes: voltage data; current data; active power data; electricity consumption data; instantaneous flow rate data; and cumulative water consumption data.

[0014] The edge computing node collects the above data according to a preset sampling period and forms a well pump operation data sequence according to the collection time order.

[0015] Step S2: Well pump operation status identification: The edge computing node identifies the well pump operation status based on the current change rate, power change amplitude, and flow change characteristics in the operation data sequence, and divides the well pump operation process into: start-up stage; stable pumping stage; shutdown stage;

[0016] The well pump start-up phase is identified by detecting the rate of change of current. When the rate of change of current exceeds a preset threshold or the current value exceeds a preset multiple of the rated current within a continuous sampling period, the well pump is determined to have entered the start-up phase.

[0017] The stable pumping phase is identified by the following conditions:

[0018] Within a continuous sampling period, the current fluctuation amplitude is less than the first preset threshold, the active power fluctuation amplitude is less than the second preset threshold, and the instantaneous flow rate is greater than the minimum flow rate threshold, and the duration of the above conditions reaches a preset time threshold.

[0019] When the detected current value is lower than the minimum operating current threshold, the well pump is determined to enter the shutdown stage.

[0020] Step S3: Identification of non-pumping energy consumption during startup: When the well pump is identified as being in the startup phase, the edge computing node calculates the startup energy consumption based on the power consumption data within the corresponding time period of the startup phase.

[0021] The starting power is obtained by integrating the power data during the starting phase or by accumulating the power data.

[0022] Since the well pump has not yet achieved stable water output during the startup phase, the startup power consumption is marked as non-pumping energy consumption and is removed from the total power consumption in subsequent hydroelectric coupling calculations.

[0023] Step S4: Effective energy consumption statistics during the stable pumping phase: When the well pump is identified as being in the stable pumping phase, the edge computing node accumulates and statistically analyzes the power consumption data within the corresponding time period of the stable pumping phase to obtain the effective power consumption during the stable pumping phase.

[0024] Meanwhile, the instantaneous flow rate data within the corresponding time period of the stable pumping phase are integrated to obtain the output water volume of the stable pumping phase.

[0025] The effective power generation and water output during the stable pumping phase serve as the basic data for hydroelectric coupling calculations.

[0026] Step S5: Establishing the hydroelectric coupling relationship: The edge computing nodes establish a hydroelectric coupling relationship model based on the correspondence between the water output and the effective power during the stable pumping stage.

[0027] Specifically, the ratio between the water output during the stable pumping phase and the effective electricity generation during the corresponding time period is determined as the observed pumping efficiency value for that pumping cycle.

[0028] The pumping efficiency observation value is used to reflect the pumping capacity of the well pump per unit of electricity within the current pumping cycle.

[0029] Step S6: Recursive Update of Pumping Efficiency Parameters: In order to adapt to the impact of changes in groundwater level, well pump operating status, and pipeline resistance on pumping efficiency, this invention dynamically updates the pumping efficiency parameters through a recursive method.

[0030] Specifically, the pumping efficiency observation value obtained in the current pumping cycle is weighted and merged with the efficiency coefficient of the previous pumping cycle to obtain the efficiency coefficient of the current pumping cycle.

[0031] The weighted fusion update is controlled by a preset smoothing coefficient, which is a real number greater than 0 and less than 1, so that the updated efficiency coefficient can reflect the operating characteristics of the current pumping cycle and maintain the stability of the pumping efficiency parameters.

[0032] Step S7: Intelligent Water Consumption Calculation: The edge computing node calculates the water consumption for this pumping cycle based on the updated efficiency coefficient and the effective power generation during the stable pumping phase, and generates the corresponding water consumption calculation results.

[0033] The water usage calculation results can be stored locally on edge computing nodes and uploaded to the cloud platform for centralized management via a communication module.

[0034] In a preferred embodiment, the edge computing node further includes a local data caching unit and a local computing unit.

[0035] The local data cache unit is used to store the collected well pump operation data sequence and the corresponding timestamp information.

[0036] The local accounting unit is used to perform the calculations from steps S2 to S7 above according to the running data sequence, and to store the generated water usage accounting results in the local data cache unit.

[0037] With the above structure, even if the communication module is temporarily unable to establish a communication connection with the cloud platform, the edge computing node can still complete the calculation of water consumption based on the locally stored data, and upload the calculation results to the cloud platform after the communication is restored.

[0038] Based on the above method, the present invention also provides a well power dual-control intelligent metering and accounting system based on edge computing.

[0039] The system includes: a wellhead data acquisition module; an edge computing node; a communication module; and a cloud platform; the above modules are connected through data communication.

[0040] The wellhead data acquisition module is used to monitor the operating status of the well pump in real time and to collect electrical and pumping parameters during the operation of the well pump.

[0041] The wellhead data acquisition module includes an energy metering device, a flow metering device, a voltage acquisition unit, and a current acquisition unit.

[0042] The power metering device is used to acquire power data during the operation of the well pump; the flow metering device is used to acquire instantaneous flow data and cumulative water volume data during the pumping process; and the voltage acquisition unit and current acquisition unit are used to acquire voltage and current data during the operation of the well pump.

[0043] The wellhead data acquisition module collects the above-mentioned operational data according to a preset sampling period and sends the collected operational data to the edge computing node.

[0044] Edge computing nodes are located at the well site. Even if the communication module is temporarily unable to establish a communication connection with the cloud platform, the local accounting unit can still independently execute steps S2 to S7 based on the operating data stored in the local data cache unit. The generated water usage accounting results are persistently stored in the local data cache unit and automatically re-uploaded to the cloud platform in chronological order after communication is restored, achieving consistent data links and accounting traceability between the end and the cloud. Recursive updates only need to save the data from the previous cycle. Compared with the statistics obtained in this period , It can complete the calculation in a timely manner, with both storage and computational overhead being constant, making it suitable for real-time operation on embedded edge devices at the well site.

[0045] The edge computing node includes: a data acquisition unit, an operation status identification unit, a startup energy consumption identification unit, a pumping stable section identification unit, a hydroelectric coupling calculation unit, a dynamic calibration unit, a local accounting unit, and a local data cache unit.

[0046] in:

[0047] The data acquisition unit is used to receive well pump operation data sent by the wellhead data acquisition module and construct a well pump operation data sequence according to the acquisition time order;

[0048] The operation status identification unit is used to identify the well pump operation status based on the current change rate, power change amplitude, and flow rate change characteristics, and to divide the well pump operation process into stages.

[0049] The startup energy consumption identification unit is used to calculate the electricity generated during the startup phase when the system is identified as a startup phase, and to mark this portion of electricity as non-pumping energy consumption.

[0050] The pumping stability section identification unit is used to identify the stable pumping stage based on the operation status identification results, and to determine the data statistics interval corresponding to the stable pumping stage;

[0051] The hydroelectric coupling calculation unit is used to calculate the pumping efficiency observation value based on the correspondence between the water output and the effective power during the stable pumping stage.

[0052] The dynamic calibration unit is used to recursively update the pumping efficiency parameters based on efficiency observations from multiple pumping cycles.

[0053] The local accounting unit is used to calculate the water consumption for pumping based on the updated efficiency coefficient and the effective power generation during the stable pumping phase, and to generate the corresponding water consumption accounting results.

[0054] The local data cache unit is used to store well pump operation data sequences and generated water usage calculation results.

[0055] The communication module is used to enable data communication between edge computing nodes and the cloud platform.

[0056] The communication module can employ one or a combination of the following communication methods: NB-IoT communication; 4G communication; LoRa communication;

[0057] The communication module is used to send well pump operation data and water usage calculation results generated by the edge computing node to the cloud platform.

[0058] When the communication connection is normal, the edge computing node can upload the running data and calculation results to the cloud platform in real time for centralized management.

[0059] The cloud platform is used for centralized storage and unified management of operational data and water usage accounting results from edge computing nodes.

[0060] The cloud platform includes a data storage module and a data management module.

[0061] in:

[0062] The data storage module is used to store well pump operation data and water usage calculation results uploaded by edge computing nodes;

[0063] The data management module is used to perform statistical analysis on the operating data and water consumption calculation results of multiple wells, and to realize remote monitoring and operation management of the well operating status.

[0064] In this invention, each functional unit in the system is used to implement the above method steps.

[0065] In the system architecture of this invention, each functional unit forms a data processing pipeline for pumping metering and accounting through edge computing nodes, thereby realizing real-time analysis of well pump operation data and water consumption accounting.

[0066] Specifically, the wellhead data acquisition module, as the system's data ingestion layer, is responsible for periodically collecting data on voltage, current, active power, electricity, and flow rate during the operation of the well pump, and sending the collected raw operating data to the edge computing nodes in chronological order.

[0067] Edge computing nodes, serving as the system's edge computing layer, are used to process collected runtime data in real time. Internally, each edge computing node consists of multiple functional units forming a sequentially executed data processing chain.

[0068] The data acquisition unit is used to receive the operating data sent by the wellhead data acquisition module and perform time-series processing on the operating data to construct a well pump operating data sequence.

[0069] The operation status identification unit is used to determine the operation status based on the operation data sequence. By jointly analyzing the current change rate, power fluctuation amplitude and flow change characteristics, the well pump operation process is divided into stages.

[0070] The start-up energy consumption identification unit is used to calculate the energy consumption data within the corresponding time period when the start-up phase is identified, and to mark this part of the energy consumption as non-pumping energy consumption;

[0071] The pumping stability stage identification unit is used to identify the stable pumping stage based on the operation status identification results, and to determine the data statistics window corresponding to the stable pumping stage;

[0072] The hydroelectric coupling calculation unit is used to perform statistical processing on the flow rate data and power data during the stable pumping stage based on the data statistics window, so as to calculate the observed value of pumping efficiency.

[0073] The dynamic calibration unit is used to recursively update the pumping efficiency parameters based on the efficiency coefficients of historical pumping cycles and the efficiency observations of the current pumping cycle, so as to form a continuously evolving hydroelectric coupling relationship model.

[0074] The local accounting unit is used to calculate the water consumption for this pumping cycle based on the updated efficiency coefficient and the effective power generation during the stable pumping phase, and to generate the corresponding water consumption accounting results.

[0075] During the aforementioned data processing, the local data cache unit is used to persistently store the running data sequence and calculation results, thereby ensuring that the system can still complete the pumping metering and calculation tasks when the communication link is temporarily unavailable.

[0076] The communication module, as the system's data synchronization layer, is used to upload the operational data generated by the edge computing nodes and the water usage calculation results to the cloud platform.

[0077] The cloud platform, as the system's data management layer, is used to centrally store and uniformly manage operational data and accounting results from multiple wells, and to provide data support for subsequent operational monitoring and statistical analysis.

[0078] Through the above system architecture, this invention shifts the pumping metering and accounting tasks from the traditional cloud processing mode to the edge computing node at the well site, enabling the well pump operation data to be collected, processed, model updated, and water usage calculated locally, thus forming an edge intelligent metering architecture for well-electric dual control scenarios.

[0079] Compared with existing technologies, the edge computing-based intelligent metering and accounting system for dual-control well power supply provided by this invention has the following advantages:

[0080] First, this invention identifies the operational stages of the well pump, dividing the pump operation into a startup stage, a stable pumping stage, and a shutdown stage. It also identifies and marks the electricity generated during the startup stage, thereby eliminating non-pumping energy consumption during the hydroelectric coupling calculation. Since well pumps typically generate a large starting current surge during the startup stage, before a stable water output is achieved, eliminating the electricity generated during this stage effectively avoids interference from non-pumping energy consumption in the hydroelectric conversion relationship, thus improving the accuracy of water consumption calculation.

[0081] Secondly, this invention statistically analyzes the effective power generation and water output during the stable pumping phase, calculates the observed pumping efficiency based on the statistical results, and dynamically adjusts the pumping efficiency parameters using a recursive update mechanism. This allows the hydroelectric coupling relationship to be continuously updated according to changes in the well pump's operating conditions. Since factors such as changes in groundwater level, pump wear, and pipeline resistance can cause pumping efficiency to vary between different pumping cycles, dynamically updating the efficiency parameters makes the hydroelectric conversion relationship closer to the actual operating state, thereby improving the system's adaptability to complex pumping conditions.

[0082] Furthermore, this invention deploys edge computing nodes at the well site, enabling the collection, analysis, and water consumption calculation of well pump operation data locally. This decentralizes the traditional cloud-based water consumption calculation process to the well side. This edge computing architecture reduces reliance on real-time cloud platform calculations, allowing for water consumption calculation even when communication links are unstable or temporarily interrupted, thus improving system stability.

[0083] Furthermore, in response to the high-frequency field pain point of flowmeter failure, an evidence-based alternative estimation mechanism based on dynamic efficiency coefficient is proposed. A complete anomaly marking system ensures the traceability of accounting data. Compared with the estimation method using fixed coefficients at the initial installation stage, it can effectively reduce the systematic deviation introduced by efficiency drift and ensure the continuous availability of the well-electric dual control system and the relative reliability of the accounting results in the case of flowmeter failure.

[0084] Therefore, by combining operational phase identification, non-pumping energy consumption elimination, dynamic updating of pumping efficiency, and edge computing accounting mechanism, this invention achieves refined metering of water consumption for well pumping, thereby improving the metering accuracy and system reliability of the well-electric dual control system. Attached Figure Description

[0085] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. Obviously, the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without creative effort. In the drawings:

[0086] Figure 1 This is a flowchart of a well power dual-control intelligent metering and accounting method based on edge computing;

[0087] Figure 2 This is an architecture diagram of a well-electricity dual-control intelligent metering and accounting system based on edge computing. Detailed Implementation

[0088] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0089] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0090] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0091] Example 1

[0092] This embodiment provides a deployment method for a well power dual-control intelligent metering and accounting system based on edge computing. (Architecture reference) Figures 1-2 As shown.

[0093] The system of this invention was deployed at an agricultural irrigation well site. The well is used for drawing water for farmland irrigation. The well is equipped with an electric water pump as a pumping device to extract groundwater to the surface irrigation network.

[0094] In this embodiment, the well pump is a three-phase submersible pump with a rated power of 15kW and a well depth of 80m. The well pump is connected to a three-phase power supply through an electrical control cabinet, and its start and stop are controlled by a control switch.

[0095] A dual-control intelligent metering and accounting system for well electricity is installed at the well site. This system includes:

[0096] Wellhead data acquisition module, edge computing node, communication module, and cloud platform.

[0097] The wellhead data acquisition module is installed inside the well electrical control cabinet and is used to collect electrical parameters and pumping parameters during the operation of the well pump.

[0098] The wellhead data acquisition module includes:

[0099] Electricity metering device, flow metering device, voltage acquisition unit, and current acquisition unit.

[0100] in:

[0101] The power metering device is installed in the well pump power supply circuit to acquire power data and active power data in real time during the well pump operation.

[0102] The voltage acquisition unit and the current acquisition unit are used to acquire three-phase voltage data and three-phase current data during the operation of the well pump;

[0103] The flow metering device is installed on the pumping pipeline to collect instantaneous flow data and cumulative water volume data during the pumping process.

[0104] The wellhead data acquisition module collects the above-mentioned operational data according to a preset sampling period and sends the collected operational data to the edge computing node.

[0105] In this embodiment, the sampling period can be set to 1 second to 5 seconds.

[0106] The edge computing node is located inside the well control cabinet and is connected to the wellhead data acquisition module to receive well pump operation data and process the operation data locally.

[0107] The edge computing node includes a processor, memory, and communication interface.

[0108] The processor is used to execute program instructions stored in memory, thereby enabling well pump operation data analysis and water consumption calculation functions.

[0109] The memory is used to store runtime data and system programs.

[0110] The communication interface is used to enable data interaction between the edge computing node, the wellhead data acquisition module, and the communication module.

[0111] In this embodiment, the edge computing node can be implemented using embedded industrial control equipment, such as an industrial controller based on an ARM architecture processor.

[0112] The following functional units are implemented internally by program modules in an edge computing node:

[0113] The system includes a data acquisition unit, an operation status identification unit, a startup energy consumption identification unit, a pumping stability section identification unit, a hydroelectric coupling calculation unit, a dynamic calibration unit, and a local accounting unit.

[0114] The aforementioned functional units form a continuous data processing flow through program logic, thereby enabling real-time analysis of well pump operation data and calculation of pumping water consumption.

[0115] The communication module is used to enable data communication between edge computing nodes and the cloud platform.

[0116] In this embodiment, the communication module can be implemented using wireless communication, such as an NB-IoT communication module, a 4G communication module, or a LoRa communication module.

[0117] The communication module connects to the edge computing node and is used to send the well pump operation data and water usage calculation results generated by the edge computing node to the cloud platform.

[0118] The cloud platform is used for centralized management of operational data and water usage calculation results from multiple wells.

[0119] The cloud platform includes a data storage module and a data management module.

[0120] The data storage module is used to store the runtime data uploaded by the edge computing nodes and the water usage calculation results;

[0121] The data management module is used to perform statistical analysis on well operation data and to monitor the well operation status and manage water usage data.

[0122] During system operation, the wellhead data acquisition module periodically collects well pump operation data and sends the collected operation data to the edge computing node.

[0123] After receiving the operational data, the edge computing node processes the data locally and divides the pumping process into stages based on the well pump operation status identification results, thereby realizing intelligent accounting of pumping water consumption.

[0124] The generated operational data and water usage calculation results are uploaded to the cloud platform for centralized management via the communication module.

[0125] Through the above system structure, this embodiment can realize real-time analysis of the well pumping process and water consumption calculation through edge computing nodes at the well site, thereby forming an edge intelligent metering system for well and electricity dual control application scenarios.

[0126] In this embodiment, in the event of a malfunction or data anomaly in the flow metering device, the edge computing node can utilize the continuously updated efficiency coefficients in the dynamic calibration unit. A well-founded alternative estimate of water consumption for pumping is required.

[0127] Specifically, when the edge computing node detects that the instantaneous flow data reported by the flow metering device is continuously zero, missing, or exceeds a reasonable range, it determines that the flow metering device is in an abnormal state and marks the pumping cycle as a "flow meter abnormal cycle".

[0128] During the abnormal flow meter cycle, the edge computing node no longer calculates the pumping efficiency observation value based on the measured flow data, nor does it adjust the efficiency coefficient. Instead of performing the current cycle update, it uses the most recent effective efficiency coefficient stored in the dynamic calibration unit. And based on the effective electricity obtained from the stable pumping phase The alternative estimated water consumption is calculated according to the following formula:

[0129] ;

[0130] in This represents the alternative estimated water consumption for the nth pumping cycle under the condition of a flow meter malfunction. The efficiency coefficient is updated from the previous effective period. This represents the effective electricity generated during the stable pumping phase of this cycle.

[0131] because Based on the multi-cycle recursive results of measured water output and effective electrical energy during the stable pumping phase during normal operation of the flow meter, this fully reflects the actual pumping efficiency of the well under the current operating conditions. Therefore, it is based on... The results of the alternative estimation have clear operating conditions as a basis, and compared with the estimation using the initial fixed coefficient, it can effectively reduce the systematic deviation introduced by efficiency drift.

[0132] When generating water usage accounting results, the local accounting unit uses alternative estimated water usage. An anomaly marker is added to distinguish the calculation results from those under normal flow meter conditions. This marker is transmitted synchronously when the communication module uploads the data to the cloud platform, so that the cloud platform can classify and manage the calculation data for abnormal periods and conduct manual verification.

[0133] In this embodiment, the edge computing node identifies the well pump operating status based on the well pump operating data collected by the wellhead data acquisition module, and divides the well pump operating process into the start-up stage, the stable pumping stage, and the shutdown stage based on the identification results.

[0134] Well pump operating data includes: voltage data; current data; active power data; instantaneous flow rate data; and electrical energy data.

[0135] The above data is collected by the wellhead data acquisition module according to the preset sampling period and formed into a well pump operation data sequence according to the acquisition time sequence.

[0136] In this embodiment, the sampling period can be set to 1 second.

[0137] Edge computing nodes identify the well pump startup phase by analyzing current variation characteristics in the running data sequence.

[0138] Specifically, within each sampling period, the difference between the current value at the current sampling moment and the current value at the previous sampling moment is calculated, and the rate of change of current is calculated based on the sampling period.

[0139] When the detected rate of change of current exceeds the preset rate of change threshold, the edge computing node determines that the well pump has entered the startup phase.

[0140] In a preferred embodiment, when the rate of change of current is greater than a preset proportion of the rated current, it can be determined that the well pump has entered the start-up stage.

[0141] For example, when the rate of change of current exceeds 30% per second of the rated current, the well pump can be considered to be in the startup process.

[0142] Furthermore, during the well pump startup phase, a rapid increase in current often occurs before the flow rate has stabilized. Therefore, detecting a rapid increase in current and an instantaneous flow rate below the minimum flow threshold can also serve as an auxiliary condition to confirm the startup phase.

[0143] After the well pump has started up and entered a stable operating state, the well pump operating parameters usually remain relatively stable.

[0144] Edge computing nodes continuously monitor the operational data sequence and identify stable pumping phases based on the following criteria:

[0145] Within multiple consecutive sampling periods: the current fluctuation amplitude is less than the first preset threshold; the active power fluctuation amplitude is less than the second preset threshold; and the instantaneous flow rate is greater than the minimum flow rate threshold.

[0146] When the above conditions are met continuously within the preset time window, the edge computing node determines that the well pump has entered the stable pumping stage.

[0147] In one embodiment, the time window can be set to 30 to 120 seconds.

[0148] After identifying the stable pumping phase, the edge computing node determines the corresponding operational data of this phase as valid pumping data and uses it for subsequent hydroelectric coupling calculations.

[0149] When the well pump stops running, the operating current of the well pump will drop rapidly to near zero.

[0150] Edge computing nodes identify downtime phases by monitoring current values ​​in the running data sequence.

[0151] When the detected current value is lower than the minimum operating current threshold and continues for multiple sampling cycles, the edge computing node determines that the well pump has entered the shutdown phase.

[0152] During the shutdown phase, the well pump does not generate pumping flow, therefore the corresponding operating data during this phase will not be included in the pumping efficiency calculation.

[0153] Through the above-described operational status identification process, the edge computing node can divide the well pump operation process into:

[0154] The process consists of three phases: startup, stabilization pumping, and shutdown.

[0155] In subsequent data processing:

[0156] The electricity consumption during the startup phase is identified as non-pumping energy consumption; the operating data during the stable pumping phase is used for pumping efficiency calculation and water consumption accounting; the data during the shutdown phase is not included in the pumping calculation.

[0157] Through the above-mentioned operational phase identification mechanism, this embodiment can provide an accurate data foundation for subsequent hydropower coupling calculations, thereby improving the accuracy of water pumping and water consumption calculations.

[0158] In this embodiment, after identifying the stable pumping stage of the well pump, the edge computing node performs statistical processing on the operating data of the stable pumping stage and establishes a water-electricity coupling relationship based on the statistical results, thereby realizing intelligent calculation of water consumption.

[0159] After identifying the stable pumping phase, the edge computing node uses the corresponding operating time period of the stable pumping phase as the data statistics window to perform statistical processing on the operating data within that time period.

[0160] Specifically:

[0161] Edge computing nodes perform integral calculations on the instantaneous flow data during the stable pumping phase to obtain the output water volume during the stable pumping phase.

[0162] Meanwhile, the electricity data during the stable pumping phase are accumulated and statistically analyzed to obtain the effective electricity during the stable pumping phase.

[0163] In this embodiment:

[0164] The outflow during the stable pumping phase is recorded as follows: The effective electricity generated during the stable pumping phase is recorded as follows: ; where n represents the current pumping cycle.

[0165] Edge computing nodes calculate pumping efficiency observations based on the water output and effective power generation during the stable pumping phase.

[0166] In this embodiment, the observed pumping efficiency can be expressed as:

[0167] ;

[0168] in:

[0169] This represents the observed pumping efficiency value for the nth pumping cycle.

[0170] This indicates the cumulative water output during the stable pumping phase of the pumping cycle.

[0171] This indicates the cumulative effective electricity generated during the stable pumping phase of the pumping cycle.

[0172] Pumping efficiency observations are used to reflect the pumping capacity of a well pump per unit of electricity during the current pumping cycle.

[0173] Because factors such as changes in groundwater level, pump operating status, and pipeline resistance can cause pumping efficiency to vary between different pumping cycles, it is necessary to dynamically update the pumping efficiency parameters.

[0174] In this embodiment, the edge computing nodes update the pumping efficiency parameters using a recursive method.

[0175] Specifically, the pumping efficiency observation value of the current pumping cycle is weighted and fused with the efficiency coefficient of the previous pumping cycle to obtain the efficiency coefficient of the current pumping cycle.

[0176] In one implementation, the efficiency coefficient update can be expressed as:

[0177] ;

[0178] in:

[0179] This represents the efficiency coefficient after the nth pumping cycle.

[0180] This represents the efficiency coefficient for the (n-1)th pumping cycle;

[0181] This represents the observed pumping efficiency value for the current pumping cycle.

[0182] α represents the smoothing coefficient, and satisfies: 0 < α < 1.

[0183] The smoothing coefficient is used to adjust the weighting relationship between historical efficiency coefficients and current observations.

[0184] When α is large, the updated efficiency coefficient depends more on the historical efficiency coefficient, thus making the efficiency parameter change more stable;

[0185] When α is small, the updated efficiency coefficient depends more on the observations of the current pumping cycle, thus reflecting changes in the well pump's operating status more quickly.

[0186] In one embodiment, the smoothing coefficient α can be set between 0.6 and 0.9.

[0187] After obtaining the updated efficiency coefficient Then, the edge computing nodes calculate the water consumption for this pumping cycle based on the effective power generation during the stable pumping phase.

[0188] In one embodiment, the water consumption for this pumping cycle can be expressed as:

[0189] ;

[0190] in: This represents the water consumption during the nth pumping cycle. The calculated water consumption result can be stored locally by the edge computing node and uploaded to the cloud platform for unified management via the communication module.

[0191] Through the above-mentioned hydroelectric coupling calculation and efficiency recursive update mechanism, this embodiment can dynamically adjust the pumping efficiency parameters according to the actual operation of the well pump, so that the hydroelectric coupling relationship can adapt to changes in groundwater level and equipment operating status, thereby improving the accuracy of pumping water consumption calculation.

[0192] In this embodiment, the edge computing node processes the well pump operation data in real time at the well site and completes the local calculation of water consumption. At the same time, it realizes data synchronization with the cloud platform through the communication module.

[0193] During system operation, the wellhead data acquisition module collects well pump operation data according to a preset sampling period and sends the collected operation data to the edge computing node.

[0194] After receiving the running data, the edge computing node writes the running data into the running data sequence according to the time order, and then executes the following processing flow in sequence:

[0195] First, the operation status identification unit identifies the well pump operation status based on the current change rate, power change amplitude, and flow change characteristics in the operation data sequence, and divides the well pump operation process into the start-up stage, the stable pumping stage, and the shutdown stage.

[0196] Secondly, when the startup energy consumption identification unit identifies the startup phase, it calculates the electricity generated during the startup phase and marks this portion of electricity as non-pumping energy consumption.

[0197] Subsequently, when the pumping stability stage is identified, the pumping stability stage identification unit determines the data statistics window corresponding to the stable pumping stage and performs statistical processing on the operating data within that time window.

[0198] The hydroelectric coupling calculation unit calculates the pumping efficiency observation value based on the water output and effective power generation during the stable pumping stage, and uses the pumping efficiency observation value as the input parameter of the pumping efficiency model.

[0199] The dynamic calibration unit recursively updates the pumping efficiency parameters based on the efficiency observations of the current pumping cycle and the efficiency coefficients of historical pumping cycles.

[0200] Finally, the local accounting unit calculates the water consumption for this pumping cycle based on the updated efficiency coefficient and the effective power generation during the stable pumping phase, and generates the corresponding water consumption accounting results.

[0201] In this embodiment, the edge computing node also includes a local data cache unit.

[0202] The local data cache unit is used to store well pump operation data sequences and generated water usage calculation results.

[0203] During each sampling period, the edge computing node writes the collected running data into the local data cache unit and saves the corresponding timestamp information in chronological order.

[0204] Meanwhile, after completing the water consumption calculation for the pumping cycle, the local accounting unit writes the generated water consumption calculation results to the local data cache unit for persistent storage.

[0205] Through the above mechanism, even if the communication link is temporarily unavailable, the edge computing node can still complete the pumping metering and accounting tasks locally.

[0206] The communication module is used to enable data communication between edge computing nodes and the cloud platform.

[0207] When the communication connection is normal, the communication module uploads the running data cached by the edge computing node and the water usage calculation results to the cloud platform.

[0208] In one embodiment, the communication module can upload locally cached data in batches at preset time intervals.

[0209] In another embodiment, the communication module can upload the corresponding accounting results immediately after the pumping cycle is detected to have ended.

[0210] After receiving the data uploaded by the edge computing node, the cloud platform writes the data into the cloud platform's data storage module, and the data management module manages the data in a unified manner.

[0211] In the event that the communication link is temporarily unavailable, the edge computing node continues to perform well pump operation data processing and water consumption calculation locally.

[0212] After communication is restored, the communication module uploads the data that was not uploaded from the local data cache unit to the cloud platform in chronological order, thereby achieving data retransmission.

[0213] Through the above mechanism, this embodiment can ensure that the system can still complete water pumping metering and water usage calculation normally even when communication is unstable.

[0214] Example 2

[0215] This embodiment provides a set of optional parameter configurations and an example of a complete pumping cycle. It also provides a more in-depth explanation of the mechanisms behind the formulas used in the hydroelectric coupling calculation and efficiency recursive update, illustrating the feasibility and stability of the algorithm on edge computing nodes. The following parameters are merely examples of one or a preferred implementation method; relevant thresholds and coefficients can be adjusted according to well type, pump model, pipeline conditions, and sensor accuracy.

[0216] The sampling period Δt of the edge computing node can be set to 1 second; voltage, current, active power, and instantaneous flow rate form a time-sorted data record in each sampling period, and the record must contain at least the timestamp t, three-phase current I(t), active power P(t), instantaneous flow rate q(t), and cumulative energy consumption. With cumulative water volume To facilitate stable operation at the edge, the edge computing nodes perform lightweight preprocessing on the collected data: short-window moving averages are used for current and power to suppress transient noise (e.g., window length of 3 to 5 sampling points), and jitter reduction processing is used for flow signals to suppress flow meter pulse jitter, while retaining the original values ​​for traceability.

[0217] The start-up phase identification uses a joint criterion of current change rate and current amplitude: calculation and will The current is compared with a preset threshold; simultaneously, I(t) is compared with a multiple threshold of the rated current. For example, the current change rate threshold can be 0.3 / s of the rated current, and the current multiple threshold can be 3 times the rated current. The engineering implication is that during the stable pumping phase, the motor enters a stable load range, and the statistical fluctuations in current and power are significantly smaller than during the start-up and shutdown transition phases. Simultaneously, the flow signal is positive and continuous, avoiding the inclusion of data from motor operation without water output or intermittent water output into the effective pumping statistics.

[0218] Within a pumping cycle, edge computing nodes begin accumulating startup energy consumption upon identifying the startup phase, and begin accumulating effective pumping energy consumption and effective water output upon identifying the stable pumping phase. The startup energy consumption can be calculated using discrete integration of the power sequence: The effective electricity during the stable pumping phase is also obtained using discrete integration or by directly calculating the difference in accumulated electricity from the meter: ,in and The start and end timestamps are used to define the statistical window for the stable pumping phase. The outflow rate is calculated using the discrete integral of the flow rate sequence. Alternatively, the cumulative water volume difference can be used: By strictly limiting the statistical window to the stable pumping phase, the algorithm achieves the constraint of "modeling only the effective pumping section" at the data level, thereby reducing the contamination of the model by unstable factors such as start-up shock, shutdown and back-pumping, and valve vibration from the source.

[0219] To obtain the effective power during the stable pumping phase of the current pumping cycle. With water output Then, the edge computing nodes calculate the observed pumping efficiency. This observation, in a physical sense, corresponds to "the effective water output per unit of electrical energy," and its reciprocal can be understood as the electrical energy required per unit of water output. Since the energy chain of a well pump system can be simplified as: electrical energy → motor mechanical energy → pump hydraulic energy → pipeline losses → effective water output, motor efficiency and pump efficiency can be considered to change slowly over a short timescale. However, factors such as groundwater level, pipeline resistance, valve opening, and grid voltage fluctuations can cause changes. Fluctuations occur between different pumping cycles. Especially in agricultural irrigation, the pumping cycle often exhibits a mixed pattern of "high-frequency short pumping / low-frequency long pumping," where factors such as flow meter pulsation, sand content in the well, and cavitation boundaries can cause fluctuations in a single pumping cycle. It includes measurement noise and operating noise. If directly... Using the conversion factor for the next cycle will cause the factor to oscillate frequently, making the accounting results unstable and even causing business risks such as "excessive differences in the accounting caliber of two adjacent pumping operations for the same farmer".

[0220] To address this, the present invention introduces a recursive update mechanism on edge computing nodes to improve efficiency coefficients. As a hydroelectric coupling parameter that evolves sustainably, and updated using an exponential smoothing method: Where 0 < α < 1. The underlying function of this formula can be understood on three levels. First, from a signal processing perspective, this recursion is essentially a first-order low-pass filter (a discrete form of exponential moving average) for the observed sequence. Smoothing is performed to suppress high-frequency noise and preserve low-frequency trends; the larger the α, the stronger the filtering and the more stable the output. Secondly, from a statistical estimation perspective... This can be considered as true efficiency. Superimposed noise Observations: ,in This represents sensor error and short-term operating condition disturbances; the recursive update is equivalent to weighted fusion of historical estimates and new observations in the online estimation process that continuously introduces new samples, making... This provides a robust estimate of θ, avoiding the dominance of parameters by a single anomalous observation. Third, from a system dynamics perspective, the actual changes in well pump efficiency are usually "slowly variable," such as intraday water level changes, seasonal declines, and pump wear accumulation, which do not abruptly change within a cycle. Exponential smoothing perfectly matches this slow-variability assumption, controlling the memory length through the "forgetting factor" α, which can track slow drifts without being overly sensitive to transient disturbances.

[0221] The recurrence relation can be expanded as follows: It is evident that the weight of older data decays exponentially with α, resulting in a finite effective memory length. In engineering, the effective memory length can be approximated as being on the order of 1 / (1-α) cycles. For example, when α=0.8, it is approximately 5 cycles, and when α=0.9, it is approximately 10 cycles. This provides an interpretable basis for parameter selection for different irrigation patterns.

[0222] To further ensure the numerical stability and service availability of recursive updates at the edge, this embodiment presents a set of preferred constraint processing methods, all of which are engineering implementations of the same recursive mechanism: when Below the minimum energy consumption threshold The system is deemed to have insufficient valid samples for this pumping cycle, and therefore skips the update for this cycle. = ;when When the water flow rate is below the minimum discharge threshold and the current is running simultaneously, this cycle can be marked as "suspected no-flow power consumption" and not used for updates. To avoid introducing sensor failures or pipeline anomalies into the model;

[0223] when relatively When the deviation exceeds a preset proportional threshold, "limited update" can be used, that is... The data is truncated to an acceptable range before being used in the recursion to resist obvious outliers. The underlying motivation for the above constraints is to ensure the robustness of the online estimation process: in situations where resources are limited on the edge and field conditions are complex, priority should be given to ensuring that the model is not broken down by outlier data, and only then should a fast response be considered.

[0224] After obtaining the updated efficiency coefficient Subsequently, this embodiment adopts Calculate the water consumption for this pumping cycle, of which To stabilize the effective power consumption during the pumping phase. The term "effective power consumption" is used because the energy consumption during the startup phase has been identified and labeled as non-pumping energy consumption. This separates the "energy consumption for generating effective water output" from the "startup impact / transition losses" in terms of energy scope, preventing the mapping of non-pumping energy consumption to false water output. This closed-loop approach ensures... Statistically, this is closer to "water output per unit of electricity under stable pumping conditions," thus improving cross-cycle portability. Furthermore, because... It is derived recursively from observations during the steady-state phase that when the steady-state operating conditions of the well pump change slowly (e.g., a gradual drop in water level leading to...) (gradually decrease) It will then descend smoothly, making The calculations remain consistent with real-world conditions over long timescales without experiencing drastic fluctuations due to measurement noise in a single cycle.

[0225] This embodiment further provides example data for a single pumping cycle to illustrate how the above mechanism operates at the edge. The edge computing node identifies a stable pumping phase lasting 600 seconds within a certain pumping cycle, and statistically obtains... , =5.4m³, then If the previous cycle =2.8m³ / kWh, take After the update Therefore, the water consumption for this period is calculated. If the next cycle results in a drop in water level... ,get ,but As can be seen, the coefficient is gradually reduced as the working conditions change, and the calculation results change continuously and smoothly, avoiding long-term deviations caused by using fixed conversion coefficients, and also avoiding short-term fluctuations caused by directly using η.

[0226] In terms of edge implementation, recursive updates only need to save the previous cycle. Compared with the statistics obtained in this period The calculation can be completed in a single operation with constant storage and computational overhead, making it suitable for real-time operation on embedded edge devices at the well site. At the end of each pumping cycle, the edge computing node stores the start and end timestamps of the cycle, the statistical window for the stable pumping phase, and... as well as The data is written to a local data cache unit and synchronized to the cloud platform in chronological order when the communication link is available, achieving consistent data links and accounting traceability between the end and the cloud. Through the above parameter configuration, data statistics method, and recursive update mechanism, this embodiment demonstrates the complete operating loop of the system of the present invention in a real well scenario, and explains the underlying mechanism of the formula in suppressing noise, tracking slow changes, and maintaining accounting stability.

[0227] Example 3

[0228] This embodiment aims to demonstrate a special case. Based on Embodiment 1 and Embodiment 2, it provides a detailed explanation of the system operation mechanism when the flow metering device at the well site malfunctions or experiences abnormal data, in order to demonstrate the complete processing flow and feasibility of the present invention in the case of flow meter malfunction.

[0229] In this embodiment, the edge computing node monitors the instantaneous flow data reported by the flow metering device in real time and determines the abnormal state of the flow meter by combining current and power data.

[0230] Specifically, the edge computing nodes perform a validity check on the instantaneous traffic data in each sampling period. The validity check conditions include the following:

[0231] Condition 1: Continuous Zero Flow Rate Detection. When the edge computing node detects that the instantaneous flow rate data is continuously zero, and the current data shows that the well pump is in a stable operating state during the same time period (i.e., the current value is higher than the minimum operating current threshold and the current fluctuation amplitude meets the stable operating conditions), it is determined that the flow meter may be stuck or in a signal interruption state. The continuous judgment time window for the above condition can be set to 60 seconds to 180 seconds to avoid misjudging the brief period of no flow rate in the early stage of pumping before pipeline pressure is established as a flow meter malfunction.

[0232] Condition 2: Detection of missing traffic data. When the edge computing node fails to receive valid data reports from the traffic metering device within multiple consecutive sampling periods, and the preset communication timeout threshold is exceeded, the traffic metering device is determined to be in a state of communication interruption or abnormal power supply.

[0233] Condition 3: Flow data exceeding range anomaly determination. When the edge computing node detects that the instantaneous flow data exceeds the rated range limit of the well flow meter, or that the instantaneous flow data is negative, the flow metering device is determined to be in an abnormal reading state.

[0234] Condition 4: Flow rate and power consistency verification. During the stable pumping phase, there is a physical constraint relationship between the pumping power and the outflow rate of the well pump, meaning that their trends are correlated under normal operating conditions. When the edge computing node detects that the active power is within the normal stable operating range, but the instantaneous flow rate reading is significantly lower than the normal range under historical operating conditions for an extended period, this situation can be used as an auxiliary criterion for abnormally low flow meter readings.

[0235] In this embodiment, when any of the above conditions is triggered and continues to meet the preset judgment time, the edge computing node marks the current pumping cycle as "flow meter abnormal cycle" and records the abnormal trigger timestamp, abnormal type identifier and flow reading at the time of trigger in the local data cache unit.

[0236] During the abnormal flow meter cycle, the edge computing node switches to alternative estimation mode. The specific processing procedure is as follows.

[0237] When an abnormal state of the flow meter is triggered, the edge computing node stops calculating the pumping efficiency observation value for the current pumping cycle, and also does not update the efficiency coefficient stored in the dynamic calibration unit for the current cycle. Instead, it directly uses the most recent effective efficiency coefficient stored in the dynamic calibration unit. This serves as the basis for the alternative estimation in this cycle.

[0238] The reason for continuing to use The reason for not using a fixed coefficient is as follows: During normal operation of the flow meter, the results of multiple pumping cycles based on the measured water output and effective power during the stable pumping phase fully reflect the actual pumping efficiency level of the well in the current operating phase. The deviation between the flow meter and the current true efficiency is usually much smaller than the deviation between the fixed coefficient at the initial installation stage and the current efficiency. This advantage is even more significant when the well has been running for a long time and the groundwater level has changed significantly.

[0239] During the abnormal flow meter cycle, the edge computing node still identifies the well pump's operating status according to the normal process, and distinguishes between the start-up phase and the stable pumping phase based on current and power data. After identifying the stable pumping phase, the edge computing node accumulates and statistically analyzes the power data during the stable pumping phase to obtain the effective power during the stable pumping phase. At the same time, the starting power generated during the startup phase will be used to... Continue to mark as non-pumping energy consumption and remove from the list. The above power consumption statistics process is unrelated to whether the flow meter is functioning properly and is not affected by any abnormal flow meter conditions; therefore, the effective power consumption... The calculation results remain reliable even in the event of a flow meter failure.

[0240] Effective electricity generated during the stable pumping phase Then, the edge computing nodes calculate the alternative estimated water consumption for abnormal flow meter cycles using the following formula:

[0241] ;

[0242] in This represents the alternative estimated water consumption for the nth pumping cycle under the condition of a flow meter malfunction. The most recent effective efficiency coefficient stored in the dynamic calibration unit. This represents the effective electricity generated during the stable pumping phase of this cycle.

[0243] When the flow metering device is in an abnormal state for multiple consecutive pumping cycles, the edge computing node uses the efficiency coefficient stored in the current dynamic calibration unit for replacement estimation in each abnormal cycle, and does not update the efficiency coefficient during the period of abnormal state.

[0244] Considering the time-sensitivity of efficiency coefficients, when the number of consecutive abnormal cycles exceeds a preset threshold (e.g., 10 consecutive pumping cycles), the edge computing node can record an "efficiency coefficient timeliness alarm" in its local data cache unit. This indicates that the flow meter of the well has been malfunctioning for an extended period, and the reliability of the current alternative estimation results has decreased over time, requiring manual intervention for equipment maintenance or coefficient recalibration. The alarm information is uploaded synchronously when the communication module establishes a connection with the cloud platform, triggering a corresponding maintenance work order from the cloud platform's data management module.

[0245] When the edge computing node detects that the flow metering device has resumed reporting valid flow data and the flow data passes the reasonableness check, the edge computing node automatically exits the alternative estimation mode and resumes the normal pumping efficiency observation calculation and efficiency coefficient recursive update process at the beginning of the next complete pumping cycle. The efficiency observation value obtained in the first complete pumping cycle after the flow meter resumes operation is... will with The efficiency coefficient is updated by weighted fusion according to the preset smoothing coefficient α, so that the efficiency coefficient can smoothly transition to the normal update state after the flow meter is restored, and avoid the observation value of the first cycle after restoration from having too large an impact on the efficiency coefficient due to the possible short-term instability when the flow meter is put back into use.

[0246] When generating water usage accounting results, the local accounting unit uses alternative estimated water usage for abnormal flow meter cycles. An additional exception flag field must be provided, which must contain at least the following information:

[0247] Anomaly type identifier is used to distinguish different fault types such as persistent zero-value anomalies, missing data, over-range anomalies, and consistency check anomalies; anomaly start and end timestamps are used to record the start and end times of the flowmeter's abnormal state; efficiency coefficient values ​​are retained. The information includes the most recent valid update time, to verify the timeliness of the efficiency parameters on which the alternative estimation is based; and an alternative estimation flag, used to distinguish it from the calculation results generated based on measured flow data under normal flow meter conditions.

[0248] The above additional marker fields and alternative estimated water consumption The data is written to the local data cache unit for persistent storage and is simultaneously transmitted to the cloud platform data storage module when the communication module uploads the data to the cloud platform. After receiving the accounting data with the alternative estimation identifier, the cloud platform data management module classifies and manages this data separately and supports manual review and data correction operations, thereby ensuring the integrity and traceability of the groundwater use ledger.

[0249] To further illustrate the feasibility of the above mechanism, this embodiment provides a set of example data.

[0250] After a well has been operating normally for several pumping cycles, the current efficiency coefficient stored in the dynamic calibration unit... =2.75 m³ / kWh. At the start of the nth pumping cycle, the edge computing node detected that the instantaneous flow rate data remained zero, while the concurrent current data showed that the well pump was operating stably, with a current value of 92% of the rated current and a power fluctuation amplitude less than the preset power threshold, lasting for more than 120 seconds. Based on this, the edge computing node determined that the flow meter was in a stuck abnormal state, marked this cycle as a flow meter abnormal cycle, and switched to alternative estimation mode.

[0251] Edge computing nodes continue to identify the stable pumping phase of the current cycle and calculate the effective electricity generated during the stable pumping phase. =1.8kWh, power consumption during startup =0.05kWh has been marked as non-pumped energy consumption and removed.

[0252] Calculate the estimated water consumption for this period using the alternative estimation formula:

[0253] ;

[0254] Local accounting unit will =4.95m³, along with the anomaly type identifier "Continuous Zero Flow Value Anomaly", anomaly start and end timestamps, and the retained efficiency coefficient value of 2.75. The data, along with its most recent valid update time, is written to the local data cache unit. After establishing a connection with the cloud platform, the communication module synchronously uploads the above data. The cloud platform's data management module categorizes this accounting record into the "alternative estimation" category and generates a flow meter anomaly alarm, prompting maintenance personnel to inspect the well flow meter.

[0255] Through the above processing flow, this embodiment demonstrates how the system of the present invention, when the flow metering device malfunctions, relies on the dynamic efficiency coefficient. This enables a complete operational loop that achieves evidence-based alternative estimations and ensures the traceability of accounting data through a comprehensive anomaly marking mechanism. This ensures the continuous availability of the well-electric dual control system and the relative reliability of the accounting results in the high-frequency field pain point scenario of flowmeter failure.

[0256] Example 4

[0257] Based on Examples 1 to 3, this embodiment proposes a dual-state recursive calculation method that introduces efficiency trend state variables into the dynamic calibration unit to address the scenario where well pumping efficiency exhibits directional drift during long-term operation. This method enables edge computing nodes to explicitly model the direction and rate of efficiency change while tracking efficiency levels, thereby gaining the ability to perceive efficiency drift trends. Furthermore, it reduces the systematic bias introduced by trend lag in flowmeter fault replacement estimation scenarios.

[0258] In the systems described in Examples 1 to 3, the dynamic calibration unit updates the efficiency coefficient using a first-order exponential smoothing recursive formula:

[0259] ;

[0260] This recursive formula exhibits good noise suppression in scenarios with random efficiency fluctuations or slow, non-directional drift. However, in the following two representative operating conditions in agricultural irrigation wells, the above first-order recursive model suffers from structural limitations:

[0261] The first type of operating condition is the seasonal and continuous decline of the groundwater level. In arid northern regions, from the start of the peak irrigation season to the end of the season, the groundwater level may drop by several meters or even more than ten meters over several weeks. This leads to a continuous increase in pump head and a continuous decrease in the amount of water output per unit of electricity, resulting in a monotonically decreasing efficiency coefficient across multiple pumping cycles. In this scenario, the first-order recursive estimate lags behind the actual efficiency due to the smoothing effect, causing the system to continuously overestimate water consumption throughout the entire irrigation season.

[0262] The second type of operating condition is the cumulative wear of the pump impeller. When a pump operates in a sandy water environment for an extended period, impeller wear is an irreversible cumulative process. While the efficiency change within a single cycle is minimal, the cumulative efficiency drop over several months can reach 10% to 20%. In this scenario, because the efficiency observation value for each cycle differs very little from the efficiency coefficient of the previous cycle, first-order recursion is insufficient to detect the trend and cannot determine its sustainability, thus missing the opportunity to trigger maintenance and repairs through trend warnings.

[0263] The common characteristic of the two types of operating conditions mentioned above is that the change in efficiency is not random noise, but a directional and persistent trend drift. If only the efficiency level is estimated without modeling the trend, the estimated value of the system will always lag behind the true efficiency, and the lag will accumulate as the trend duration increases. Simply reducing the smoothing coefficient α can speed up the response, but at the same time it will reduce the ability to suppress random noise. There is an irreconcilable contradiction between the two objectives.

[0264] To address the aforementioned issues, this embodiment introduces a second state variable into the dynamic calibration unit, which is based on the original efficiency level estimate. Based on this, add an efficiency trend estimate. This constitutes a two-state description of the efficiency time series.

[0265] The meaning is consistent with that of Example 1, namely, the pumping efficiency of the well pump in the current pumping cycle (water output per unit of electricity, in units of...). The recursive estimate of ) is used as the conversion parameter for calculating water consumption in this cycle.

[0266] This is a newly introduced efficiency trend estimate in this embodiment, representing the change in efficiency level between adjacent pumping cycles, with the same unit. (per pumping cycle). This indicates that efficiency is on the rise. This indicates that efficiency is declining. The size reflects the strength of the trend.

[0267] During the system initialization phase The method for determining the efficiency is the same as in Example 1, that is, during the equipment installation and commissioning phase, data from the stable pumping stage of no less than three complete pumping cycles are collected, and the average of the efficiency observations for each cycle is used as the... Initial trend value Setting it to zero means that the system does not make any prior assumptions about the direction of efficiency changes in the initial state, and relies on the recursive accumulation of subsequent operating data to gradually build the ability to perceive trends.

[0268] After each steady-state pumping phase, the hydroelectric coupling calculation unit first calculates the efficiency observation value for the current pumping cycle according to the method in Example 1:

[0269] ;

[0270] in To stabilize the cumulative water output during the pumping phase, The calculation method for the cumulative effective electricity during the stable pumping phase is exactly the same as in Example 1.

[0271] Subsequently, the dynamic calibration unit completes the dual-state update sequentially according to the following two formulas.

[0272] Step 1, Efficiency Level Update:

[0273] ;

[0274] This formula incorporates historical efficiency coefficients in weighted fusion. Replace with That is, the sum of the efficiency level and trend of the previous cycle is used as the prior prediction of the efficiency of the current cycle, and then compared with the measured values. Weighted fusion is performed. The engineering implications of this modification are: when estimating the efficiency of the current cycle, the assumption that the efficiency is the same as that of the previous cycle (i.e., the zero-trend assumption) is no longer made. Instead, the trend direction identified in the previous cycle is incorporated into the prior prediction, so that the estimated value will approach the true efficiency in advance when the trend continues, thereby reducing lag error.

[0275] Step 2, Efficiency Trend Update:

[0276] ;

[0277] in The trend smoothing coefficient satisfies... Independent of efficiency level smoothing coefficient It is specifically designed to control the sensitivity and smoothness of trend estimation. This represents the actual change in efficiency level during the current period, which is used as the current trend observation to participate in the recursive update of the trend. The structure of this formula is consistent with the efficiency level recursive formula. Essentially, it performs first-order exponential smoothing on the trend quantity itself, so that the trend estimate can both follow continuous drift and not be overly sensitive to changes in efficiency level caused by single fluctuations.

[0278] Regarding parameter settings, The typical value range is 0.75 to 0.85, which is consistent with the recommended range of the smoothing coefficient in Example 1. The typical value range is 0.50 to 0.70, and it is intentionally set to be lower than 0.70. The reason is that the trend quantity changes more slowly and less than the efficiency level itself. This makes trend estimation more sensitive to recent changes and can quickly identify reversal signals when efficiency trends reverse (such as when efficiency recovers after pump maintenance or when groundwater levels recover at the end of the irrigation season), without carrying over the inertia of historical trends after the trend has changed.

[0279] After completing the dual-state update, the dynamic calibration unit generates a predicted value for the efficiency level of the next pumping cycle:

[0280] ;

[0281] The calculation requires no new sensor input, relying only on the two state variables updated in the current cycle, with a computational overhead of a single addition operation. This prediction has the following two engineering applications in this embodiment.

[0282] Firstly, it is used as a substitute estimate when the flow metering device malfunctions. In the flow meter malfunction scenario described in Example 3, the system uses the most recent effective efficiency coefficient for substitute estimation. In this example, the substitute estimation formula is modified as follows:

[0283] ;

[0284] Compared to Example 3, which directly uses This approach incorporates the trend direction identified in the previous period into the alternative estimation, ensuring that the estimated efficiency coefficient automatically falls below historical levels during periods of continuous efficiency decline and automatically rises above historical levels during periods of continuous efficiency recovery. This reduces systematic biases introduced by trend lag. The generated alternative estimated water consumption is also marked with anomalies, with the marking content consistent with that in Example 3, and is synchronously uploaded to the cloud platform after communication is restored.

[0285] Secondly, it is used for early warning of a continuous decline in efficiency. After each pumping cycle, the dynamic calibration unit updates the trend status quantity. Monitoring is being conducted. When... When the value of edge computing node generation efficiency continuously declines and an alarm is triggered for multiple consecutive pumping cycles, the following condition must be met: ;in The trend alarm threshold can be set at 1% to 3% of the historical average efficiency coefficient per cycle. To continuously determine the number of pumping cycles, a typical value is 3 to 5 pumping cycles, used to exclude temporary negative values ​​in trend estimates caused by underestimation of efficiency in a single instance. The continuous decline in efficiency is detected along with the current... , The specific values ​​and trigger timestamps are recorded in the local data cache unit and uploaded to the cloud platform when the communication link is available. This allows the data management module to classify and label the data and conduct manual verification, providing data for maintenance personnel to determine whether pump inspections or pipe dredging are necessary. This early warning mechanism relies entirely on the local computation of the edge computing nodes and does not depend on real-time responses from the cloud platform. It can continue to accumulate alarm records locally even during communication interruptions.

[0286] In the dual-state recursion mechanism, this embodiment also extends and adapts the robustness constraints described in Embodiment 2.

[0287] When the effective power Below the minimum energy consumption threshold If the current period's valid samples are insufficient, skip the update for this period. , Both state variables remain unchanged.

[0288] When efficiency observation Relative to prior predictions When the deviation exceeds a preset proportional threshold, After limiting and truncating the amplitude, it is used to participate in the recursion to prevent a single abnormal observation from simultaneously polluting both the efficiency level and the trend of the two state quantities.

[0289] When the flow meter is detected to be in an abnormal state as defined in Example 3, no updates to the dual-state quantities are made in this cycle. , This ensures that the continuity of trend estimation is not affected by flow meter failure.

[0290] Taking the scenario of a groundwater level continuously declining after the peak irrigation season begins at an agricultural irrigation well as an example, the basic parameters of the well remain consistent with those in Example 1: the well pump is a three-phase submersible pump with a rated power of 15 kW, the well depth is 80 m, and the sampling period is... The system's two-state recursive parameters are set as follows: initial efficiency level. Initial value of trend The efficiency level smoothing coefficient is 0. Trend smoothing coefficient Trend alert threshold Continuous determination of the number of cycles .

[0291] This example demonstrates four complete pumping cycles, each consisting of a startup phase, a stable pumping phase, and a shutdown phase. The following sections detail the data acquisition, statistical processing, and two-state recursion process for each cycle.

[0292] In the first cycle, the edge computing node detected that the well pump started at time t = 0 s. During the period from t = 0 s to t = 18 s, the rate of change of current exceeded 0.3 times the rated current per second, which was determined to be the start-up phase. The accumulated power during the start-up phase is shown below;

[0293] Startup power consumption: It is marked as non-pumping energy consumption and does not participate in hydroelectric coupling;

[0294] Starting at t = 18 s, the three-phase current fluctuation amplitude is continuously below 3% of the rated current for 60 s, the active power fluctuation amplitude is below 2% of the rated power, and the instantaneous flow rate is continuously above 2.0. The conditions for identifying the stable pumping phase are met, and the statistical window is determined as follows: to The duration is 2400 seconds.

[0295] Statistics during the stable pumping phase:

[0296] Effective battery capacity: ;

[0297] Water output: ;

[0298] Hydroelectric coupling calculation:

[0299] Efficiency observations: ;

[0300] Dynamic calibration unit dual-state recursion:

[0301] ;

[0302] Efficiency level update:

[0303]

[0304] Trend observations:

[0305] Trend Update:

[0306] ;

[0307] Water consumption for this period: ;

[0308] Write to the local data cache unit;

[0309] In the second cycle, due to the continued peak irrigation season, the groundwater level dropped further compared to the first cycle, and the pump head increased. The stable pumping phase in this cycle lasted for 2400 seconds. Statistical data showed that:

[0310] Effective battery capacity: ;

[0311] Water output: ;

[0312] Efficiency observations: ;

[0313] Dynamic calibration unit dual-state recursion:

[0314] Prior prediction:

[0315] Efficiency level update:

[0316]

[0317] Trend observations: ;

[0318] Trend Update:

[0319] ;

[0320] Water consumption for this period: ;

[0321] Write to the local data cache unit;

[0322] In the third cycle, the groundwater level dropped further. Statistics from the stable pumping phase of this cycle show:

[0323] Effective battery capacity: ;

[0324] Water output: Q3 = 5.960 m³;

[0325] Efficiency observations: ;

[0326] Dynamic calibration unit dual-state recursion:

[0327] Prior prediction: ;

[0328] Efficiency level update:

[0329]

[0330] Trend observations:

[0331] Trend Update:

[0332] ;

[0333] Water consumption for this period:

[0334] Write to the local data cache unit;

[0335] In the fourth cycle, the groundwater level continued to decline. Statistics from the stable pumping phase of this cycle show:

[0336] Effective energy capacity: E4 = 2.000 kWh;

[0337] Water output: ;

[0338] Efficiency observations:

[0339] Dynamic calibration unit dual-state recursion:

[0340] Prior prediction:

[0341] Efficiency level update:

[0342]

[0343] Trend observations:

[0344] Trend Update:

[0345] ;

[0346] Water consumption for this period:

[0347] Alarm triggered, The consecutive number of cycles reaches 3, satisfying the condition. = 3, triggering an alarm for continuous efficiency decline.

[0348] Edge computing nodes will write the following alarm records to their local data cache units: alarm type: continuous efficiency decline, trigger timestamp (end of the fourth cycle), current efficiency level. = 3.048 Current trend estimate = -0.042 The alarm triggers continuously for 3 cycles. Recommended maintenance actions include checking the groundwater level and pump operating status. This alarm record will be uploaded to the cloud platform data management module along with the accounting data when the communication link is available.

[0349] The complete recursive process for the four cycles is summarized in the table below. The column for prior predictions shows the predicted efficiency of the dynamic calibration unit for the current cycle after the introduction of the trend term. This can be compared with the actual efficiency observations to intuitively reflect the effect of the trend correction.

[0350] Periodic Summary Table:

[0351] ;

[0352] Assuming that 30 seconds after the start of the fourth cycle of stable pumping, the data reported by the flow metering device is interrupted, triggering the flow meter anomaly judgment condition as defined in Example 3, the edge computing node marks this cycle as the flow meter anomaly cycle, no longer calculates the efficiency observation value based on the actual water output, and does not update the dual state quantity.

[0353] This embodiment uses prior predictions with trend correction for alternative estimation:

[0354] Trend correction efficiency: ;

[0355] This embodiment uses an alternative estimation: ;

[0356] In Example 3, the substitution estimation result of the efficiency coefficient from the previous cycle is directly used as follows:

[0357] Example 3 Alternative Estimation:

[0358] The actual water consumption for this period is as follows (based on the calculation results when the flow meter is functioning normally):

[0359] Actual accounting results: ;

[0360] The following table shows a comparison of the deviations between the two alternative estimation methods and the actual accounting results:

[0361] Deviation Comparison Table:

[0362] ;

[0363] The estimation deviation in this embodiment is smaller than that in Embodiment 3. The relationship between the deviation reduction and the trend correction term is as follows:

[0364] Contribution of trend correction: ;

[0365] Trend correction term Contributed to alternative estimates The downward correction amount precisely corresponds to the efficiency reduction that should be reflected in this cycle due to the continuous downward trend in efficiency. If the flowmeter failure continues into the fifth or sixth cycle, and the efficiency continues to decline along the trend, the cumulative deviation difference between the two schemes will further widen. The accuracy advantage brought by the continuous correction capability of the trend term in this embodiment will accumulate linearly with the duration of the failure, and has more significant engineering value in long-term flowmeter failure scenarios.

[0366] Each pumping cycle includes two additional addition and subtraction operations (to calculate prior predictions). + and trend observations - In addition to one multiplication and one addition (for the recursive update of the trend state), a total of four floating-point operations are added. Regarding storage overhead, this adds to the existing efficiency level of floating-point operations. Based on this, add a trend state quantity. The additional storage overhead is a floating-point number. This incremental overhead is negligible for edge computing nodes commonly found in well sites, based on ARM Cortex-M series or equivalent embedded processors. It does not change the overall system architecture, add additional sensor configurations, or alter the data acquisition frequency and communication protocol. It can be implemented through software updates on existing hardware platforms.

[0367] At the end of each pumping cycle, the edge computing node will record the start and end timestamps of the cycle, the statistical window of the stable pumping phase, and... , as well as The data is written to the local data cache unit and synchronized to the cloud platform in chronological order when the communication link is available, so as to achieve consistent data link and accounting traceability between the end and the cloud.

[0368] By introducing the aforementioned dual-state recursive mechanism, edge computing nodes can achieve more accurate efficiency level estimation, more evidence-based flow meter fault replacement estimation, and early warning function for efficiency decline trends in long-term operation scenarios, thereby improving the accounting reliability and operation and maintenance response capability of the well-electric dual control system under complex irrigation conditions.

Claims

1. An edge-computing-based well electric dual-control intelligent metering and accounting method, characterized in that, Includes the following steps: S1. An edge computing node is set at the wellhead, and well pump operation data is collected through an energy metering device and a flow metering device. The operation data includes voltage, current, active power, electricity, instantaneous flow rate and cumulative water volume. S2. The edge computing node identifies the well pump operating status based on the joint criteria of current change rate, power fluctuation amplitude and flow change characteristics, and divides the well pump operation process into the start-up stage, stable pumping stage and shutdown stage. S3. When the start-up phase is identified, the electricity consumption during the start-up phase period is integrated or accumulated to obtain the start-up electricity consumption, and the start-up electricity consumption is marked as non-pumping energy consumption and removed from the total electricity consumption in subsequent calculations. S4. When the pumping phase is identified as a stable pumping phase, the electricity data within this time period is accumulated and statistically analyzed to obtain the effective electricity of the stable pumping phase; at the same time, the instantaneous flow data of this phase is integrated to obtain the water output of the stable pumping phase. S5. The ratio of water output to effective electricity during the stable pumping phase is determined as the observed pumping efficiency value for the current pumping cycle. S6. The efficiency coefficient is dynamically updated using an exponential smoothing recursive formula. S7. Calculate the water consumption for this pumping operation based on the updated pumping efficiency parameters and the effective power consumption, and complete the water consumption calculation and local storage in the edge computing node.

2. The intelligent metering and accounting method for well power dual control based on edge computing according to claim 1, characterized in that: In step S2, the identification conditions for the start-up phase are: the rate of change of current is greater than the preset rate of change threshold within a continuous sampling period, or the current value exceeds a preset multiple of the rated current; the identification conditions for the stable pumping phase are: within a preset time window, the current fluctuation amplitude is less than the first preset threshold, the active power fluctuation amplitude is less than the second preset threshold, and the instantaneous flow rate is continuously greater than the minimum flow rate threshold; the identification conditions for the shutdown phase are: the current value is lower than the minimum operating current threshold and continues for multiple sampling periods.

3. The intelligent metering and accounting method for well power dual control based on edge computing according to claim 1, characterized in that: In step S6, the efficiency coefficient initial value Determined by one of the following methods: During the equipment installation and commissioning phase, collect stable pumping phase data for no less than three complete pumping cycles, and use the average of the efficiency observations for each cycle as the mean. Alternatively, industry reference values ​​that match the well pump model and well depth can be used as... .

4. The intelligent metering and accounting method for well power dual control based on edge computing according to claim 1, characterized in that: Step S6 also includes the following robustness constraint handling: a) Effective electricity during the stable pumping phase Below the minimum energy consumption threshold If the current period's valid samples are insufficient, skip the update for this period. ; b) When the efficiency observation value Compared to When the deviation exceeds a preset proportional threshold, a limiting update is used. Cut off to the allowable range before participating in the recursion; c) When it is detected that the well pump is running but the water output is lower than the minimum water output threshold, the current cycle will be marked as suspected no-flow power consumption and will not be included in the efficiency coefficient update.

5. The intelligent metering and accounting method for well power dual control based on edge computing according to claim 1, characterized in that: In step S6, the smoothing coefficient The value range is from 0.6 to 0.9, corresponding to an effective memory length of... One pumping cycle.

6. The intelligent metering and accounting method for well power dual control based on edge computing according to claim 1, characterized in that, It also includes flow meter failure replacement estimation steps: When the edge computing node detects that the flow metering device meets any of the following abnormal conditions, it determines that the flow metering device is in a fault state and marks the current pumping cycle as a flow meter abnormal cycle: a) The instantaneous flow rate data remains zero, and the concurrent current data shows that the well pump is in a stable operating state for a duration exceeding the preset judgment time window; b) No valid data is received from the flow metering device for several consecutive sampling periods, exceeding the preset communication timeout threshold; c) Instantaneous flow rate data exceeds the upper limit of the rated range or shows a negative value; d) The active power is within the normal and stable operating range, while the flow rate reading is significantly lower than the normal range under the same historical operating conditions for an extended period of time; During the flow meter's abnormal cycle, the edge computing node uses the most recent effective efficiency coefficient stored in the dynamic calibration unit. Calculate the alternative estimated water consumption using the following formula: ; in To replace estimating water consumption, The effective electricity generation during the stable pumping phase of this cycle is calculated; an anomaly marker is added to the generated accounting results, and the anomaly marker includes at least the anomaly type identifier, the anomaly start and end timestamps, the efficiency coefficient value used and its most recent valid update time, and is uploaded to the cloud platform synchronously with the accounting results. When the number of consecutive abnormal cycles exceeds a preset threshold, the edge computing node records an efficiency coefficient timeliness alarm in the local data cache unit.

7. The intelligent metering and accounting method for well power dual control based on edge computing according to claim 1, characterized in that: The edge computing node is equipped with a local data caching unit and a local accounting unit. The local data cache unit is used to store the collected well pump operation data and corresponding time series information. When the communication module fails to establish a communication connection with the cloud platform, the local accounting unit independently executes steps S2 to S7 based on the locally cached running data, and persistently stores the accounting results in the local data cache unit. Once communication is restored, the results are automatically retransmitted to the cloud platform.

8. A well power dual-control intelligent metering and accounting system based on edge computing, characterized in that, include: The wellhead data acquisition module is used to collect well pump operating data, including voltage, current, active power, electricity, instantaneous flow rate, and cumulative water volume. An edge computing node, connected to the wellhead data acquisition module, is used to receive the operating data and identify the operating status of the well pump, dividing the well pump operation process into the start-up stage, the stable pumping stage, and the shutdown stage. The edge computing node is also used to calculate the startup power consumption during the startup phase and mark it as non-pumping energy consumption, and to count the effective power consumption and calculate the corresponding water output during the stable pumping phase, thereby establishing a hydroelectric coupling relationship model and updating the pumping efficiency parameters. The communication module, connected to the edge computing node, is used to upload the running data and calculation results to the cloud platform; The cloud platform, connected to the communication module, is used to store and manage the running data and calculation results.

9. The intelligent metering and accounting system for well power dual control based on edge computing according to claim 8, characterized in that: The edge computing node includes: a data acquisition unit, an operation status identification unit, a startup energy consumption identification unit, a pumping stable section identification unit, a water-electricity coupling calculation unit, a dynamic calibration unit, and a local accounting unit; the dynamic calibration unit uses an exponential smoothing recursive formula to dynamically update the efficiency coefficient; the local accounting unit independently completes the calculation of pumping water consumption based on local cached data when communication is interrupted.

Images