Distributed pumping unit group regenerative power generation and energy storage dispatching platform based on internet of things

Abstract

The application relates to the field of pumping unit power generation, and discloses a distributed pumping unit group regenerative power generation energy storage scheduling platform based on the Internet of Things, which comprises the following steps: in the operation process of a pumping unit, data is collected through a deployed multi-modal sensing terminal; when abnormal data is detected, a short-period compensation is performed on the abnormal data by using a time series interpolation and sample comparison mechanism to restore the integrity of the operation state; based on the power change trend, load response frequency and non-critical working condition characteristics in the pumping unit operation data, a time section with residual kinetic energy release is identified; according to the current state of the pumping unit and the available capacity of the energy storage equipment, an energy transmission path is dynamically allocated, and a multi-parameter scheduling criterion is adopted; corresponding control instructions are generated according to the scheduling results, energy recovery operations are performed, energy storage feedback data is collected, the execution effect is evaluated, scheduling parameters are adjusted, and a closed-loop optimization mechanism is formed. The application has the advantages of improving the energy recycling efficiency.

Description

Technical Field

[0001] This invention relates to the field of oil pumping unit power generation, specifically to a distributed oil pumping unit cluster regenerative power generation and energy storage scheduling platform based on the Internet of Things. Background Technology

[0002] Traditional oil pumping unit operation and management systems typically rely on manual inspections or timed control for equipment start-up, shutdown, and load adjustment, making it difficult to achieve unified scheduling and dynamic optimization of oil pumping unit groups across different areas. With the development of IoT technology, more and more oil pumping units are being connected to distributed energy systems. These systems use sensing terminals to collect real-time data on power status and operating cycles, and combine this data with regional energy storage units to achieve regenerative power generation and unified scheduling of remaining kinetic energy. However, existing systems generally suffer from a technical issue: due to the complex network environment in remote oilfields, data acquisition from some oil pumping units experiences delays or packet loss. This means that the operating status information used by the platform for regenerative power generation and energy storage scheduling calculations is not updated in real time, thus affecting the accuracy and timeliness of the scheduling strategy. Therefore, it is essential to design an IoT-based distributed oil pumping unit group regenerative power generation and energy storage scheduling platform to improve energy recovery and utilization efficiency. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides an IoT-based distributed oil pumping unit cluster regenerative power generation and energy storage scheduling platform, which has the advantage of improving energy recovery and utilization efficiency and solves the problems mentioned in the background technology.

[0004] To achieve the aforementioned goal of improving energy recovery and utilization efficiency, this invention provides the following technical solution: a distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform based on the Internet of Things, comprising:

[0005] Status awareness module: During the operation of the pumping unit, data is collected through deployed multimodal sensing terminals. When abnormal data is detected, it is determined whether status repair processing needs to be triggered. If so, the status repair module is entered.

[0006] State Repair Module: Utilizes time-series interpolation and sample comparison mechanisms to perform short-cycle compensation on abnormal data, restoring the integrity of the operating state. The confidence level is used to determine whether the compensation result is lower than the threshold. If so, the module proceeds to the energy efficiency identification module.

[0007] Energy efficiency identification module: Based on the power change trend, load response frequency and non-critical operating condition characteristics in the pumping unit operation data, it identifies the time period where there is residual kinetic energy release, and judges whether the identification result meets the conditions for regenerative power generation. If it does, it enters the collaborative scheduling module.

[0008] Coordinated scheduling module: Based on the current status of the pumping unit and the available capacity of the energy storage equipment, it dynamically allocates energy transmission paths, adopts multi-parameter scheduling criteria, and determines whether to execute the energy storage task. If so, it enters the energy storage control module.

[0009] Energy storage control module: Generates corresponding control commands based on scheduling results, executes energy recovery operations, collects energy storage feedback data, evaluates the execution effect, adjusts scheduling parameters, and forms a closed-loop optimization mechanism.

[0010] Preferably, the process of determining whether to trigger state repair processing is as follows:

[0011] After collecting the operating data of the oil pumping unit, the key parameters are judged for time-series consistency through a multi-dimensional anomaly detection algorithm, and the data anomaly score is calculated by combining the difference between the current data and the historical stable operating data.

[0012] If the anomaly score exceeds the set tolerance threshold, and the duration of consecutive anomaly segments exceeds the minimum processing cycle threshold, then state repair processing is triggered.

[0013] Preferably, the process of short-cycle compensation for outlier data using time-series interpolation and sample comparison mechanisms is as follows:

[0014] For data with detected abnormal segments, historical normal periodic data within the sliding window is retrieved, and local interpolation compensation is performed based on the sampling frequency, data periodicity, and the trend of adjacent point values.

[0015] The standard operating mode data stored in the feature template library is called, the residual distribution between the current data segment and similar patterns is compared, and the sample segment with the smallest residual is selected for weight fusion repair.

[0016] The final generated compensation data segment should maintain a continuous trend of change with the adjacent non-abnormal segments.

[0017] Preferably, the process of determining whether the compensation result is lower than the threshold based on the confidence level is as follows:

[0018] The similarity between the compensated data and the original data segments before and after is compared, and the data trend matching degree is calculated by using the sliding window dynamic time warping algorithm.

[0019] A confidence score is generated by combining the degree of deviation between the compensated value and the original sampled value in the local statistical distribution;

[0020] If the score is below the confidence threshold, it indicates that the compensation is unstable or fluctuates drastically.

[0021] Preferably, the process of determining whether the identification result meets the conditions for renewable power generation is as follows:

[0022] Extract the negative power segment per unit time, the transient instantaneous current drop rate, and the equipment inertial shedding response parameters from the power curve;

[0023] The three indicators are matched with the releasable kinetic energy indicators set in the threshold library. If the matching degree exceeds the set standard and there is a predicted short-term low load operation period, the current period is determined to meet the conditions for regenerative power generation.

[0024] The pumping unit is marked as a candidate device to participate in the scheduling and then enters the collaborative scheduling module.

[0025] The preferred process for dynamically allocating energy transmission paths is as follows:

[0026] After obtaining candidate pumping units and their current remaining power capacity, the system calls the geographic location database and network topology information to calculate the communication delay between energy storage units in each region and the line loss on the energy transmission path.

[0027] By combining the remaining capacity, charging dynamic response speed and receiving power curve parameters of energy storage units in each region, a comprehensive efficiency model is constructed to generate the corresponding energy path efficiency score.

[0028] Sort all available paths from highest to lowest score, prioritize the path with the highest score to establish the energy transmission link between the pumping unit and the regional energy storage unit, and record the selected path as the current scheduling node.

[0029] Preferably, the process for determining whether to perform an energy storage task is as follows:

[0030] Based on the current remaining kinetic energy release potential of the pumping unit and the real-time status of the regional energy storage unit, the expected efficiency of this energy recovery is calculated;

[0031] If the expected efficiency is greater than the minimum energy utilization threshold set by the system, and there are no conflicting scheduling tasks in the current system, then an energy storage execution command is issued.

[0032] If the expected efficiency is less than or equal to the minimum energy utilization threshold set by the system, and there are conflicting scheduling tasks in the current system, then no energy storage execution command will be issued.

[0033] Preferably, the process of forming a closed-loop optimization mechanism is as follows:

[0034] After completing the scheduling command, the energy storage control module sends back execution feedback information in real time, including actual recovered power, energy storage efficiency and response time.

[0035] The system compares and analyzes the feedback results with the expected results of the scheduled tasks, identifies the sources of deviation, and updates the scheduling parameter library.

[0036] If deviations or a downward trend in energy utilization occur repeatedly, the system will automatically optimize the scheduling rules and path scoring model to achieve adaptive iterative updates of the scheduling strategy.

[0037] Compared with existing technologies, the present invention provides an IoT-based distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform, which has the following advantages:

[0038] This invention deploys multimodal sensing terminals during the operation of an oil pumping unit to collect multi-dimensional operational data such as power, current, and vibration in real time. Combined with anomaly detection algorithms, it enables rapid identification and judgment of data anomalies. When anomaly characteristics meet preset trigger conditions, the state repair module is automatically activated, thereby improving the system's response speed and processing accuracy to fault symptoms and effectively ensuring the stability and continuity of the oil pumping unit's operating status. By introducing time-series interpolation and sample comparison mechanisms, short-cycle repair is performed on abnormal data segments, effectively restoring the continuity and integrity of the oil pumping unit's operating status. Simultaneously, a confidence scoring mechanism is used to assess the quality of the compensation effect, ensuring the reliability and reference value of the data after anomaly repair. The introduction of a confidence judgment mechanism improves the reliability assessment capability of the compensation results, avoiding the misuse of abnormal data to affect energy efficiency judgment. When compensation is unstable, the system can promptly switch to the energy efficiency identification process, enhancing the robustness and adaptability of overall data processing, thereby improving the response efficiency and energy utilization reliability of the entire scheduling system to abnormal operating conditions. By combining the remaining kinetic energy release potential of the pumping unit with the available capacity of the energy storage equipment, the efficiency scores of each energy path are dynamically evaluated. Multi-parameter scheduling criteria are used for optimal selection, achieving precise matching of the pumping unit and the optimal energy storage path while ensuring efficient and stable system operation. This improves the real-time performance, utilization rate, and overall scheduling intelligence of energy recovery, significantly enhancing the collaborative control capability of the distributed energy storage system. By generating control commands based on the scheduling results to execute energy recovery operations and collecting real-time energy storage feedback data, the scheduling execution effect is dynamically evaluated. If deviations or a downward trend in efficiency are detected, scheduling parameters are adjusted promptly, realizing a closed-loop adaptive optimization mechanism for the scheduling strategy. This effectively improves the accuracy and robustness of the system response, enhancing the intelligence and sustainable operation capability of energy management. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the structure of the present invention;

[0040] Figure 2 This is a schematic diagram of the method of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1: Please refer to Figure 1 As shown in the embodiment of the present invention, the distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform based on the Internet of Things includes:

[0043] Status Awareness Module: During the operation of the pumping unit, data is collected through deployed multimodal sensing terminals. When abnormal data is detected, it is determined whether status repair processing needs to be triggered. If so, the status repair module is entered.

[0044] The process by which the state awareness module determines whether state repair processing needs to be triggered is as follows:

[0045] After collecting the pumping unit's operating data, a multi-dimensional anomaly detection algorithm is used to assess the time-series consistency of key parameters such as power curves, current changes, and vibration signals. The anomaly score is calculated by combining the difference between the current data and historical stable operating data. Multi-modal sensing terminals, including but not limited to current sensors, voltage sensors, vibration sensors, and power acquisition modules, are deployed on key operating components of the pumping unit to collect multi-dimensional key parameter data in real time during operation. The collected data mainly includes power curves, three-phase currents, structural vibration signals, and other auxiliary operating parameters. All collected data is recorded according to timestamps and synchronously sent to the edge gateway or central dispatch and control center. A historical stable operating status database is constructed, selecting data segments of long-term continuous fault-free operation of the pumping unit under various typical operating conditions as a reference sample library. The system performs windowing processing on newly collected data and extracts the changing trends and statistical characteristic values ​​of power, current, and vibration signals as input features. The anomaly degree of each monitored parameter is quantified into a score, forming the following anomaly score function, with the formula:

[0046] ;

[0047] In the formula, These are the standardized difference indices for power, current, and vibration signals, respectively. , , These are the weighting coefficients for power, current, and vibration signals, respectively; the anomaly score is a continuous value used to comprehensively reflect the degree of deviation between the current window and the steady state; and a preset tolerance judgment threshold is provided. If multiple consecutive windows > If the corresponding data segment is marked as a potential anomalous segment, a minimum processing cycle threshold is set to eliminate false judgments caused by transient noise interference. When the duration of the anomalous segment... ≥ Furthermore, if the anomaly score continuously exceeds the tolerance threshold, it is determined that an observable anomaly has occurred in the current operating status of the pumping unit.

[0048] If the anomaly score exceeds the set tolerance threshold, and the duration of consecutive anomaly segments exceeds the minimum processing cycle threshold, then state repair processing is triggered.

[0049] By deploying multimodal sensing terminals during the operation of the pumping unit, multi-dimensional operating data such as power, current, and vibration are collected in real time. Combined with anomaly detection algorithms, the system can quickly identify and judge data anomalies. When the abnormal characteristics meet the preset trigger conditions, the status repair module is automatically activated, thereby improving the system's response speed and processing accuracy to fault symptoms and effectively ensuring the stability and continuity of the pumping unit's operating status.

[0050] State Repair Module: Uses time-series interpolation and sample comparison mechanism to perform short-cycle compensation for abnormal data, restore the integrity of the operating state, and judges whether the compensation result is lower than the threshold by confidence level. If so, it enters the energy efficiency identification module.

[0051] The process of short-cycle compensation for abnormal data using temporal interpolation and sample comparison mechanisms in the state repair module is as follows:

[0052] For detected abnormal data segments, historical normal periodic data within a sliding window is retrieved, and local interpolation compensation is performed based on sampling frequency, data periodicity, and the numerical trend of adjacent points. The system first determines the time range of the current abnormal segment and extracts multiple reference data windows from historical data of similar operating conditions of the pumping unit. These reference data windows are adjacent to the time position of the abnormal segment, have similar operating states, and have no abnormal records, serving as the basis for interpolation. Each reference window contains several sampling periods, and the window length is equal to that of the current abnormal segment. For each historical normal periodic data window, the first-order difference, fluctuation trend, extreme value position, and periodic peak and valley features of the data segment are extracted and compared with the boundary point features of the abnormal segment to ensure that the interpolation compensation does not destroy the overall shape of the original sequence.

[0053] Simultaneously, the system calls upon standard operating mode data stored in the feature template library, compares the residual distribution between the current data segment and similar patterns, and selects the sample segment with the smallest residual for weighted fusion repair. The system has a built-in standard feature template library for various operating conditions of the pumping unit, where each template contains typical variation patterns of multiple key parameters within a certain time window. Based on the operating condition of the current abnormal segment, the system filters out a candidate set of standard patterns that match it; for each candidate standard template, the system calculates the point-by-point residual of the key parameters within the corresponding time period between the interpolated abnormal segment and each candidate standard template, using the following formula:

[0054]

[0055] In the formula, This represents the value of the i-th sampling point;

[0056] The final generated compensation data segment should maintain a continuous trend of change with the adjacent non-abnormal segments.

[0057] The process in the state repair module to determine whether the compensation result is lower than the threshold based on confidence level is as follows:

[0058] The compensated data is compared with the original data segments before and after it. The sliding window dynamic time warping algorithm is used to calculate the data trend matching degree. The formula is:

[0059]

[0060] In the formula, For the target time series, The target time series; The system uses a fixed-length sliding window on the long-time series, sliding one step at a time to process the subsequences within the window. This represents the squared Euclidean distance between a pair of matching points. , Perform DTW; obtain a series of DTW scores. , ,..., This is used to assess whether the trend match changes over time;

[0061] A confidence score is generated by combining the degree of deviation between the compensated value and the original sampled value in the local statistical distribution;

[0062] If the score is below the confidence threshold, it indicates that the compensation is unstable or fluctuates drastically. In this case, the analysis will proceed to the energy efficiency identification module to prevent misjudgment of the timing of energy release.

[0063] By introducing time-series interpolation and sample comparison mechanisms, abnormal data segments are repaired in a short period of time, effectively restoring the continuity of the pumping unit's operating status and the integrity of the data. At the same time, a confidence score mechanism is used to judge the quality of the compensation effect, ensuring that the data after anomaly repair has credibility and reference value. If the compensation confidence is lower than the set threshold, the system will automatically switch to the energy efficiency identification module for further analysis, thereby improving the accuracy of overall data processing and the reliability of system decision-making.

[0064] Energy efficiency identification module: Based on the power change trend, load response frequency and non-critical operating condition characteristics in the pumping unit operation data, it identifies the time period where there is residual kinetic energy release, and judges whether the identification result meets the conditions for regenerative power generation. If it does, it enters the collaborative scheduling module.

[0065] The process by which the energy efficiency identification module determines whether the identification result meets the conditions for renewable power generation is as follows:

[0066] Extract the negative power segment per unit time, the transient instantaneous current drop rate, and the equipment inertial unloading response parameters from the power curve; collect real-time power curve data of the pumping unit; within a fixed time window, statistically analyze the segments of the power curve where the power is less than zero; extract the start and end times, amplitude, and duration of these segments as candidate energy recovery features; extract the current change curve of the pumping unit from the loaded state to the unloaded state; calculate the current drop rate using the first-order difference method, the formula is:

[0067] ;

[0068] In the formula, To unload the current value from the previous moment, The sampling time interval;

[0069] The three indicators are matched with the releasable kinetic energy indicators set in the threshold library. If the matching degree exceeds the set standard and there is a predictable short-term low load operation period, the current period is determined to meet the conditions for regenerative power generation.

[0070] The pumping unit is marked as a candidate device that can participate in scheduling and then enters the collaborative scheduling module.

[0071] It should be noted that the purpose of determining whether the identification result meets the conditions for renewable power generation is as follows:

[0072] Function 1: By accurately identifying the periods during the operation of the oil pumping unit that have the potential to release kinetic energy, it ensures that regenerative power generation is only performed under conditions with high kinetic energy feedback potential, avoiding ineffective starts and maximizing the ratio of kinetic energy to electrical energy.

[0073] Function 2: Predicting regenerative power generation conditions in advance helps coordinate the energy flow and load balance between the pumping unit group and the energy storage system, preventing system fluctuations or response delays due to scheduling conflicts or timing mismatches, and maintaining stable platform operation.

[0074] Function 3: Based on the identification results, dynamically select equipment that meets the participation conditions, effectively manage the participation scope of oil pumping units and energy storage nodes, and ensure that limited energy storage resources are prioritized for the equipment with the highest recovery value, thereby improving the overall system's resource utilization rate and scheduling intelligence level.

[0075] The technical solution of this embodiment is as follows: A time-series interpolation and sample comparison mechanism is used to compensate for short-period abnormal data detected during the operation of the pumping unit. Historical normal data segments are extracted using a sliding window for local interpolation repair, while standard operating samples are introduced for residual comparison. The optimal segment is selected for weighted fusion to ensure that the compensated data is consistent with the original trend. After compensation, a confidence score is generated using a dynamic time warping algorithm and local bias analysis. If the score is lower than a set threshold, it indicates insufficient data recovery stability, and the system will automatically transfer to the energy efficiency identification module for further analysis and processing. This achieves rapid identification and accurate compensation of short-period abnormal segments in the pumping unit's operating data, effectively restoring the integrity and trend continuity of the time-series data. By introducing a confidence judgment mechanism, the reliability assessment capability of the compensation results is improved, avoiding the misuse of abnormal data that could affect energy efficiency judgment. After compensation, the system not only performs trend consistency comparison on the results but also combines a sliding window dynamic time warping algorithm and local statistical bias analysis to comprehensively generate a confidence score, ensuring that the compensated data is highly reasonable in both value and trend. When the confidence level falls below a set threshold, the system can quickly identify and compensate for unstable states, automatically switching to the energy efficiency identification process for in-depth analysis and reassessment, preventing misjudgments or misscheduling caused by data anomalies. This mechanism significantly enhances the robustness and adaptability of the overall data processing chain. Especially when facing complex scenarios such as sudden operating conditions, sensor drift, or instantaneous data fluctuations, the system can still maintain high response accuracy and decision-making effectiveness, thereby further improving the entire scheduling system's ability to identify abnormal operating conditions, response efficiency, and energy utilization reliability, achieving refined and intelligent scheduling control.

[0076] Example 2: As Figure 1 As shown, the IoT-based distributed pumping unit cluster regenerative power generation and energy storage dispatching platform also includes the following modules:

[0077] Coordinated scheduling module: Based on the current status of the pumping unit and the available capacity of the energy storage equipment, it dynamically allocates energy transmission paths, adopts multi-parameter scheduling criteria, and determines whether to execute the energy storage task. If so, it enters the energy storage control module.

[0078] The process of dynamically allocating energy transmission paths in the collaborative scheduling module is as follows:

[0079] After obtaining candidate pumping units and their current remaining power capacity, the system calls the geographic location database and network topology information to calculate the communication delay between energy storage units in each region and the line loss on the energy transmission path.

[0080] By combining the remaining capacity, charging dynamic response speed and receiving power curve parameters of energy storage units in each region, a comprehensive efficiency model is constructed to generate the corresponding energy path efficiency score.

[0081] Sort all available paths from highest to lowest score, prioritize the path with the highest score to establish the energy transmission link between the pumping unit and the regional energy storage unit, and record the selected path as the current scheduling node.

[0082] The process of determining whether to execute an energy storage task in the collaborative scheduling module is as follows:

[0083] Based on the combined remaining kinetic energy release potential of the pumping unit and the real-time status of the regional energy storage units, the expected efficiency of this energy recovery is calculated. The system obtains the following information from each energy storage unit in real time via IoT terminals: current remaining charging capacity; current maximum charging power received; current charging response time and power response rate; location within the network topology and estimated transmission loss. The kinetic energy release potential is matched with the receiving capacity of the energy storage unit to calculate the expected recovery efficiency, using the following formula:

[0084]

[0085] In the formula, To release the potential value of kinetic energy, This represents the proportion of power loss in the communication and energy transmission path. The energy receiving efficiency of the regional energy storage unit under its current state;

[0086] If the expected efficiency is greater than the minimum energy utilization threshold set by the system, and there are no conflicting scheduling tasks in the current system, then an energy storage execution command is issued.

[0087] If the expected efficiency is less than or equal to the minimum energy utilization threshold set by the system, and there are conflicting scheduling tasks in the current system, then no energy storage execution command will be issued.

[0088] By combining the current remaining kinetic energy release potential of the pumping unit with the available capacity of the energy storage equipment, the efficiency score of each energy path is dynamically evaluated. Multi-parameter scheduling criteria are used for optimal selection, which achieves precise matching of the pumping unit and the optimal energy storage path under the premise of ensuring the efficient and stable operation of the system. This improves the real-time performance, utilization rate and overall scheduling intelligence of energy recovery, and significantly enhances the collaborative control capability of the distributed energy storage system.

[0089] Energy storage control module: Generates corresponding control commands based on scheduling results, executes energy recovery operations, collects energy storage feedback data, evaluates the execution effect, adjusts scheduling parameters, and forms a closed-loop optimization mechanism.

[0090] The process of forming a closed-loop optimization mechanism in the energy storage control module is as follows:

[0091] After the energy storage control module completes the scheduling command, the system first receives and records execution feedback information in real time, including actual recovered power, energy storage efficiency, and response time. These feedback values ​​are compared with the expected targets set in the scheduling task to calculate power difference, efficiency deviation, and response delay, and to determine if they exceed the set error tolerance range. If a deviation is detected, the system analyzes the source of the deviation by considering factors such as the pumping unit release status, energy storage unit load, and path energy loss, and updates the equipment performance coefficients, path loss weights, and response delay models in the scheduling parameter library accordingly. If the system detects a decrease in energy utilization or repeated deviations in multiple consecutive scheduling operations, it automatically triggers the scheduling rule optimization mechanism. Based on historical feedback, it retrains the path scoring model and adjusts equipment scheduling priorities, ultimately achieving adaptive iteration and dynamic correction of the scheduling strategy, continuously improving energy recovery efficiency and scheduling response stability.

[0092] By generating control commands based on scheduling results to execute energy recovery operations and collecting energy storage feedback data in real time, the scheduling execution effect is dynamically evaluated. If deviations or a downward trend in efficiency are detected, scheduling parameters are adjusted in a timely manner to realize a closed-loop adaptive optimization mechanism for scheduling strategies. This effectively improves the accuracy and robustness of system response and enhances the intelligence and sustainable operation capabilities of energy management.

[0093] Example 3: Please refer to Figure 2 As shown, the IoT-based distributed oil pumping unit cluster regenerative power generation and energy storage scheduling method includes the following steps:

[0094] S1: During the operation of the oil pumping unit, data is collected through the deployed multimodal sensing terminal. When abnormal data is detected, it is determined whether to trigger state repair processing. If so, the state repair module is entered.

[0095] S2: Using time-series interpolation and sample comparison mechanisms, short-cycle compensation is performed on abnormal data to restore the integrity of the operating state. The confidence level is used to determine whether the compensation result is lower than the threshold. If so, the energy efficiency identification module is entered.

[0096] S3: Based on the power change trend, load response frequency and non-critical operating condition characteristics in the pumping unit operation data, identify the time period where there is residual kinetic energy release, and determine whether the identification result meets the conditions for regenerative power generation. If it does, enter the collaborative scheduling module.

[0097] S4: Based on the current status of the pumping unit and the available capacity of the energy storage equipment, dynamically allocate energy transmission paths, adopt multi-parameter scheduling criteria, determine whether to execute the energy storage task, and if so, enter the energy storage control module.

[0098] S5: Generate corresponding control commands based on the scheduling results, execute energy recovery operations, collect energy storage feedback data, evaluate the execution effect, adjust scheduling parameters, and form a closed-loop optimization mechanism.

[0099] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0100] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform based on the Internet of Things, characterized in that, include: Status awareness module: During the operation of the pumping unit, data is collected through deployed multimodal sensing terminals. When abnormal data is detected, it is determined whether status repair processing needs to be triggered. If so, the status repair module is entered. State Repair Module: Uses time-series interpolation and sample comparison mechanism to perform short-cycle compensation for abnormal data, restore the integrity of the operating state, and judges whether the compensation result is lower than the threshold by confidence level. If so, it enters the energy efficiency identification module. Energy efficiency identification module: Based on the power change trend, load response frequency and non-critical operating condition characteristics in the pumping unit operation data, it identifies the time period where there is residual kinetic energy release, and judges whether the identification result meets the conditions for regenerative power generation. If it does, it enters the collaborative scheduling module. Coordinated scheduling module: Based on the current status of the pumping unit and the available capacity of the energy storage equipment, it dynamically allocates energy transmission paths, adopts multi-parameter scheduling criteria, and determines whether to execute the energy storage task. If so, it enters the energy storage control module. The process of determining whether to execute an energy storage task is as follows: Based on the combined remaining kinetic energy release potential of the pumping unit and the real-time status of the regional energy storage units, the expected efficiency of this energy recovery is calculated. The following information for each energy storage unit is obtained in real-time via IoT terminals: current remaining charging capacity, current maximum charging power received; current charging response time and power response rate; location within the network topology and estimated transmission loss. The kinetic energy release potential is matched with the receiving capacity of the energy storage unit to calculate the expected recovery efficiency, using the following formula: ； In the formula, To release the potential value of kinetic energy, This represents the proportion of power loss in the communication and energy transmission path. The energy receiving efficiency of the regional energy storage unit under its current state; If the expected efficiency is greater than the minimum energy utilization threshold set by the system, and there are no conflicting scheduling tasks in the current system, then an energy storage execution command is issued. If the expected efficiency is less than or equal to the minimum energy utilization threshold set by the system, and there are conflicting scheduling tasks in the current system, then no energy storage execution command will be issued. Energy storage control module: Generates corresponding control commands based on scheduling results, executes energy recovery operations, evaluates the execution effect, adjusts scheduling parameters, and forms a closed-loop optimization mechanism.

2. The IoT-based distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform according to claim 1, characterized in that, The process of determining whether state repair processing needs to be triggered is as follows: After collecting the pumping unit's operating data, a multi-dimensional anomaly detection algorithm is used to assess the temporal consistency of key parameters. The anomaly score is then calculated by combining the differences between the current data and historical stable operating data. The formula is as follows: ； In the formula, , , These are the standardized difference indices for power, current, and vibration signals, respectively. , , These are the weighting coefficients for power, current, and vibration signal, respectively. The anomaly score is a continuous value used to comprehensively reflect the degree of deviation between the current window and the steady state. A preset tolerance judgment threshold is also included. If multiple consecutive windows > If the corresponding data segment is marked as a potential anomalous segment, a minimum processing cycle threshold is set to eliminate false positives caused by transient noise interference. When the duration of the anomalous segment... ≥ Furthermore, if the anomaly score continuously exceeds the tolerance threshold, it is determined that an observable anomaly has occurred in the current operating status of the pumping unit. If the anomaly score exceeds the set tolerance threshold, and the duration of consecutive anomaly segments exceeds the minimum processing cycle threshold, then state repair processing is triggered.

3. The IoT-based distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform according to claim 2, characterized in that, The process of using time-series interpolation and sample comparison mechanisms to perform short-cycle compensation for outlier data is as follows: For data with detected abnormal segments, historical normal periodic data within the sliding window is retrieved, and local interpolation compensation is performed based on the sampling frequency, data periodicity, and the trend of adjacent point values. The standard operating mode data stored in the feature template library is called, the residual distribution between the current data segment and similar patterns is compared, and the sample segment with the smallest residual is selected for weight fusion repair. The final generated compensation data segment should maintain a continuous trend of change with the adjacent non-abnormal segments.

4. The IoT-based distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform according to claim 3, characterized in that, The process of determining whether the compensation result is lower than the threshold based on the confidence level is as follows: The compensated data is compared with the original data segments before and after it. The sliding window dynamic time warping algorithm is used to calculate the data trend matching degree. The formula is: ； In the formula, For the target time series, For the target time series; Given the squared Euclidean distance between a pair of matching points, a fixed-length sliding window is used on a long-time series, sliding one step at a time to process the subsequences within the window. , Perform DTW; obtain a series of DTW scores. , ,..., This is used to assess whether the trend match changes over time; A confidence score is generated by combining the degree of deviation between the compensated value and the original sampled value in the local statistical distribution; If the score is below the confidence threshold, it indicates that the compensation is unstable or fluctuates drastically.

5. The IoT-based distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform according to claim 4, characterized in that, The process of determining whether the identification result meets the conditions for renewable power generation is as follows: Extract the negative power segment per unit time, the transient instantaneous current drop rate, and the equipment inertial shedding response parameters from the power curve; The three indicators are matched with the releasable kinetic energy indicators set in the threshold library. If the matching degree exceeds the set standard and there is a predicted short-term low load operation period, the current period is determined to meet the conditions for regenerative power generation. The pumping unit is marked as a candidate device to participate in the scheduling and then enters the collaborative scheduling module.

6. The IoT-based distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform according to claim 5, characterized in that, The process of dynamically allocating energy transmission paths is as follows: After obtaining candidate pumping units and their current remaining power capacity, the system calls the geographic location database and network topology information to calculate the communication delay between energy storage units in each region and the line loss on the energy transmission path. By combining the remaining capacity, charging dynamic response speed and receiving power curve parameters of energy storage units in each region, a comprehensive efficiency model is constructed to generate the corresponding energy path efficiency score. Sort all available paths from highest to lowest score, prioritize the path with the highest score to establish the energy transmission link between the pumping unit and the regional energy storage unit, and record the selected path as the current scheduling node.

7. The IoT-based distributed oil pumping unit cluster regenerative power generation and energy storage dispatching platform according to claim 1, characterized in that, The process of forming a closed-loop optimization mechanism is as follows: After completing the scheduling command, the energy storage control module sends back execution feedback information in real time, including actual recovered power, energy storage efficiency and response time. The system compares and analyzes the feedback results with the expected results of the scheduled tasks, identifies the sources of deviation, and updates the scheduling parameter library. If deviations or a downward trend in energy utilization occur repeatedly, the system will automatically optimize the scheduling rules and path scoring model to achieve adaptive iterative updates of the scheduling strategy.

Images