Ground control system and control method of electric submersible pump

Abstract

The invention belongs to the technical field of intelligent control of electric submersible pumps, and discloses a ground control system and control method of an electric submersible pump. A sensor module of the system is used for collecting operation data of the electric submersible pump in real time; the intelligent calculation module is used for analyzing and calculating the collected real-time operation data of the electric submersible pump to obtain optimized control parameters; the control execution module is used for generating a corresponding instruction according to the optimized parameters and accurately adjusting the steering, rotating speed and stroke parameters of the servo motor; the real-time monitoring and feedback optimization module is used for monitoring the system state after execution of the control execution module in real time and judging whether system operation is abnormal or not according to a real-time monitoring result; and the alarm optimization adjustment module is used for dynamically optimizing the control strategy of the electric submersible pump based on the parameter abnormal condition fed back by the real-time monitoring module after the abnormal state of the system is detected. The control efficiency of the electric submersible pump is effectively improved, and the problem that in the prior art, a control method is low in efficiency is solved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent control of submersible electric pumps, and particularly relates to a ground control system and a control method for a submersible electric pump. BACKGROUND

[0002] Submersible electric pumps are increasingly widely used in the field of oil exploitation due to their own advantages, and are mainly used to improve oil production efficiency and ensure stable production of oil wells. However, in the actual operation process of the submersible electric pump, due to the poor downhole environment and complex operation conditions, the operating parameters (such as voltage, current, temperature, stroke, etc.) of the submersible electric pump often fluctuate greatly. If these parameters cannot be monitored, analyzed and controlled in a timely and effective manner, problems such as reduced equipment efficiency and reduced service life may occur.

[0003] Most existing control systems for submersible electric pumps are based on PLC and PID control logic with fixed parameters, and the monitoring mainly focuses on basic data such as current and voltage. The types of sensors are single and the sampling frequency is low, which cannot effectively reflect the transient operating condition characteristics. There is a lack of intelligent identification and prediction capabilities, and the following problems commonly exist in the control cabinet:

[0004] The quality of key components such as soft starters and contactors is unstable; the fault positioning capability is weak, usually requiring manual intervention, and the maintenance cycle is long; the control strategy cannot be adjusted in real time according to changes in load and displacement requirements; it is impossible to achieve operating condition classification alarm and self-adaptive optimization, and the system is prone to a passive cycle of protection-shutdown-manual intervention. In addition, some technologies introduce fuzzy control for speed adjustment of reciprocating pumps, but the scheme mainly focuses on displacement feedback control and cannot achieve overall adaptive management of the submersible electric pump system.

[0005] In order to improve the operating efficiency of the submersible electric pump, prolong the service life of the equipment, reduce the maintenance cost of the equipment, and improve the overall reliability and stability of the system, it is urgent to develop a ground control system and a control method that have real-time data acquisition, intelligent calculation and analysis, dynamic adjustment control, and can timely handle abnormal situations.

[0006] Through the above analysis, the problems and defects of the prior art are:

[0007] (1) Limited monitoring means: unable to achieve comprehensive real-time monitoring of operating parameters, low sampling accuracy, and insufficient collection dimension of operating parameters.

[0008] (2) Lack of intelligent control capability: the control strategy response is slow, and cannot be optimized in real time based on historical data trends and real-time data feedback, resulting in the submersible electric pump operating in a non-optimal condition.

[0009] (3) Lack of hierarchical response strategy based on fault levels, unable to achieve fault prediction, priority control and closed-loop optimization.

[0010] (4) Lack of dynamic frequency modulation capability, it is difficult to make reasonable response to the problems such as resonance and stroke reduction caused by rapid change of downhole working conditions. SUMMARY

[0011] In order to overcome the problems in the related art, the present application discloses a surface control system and control method for submersible electric pumps, which can realize real-time monitoring, intelligent analysis and calculation, dynamic control and adjustment of the operating state of the submersible electric pump, and can cope with abnormal situations to a certain extent, improve the operating efficiency and reliability of the equipment, and the technical solution is as follows:

[0012] The present application is implemented as follows: a surface control system for a submersible electric pump, comprising:

[0013] A sensor module for real-time acquisition of operating data of the submersible electric pump;

[0014] An intelligent calculation module including an LSTM network for fault prediction and a reinforcement learning algorithm for control optimization, which analyzes and calculates the acquired real-time data of the submersible electric pump based on historical operating data and simulation data, obtains optimized control parameters, and trains and periodically adjusts the model based on the operating data of the submersible electric pump;

[0015] A control execution module for generating instructions based on the optimized parameters to adjust the steering, speed and stroke parameters of the servo motor; wherein the speed error is controlled within ±0.5%, and the stroke adjustment accuracy is not less than ±2mm; the control execution module adopts redundant control logic to automatically switch to PID algorithm when the artificial intelligence control is abnormal;

[0016] A real-time monitoring and feedback optimization module connected with the control execution module for real-time monitoring of the system state after execution of the control execution module, and determining whether the system operation is abnormal based on the real-time monitoring results;

[0017] An alarm optimization adjustment module for dynamically optimizing the control strategy of the submersible electric pump based on the parameter abnormality feedback of the real-time monitoring module after detecting the abnormal state of the system; wherein the abnormal alarm includes two levels, the first level alarm corresponds to serious abnormality, such as motor overload and stroke overrun, and the system stops immediately after triggering; the second level alarm corresponds to controllable abnormality, such as unstable liquid discharge and current fluctuation, and the system enters a frequency limiting or load reducing operating state after triggering, and if the second level alarm is not removed within the preset time limit, the system is automatically upgraded to the first level alarm.

[0018] Further, the sensor module includes a voltage sensor, a current sensor, a temperature sensor, and a stroke sensor for comprehensive real-time monitoring of the operating parameters of the electric submersible pump. The sampling frequency of the current sensor and the stroke sensor is not less than 1 kHz, the sampling frequency of the voltage sensor is not less than 500 Hz, and the sampling frequency of the temperature sensor is not less than 0.1 Hz; the sensors are packaged with corrosion-resistant materials, signal transmission uses shielded cable and differential transmission method, and temperature compensation is combined to improve measurement stability.

[0019] Further, the intelligent computing module includes a data preprocessing unit, a feature extraction unit, and an intelligent analysis and calculation unit.

[0020] The data preprocessing unit is used for standardizing the data collected by the sensors; the data collected by the sensors is completed by linear interpolation, and abnormal values are corrected by using the 3σ principle:

[0021]

[0022] In the formula, μ is the mean, σ is the variance, and x is the data collected by the sensor.

[0023] The current and voltage data are standardized by Z-score:

[0024]

[0025] In the formula, z is the standardized value, indicating the deviation of the data from the mean;

[0026] The temperature is standardized by Min-Max to map to the interval [0, 1]:

[0027]

[0028] In the formula, x' is the standardized data, max(x) is the maximum value in the data, and min(x) is the minimum value in the data;

[0029] The stroke is sine-encoded.

[0030] The feature extraction unit is used for extracting key features from the preprocessed data; statistical features (mean, variance, etc.) are extracted based on a sliding window, and time-frequency features are extracted based on wavelet changes.

[0031] The intelligent analysis calculation unit comprises an LSTM network for fault prediction and a reinforcement learning algorithm for control optimization, analyzes and calculates real-time operation data collected by the sensor based on historical operation data and simulation data, and outputs optimized servo motor control parameters; wherein the LSTM network is used for modeling the operation trend of the electric submersible pump, predicting potential fault risks, and the reinforcement learning algorithm is used for generating optimal steering timing, operation speed and stroke control strategies according to the current working condition.

[0032] Further, the control execution module comprises a steering control unit, a speed control unit and a stroke control unit, for receiving the analysis results of the intelligent calculation module in real time, and automatically adjusting the steering, speed and stroke of the electric submersible pump motor.

[0033] Further, the real-time monitoring and feedback optimization module comprises a state monitoring unit and a feedback optimization unit;

[0034] The state monitoring unit is used for continuously monitoring the running state after the execution of the control execution module, and judging whether the system running state deviates from the normal range through the mean value, variance and other indicators in the window based on the dynamic threshold model constructed based on the operation history data;

[0035] The feedback optimization unit is used for feeding back the monitoring results to the intelligent calculation module in real time to form a closed-loop optimization control.

[0036] Further, the alarm optimization adjustment module comprises an abnormal diagnosis unit and an alarm response unit;

[0037] When the system monitors an abnormal state, the abnormal diagnosis unit analyzes the fault cause based on parameters such as current harmonic characteristics, axial vibration spectrum and well temperature gradient, generates an optimization adjustment strategy; when the efficiency optimization strategy and the protective load reduction strategy conflict, the efficiency optimization strategy is executed preferentially; the alarm response unit timely issues an alarm signal and executes automatic optimization adjustment. The optimization adjustment strategy has different strategy weights under different working conditions, and in high water cut wells, the gas lock risk is preferentially suppressed, and in low yield wells, the sand sticking fault is preferentially suppressed.

[0038] Another object of the present application is to provide a ground control method for an electric submersible pump, which is used for regulating and controlling the ground control system of the electric submersible pump, and the method comprises the following steps:

[0039] S1, collecting the voltage, current, temperature and stroke operation data of the electric submersible pump in real time through a sensor;

[0040] S2, preprocessing the collected electric submersible pump operation data, and analyzing and calculating through an artificial intelligence algorithm to predict the best servo motor steering time, speed and stroke parameters;

[0041] S3, dynamically control the steering, rotation speed and stroke of the motor based on the calculated parameters;

[0042] S4, continuously monitor the system operating state in real time and perform feedback and optimization; if an abnormal condition is detected, an alarm signal is issued and the control strategy is optimized and adjusted until the system returns to normal operating state.

[0043] In step S2, the best servo motor steering time, rotation speed and stroke parameters are predicted through artificial intelligence algorithm analysis and calculation, including:

[0044] Based on the real-time data collected by the sensor, the motor current fluctuation is used to determine whether the plunger has reached the maximum stroke, and the current situation is compared with the previous cycle reversal time to determine the steering time.

[0045] According to the well fluid flow demand and stroke length and frequency requirements, combined with real-time load condition, the minimum energy consumption is calculated to meet the demand of rotation speed, fuzzy logic control algorithm is introduced to optimize the rotation speed, set flow deviation ΔQ = Q target -Q t , Q target is the target flow, Q t is the current flow value, the domain is [-20%, +20%], divided into 7 fuzzy sets: NB, NM, NS, ZO, PS, PM, PB, each fuzzy set corresponds to a group of rotation speed adjustment strategy, the control system can adaptively adjust the weight, when the water content of well fluid exceeds the set threshold, the energy efficiency weight is automatically adjusted, and the control target is corrected;

[0046] Real-time analysis of stroke displacement sensor data is performed to identify effective upstroke and downstroke, and the effective stroke determination standard is Δx ≥ 85% x theory , x theory is the theoretical displacement; the acceleration meets the pumping characteristics; when the system detects insufficient stroke, the control logic processes according to the deviation level: when the deviation is slight, the steering time is adjusted to correct it; when the deviation is moderate and severe: the stroke target value is corrected combined with the displacement trend of the previous few cycles; when the flow demand changes suddenly, the motor rotation speed is adjusted first, and when the motor rotation speed cannot meet the demand, the stroke and steering logic are adjusted; finally, the motor steering time, rotation speed and stroke are calculated.

[0047] In step S3, the steering, rotation speed and stroke of the motor are dynamically controlled based on the calculated parameters, including:

[0048] Receive the steering trigger signal calculated by the artificial intelligence algorithm, and send a positive or negative rotation instruction to the motor controller to make the motor complete the steering switching; the switching signal is received and analyzed by the PLC control unit, and the instruction is transmitted to the motor through the driver interface to realize the up-down reciprocating motion switching of the plunger;

[0049] The frequency modulation capability of the servo driver is utilized, a dynamic frequency modulation control architecture is constructed based on the displacement-load double feedback, and the rotation speed is adjusted; the control system comprises an inner loop and an outer loop, the inner loop is a load index fused by a current, a torque and a vibration signal, and the outer loop is a real-time displacement; the frequency modulation control adopts a fuzzy self-adaptive PID algorithm, dynamically adjusts an output frequency according to load changes and displacement requirements, the system has a resonance dynamic detection function, potential resonance working conditions are identified by analyzing frequency response and load fluctuation, and a hierarchical response strategy is adopted, when the condition is slight, the frequency is adjusted, and when the condition is serious, a protection mode is entered to reduce the rotation speed;

[0050] The displacement sensor feeds back the real-time position of the plunger to the PLC, and the control execution module compares the position with the historical stroke range, if it is determined that the stroke is incomplete or deviates from the preset upper and lower limits of the stroke, the reverse rotation logic is adjusted and the position trigger threshold is reset, so that the actual stroke is consistent with the target stroke, and finally the closed-loop control of the motor forward and reverse rotation, the running speed and the stroke length is realized.

[0051] In step S4, the system operation state is continuously monitored in real time and feedback and optimization are performed, including:

[0052] After detecting the abnormal state of the system, the control strategy of the submersible electric pump is dynamically optimized based on the parameter abnormality feedback of the real-time monitoring module; the fault level of the submersible electric pump is divided into three levels:

[0053] First-level fault: motor overload, high current harmonic, stroke out-of-limit causing equipment damage;

[0054] Second-level fault: unstable displacement, abnormal vibration, stroke deviation with recoverability;

[0055] Third-level fault: vibration change or deviation fault;

[0056] The alarm module judges the fault level according to the abnormal type classification, and transmits the result to the control logic module, re-evaluates the steering, rotation speed and stroke according to the historical trend and current fluctuation range of the abnormal parameter, and adjusts the established control cycle;

[0057] The optimization module feeds back the corrected control parameters to the control execution module, executes new scheduling instructions, and realizes closed-loop correction and recovery control of the system operation strategy.

[0058] In combination with all the technical solutions described above, the present application has the following beneficial effects:

[0059] First, in view of the technical problems existing in the prior art and the difficulty of solving the problems, the technical problems solved by the technical solution of the present application and the results and data obtained during the research and development process are analyzed in detail and deeply, and the technical effects brought about after solving the problems are described as follows:

[0060] 1. The system can realize real-time and high-precision monitoring of the operating parameters of the submersible electric pump, control the steering time and rotating speed of the servo motor, and improve the overall efficiency and reliability of the system based on sensor feedback, PLC and servo drive.

[0061] 2. The accuracy and response speed of parameter adjustment are improved by introducing fuzzy control and intelligent algorithm for real-time calculation and judgment of collected data.

[0062] 3. The system has self-adaptive adjustment capability, can correct the control strategy in real time according to the load and stroke state, and always keep the submersible electric pump running in high efficiency condition, thereby reducing the failure rate.

[0063] 4. The system has fault diagnosis and alarm optimization mechanism, can quickly respond to abnormal conditions and correct the strategy, reduce manual intervention and prolong the service life of the equipment.

[0064] 5. The automation degree and stability of the whole system are improved, and the operation reliability of the submersible electric pump is significantly improved.

[0065] Second, the ground control system of the submersible electric pump of the present application can monitor the operating state of the submersible electric pump, control the steering time, rotating speed and stroke of the servo motor and judge whether there is abnormal condition, and give an alarm and make optimization adjustment when abnormality is found. The present application effectively improves the control efficiency of the submersible electric pump, overcomes the problem of low efficiency of the control method in the prior art, and can realize intelligent and efficient management of the submersible electric pump.

[0066] Third, as a positive effect of the present application, it is also reflected in the following important aspects:

[0067] (1) Compared with the traditional control system based on timing maintenance, the intelligent predictive maintenance in the present application can reduce the number of unplanned shutdowns by more than 40%, and reduce the operation and maintenance cost by about 20% to 25%; the intelligent control system adjusts the operation strategy based on historical operation data and real-time trend, which can improve the stability of oil well operation, prolong the average trouble-free working time of the submersible electric pump, reduce the replacement frequency and increase the service life.

[0068] (2) The existing submersible electric pump control system at home and abroad mainly takes PID as the core, lacks high-frequency sampling, intelligent control and optimization mechanism, the real-time sampling frequency of the present application is greatly improved compared with the prior art; the LSTM network is introduced to model the operation trend, which can predict the state of the submersible electric pump in advance and identify potential faults; the reinforcement learning algorithm is used to optimize the steering, speed and stroke of the motor, which has self-adaptive ability; it supports the linkage analysis of parameters such as vibration and temperature gradient, has fault identification and fault warning capability, and fills the technical gap of single monitoring dimension.

[0069] (3) The present application ensures the integrity of the plunger movement through the sine-cosine coded displacement analysis, operation trend judgment and deviation grading control, solves the problems of insufficient stroke and deviation of the submersible electric pump;

[0070] Using the intelligent control algorithm prediction model to analyze the motor load, current fluctuation and other parameters in advance can avoid accident shutdown and solve the problem of difficult predictive maintenance.

[0071] (4) The present application breaks the technical prejudice in the industry that "artificial intelligence algorithm is not suitable for real-time industrial control". Through the combination of reinforcement learning and LSTM algorithm, the running state of the submersible electric pump is quickly perceived and accurately controlled, the control response delay is less than 50ms, which meets the real-time requirements of industrial grade. At the same time, the system designs redundant control logic to ensure automatic switching to PID control when the intelligent module is abnormal, ensuring stability and reliability, thus effectively overcoming the inherent doubts of the industry about the stability and practicality of intelligent control. BRIEF DESCRIPTION OF DRAWINGS

[0072] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure;

[0073] Figure 1 is the ground control method flow chart of the submersible electric pump provided by the embodiment of the present application. DETAILED DESCRIPTION

[0074] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below. In the following description, a lot of specific details are set forth in order to fully understand the present application. However, the present application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the scope of the present application, therefore the present application is not limited to the specific implementation disclosed below.

[0075] The innovation of the present application is that the LSTM neural network and the reinforcement learning algorithm are introduced to realize real-time analysis and fault prediction of the operating data of the electric submersible pump, and the dynamic control of the motor steering, speed and stroke is realized by combining various sensors; the system has the functions of closed-loop feedback, adaptive optimization, hierarchical alarm response and abnormal condition self-repairing, and adopts redundant control logic to ensure the stability of control, which significantly improves the operating efficiency, reliability and intelligent level of the electric submersible pump.

[0076] Embodiment 1, Figure 1 The flow chart of the ground control method of the electric submersible pump is as follows: when the system is powered on and started, the control system will activate each functional module in turn, including the data acquisition module, the intelligent calculation module, the control execution module and the real-time monitoring feedback module. The system starts running, and various sensors collect operating data in real time, including but not limited to current, voltage sensors, temperature sensors, pressure sensors and stroke displacement sensors, to comprehensively collect important parameter information of the electric submersible pump, such as current, voltage, temperature, pressure and stroke displacement. The real-time sensor data collected is transmitted to the intelligent calculation module of the ground control center through the data transmission module, and the real-time data is comprehensively analyzed and calculated to calculate the servo motor control parameters of the electric submersible pump during operation, including but not limited to the steering time, the optimal operating speed and the stroke length of the servo motor. The calculated control parameters are then sent to the control execution module, and according to the calculation results, the control execution module sends control instructions to the servo motor in real time to dynamically adjust the steering, speed and stroke action of the motor, so as to realize efficient and accurate real-time control operation of the electric submersible pump. When the control execution module is started, the system does not stop monitoring the operating state of the electric submersible pump, but monitors and feeds back the operating state in real time through the sensor, and dynamically optimizes and predicts the control strategy again through the feedback data. When the monitoring system finds that the operating data of the electric submersible pump is abnormal, the system will start the alarm program and the optimization and adjustment program to correct the abnormal operating parameters to try to eliminate the fault or abnormal state. When the system successfully eliminates the abnormality through automatic optimization and adjustment, the electric submersible pump control system will continue to return to the real-time monitoring and feedback optimization stage to continue running, so as to keep the electric submersible pump running stably and efficiently for a long time. If the automatic optimization and adjustment cannot solve the abnormality, the system prompts the on-site staff to further intervene in the inspection and maintenance. The above process continues to run in a loop.

[0077] The present application provides a ground control method of an electric submersible pump, which is used for regulating and controlling the ground control system of the electric submersible pump, and comprises the following steps:

[0078] S1, collecting voltage, current, temperature and stroke operating data of the electric submersible pump in real time through sensors;

[0079] S2, the collected submersible electric pump operation data is preprocessed, and the optimal servo motor steering time, speed and stroke parameters are predicted through artificial intelligence algorithm analysis and calculation;

[0080] S3, the steering, speed and stroke of the motor are dynamically controlled based on the calculated parameters;

[0081] S4, continuously monitor the system operation state and feedback and optimization; if abnormal condition is detected, an alarm signal is sent and the control strategy is optimized and adjusted until the system returns to normal operation state.

[0082] In step S2, the optimal servo motor steering time, speed and stroke parameters are predicted through artificial intelligence algorithm analysis and calculation, including:

[0083] Based on the real-time data collected by the sensor, the motor current fluctuation is used to determine whether the plunger reaches the maximum stroke, and the steering time is determined by comparing the previous cycle reverse time with the current situation.

[0084] According to the well fluid flow demand and the stroke length and frequency requirement, the minimum energy consumption is calculated to meet the demand of the speed, the fuzzy logic control algorithm is introduced to optimize the speed, and the flow deviation AQ = Q target -Q t , Q target is the target flow value, Q t is the current flow value, the domain is [-20%, +20%], divided into 7 fuzzy sets: NB, NM, NS, ZO, PS, PM, PB, each fuzzy set corresponds to a group of speed adjustment strategy, the control system can adaptively adjust the weight, when the water content of well fluid exceeds the set threshold, the energy efficiency weight is automatically adjusted, and the control target is corrected;

[0085] The stroke displacement sensor data is analyzed in real time, and the effective up and down stroke is identified, and the effective stroke determination standard is Δx≥85% x theory , x theory is the theoretical displacement; the acceleration meets the pumping characteristics; when the system detects insufficient stroke, the control logic processes according to the deviation degree: when the deviation is slight, the steering time is adjusted to correct; when the deviation is moderate and severe: the stroke target value is corrected according to the displacement trend of the previous several cycles; when the flow demand changes suddenly, the motor speed is adjusted first, and when the motor speed cannot meet the demand, the stroke and steering logic are adjusted; finally, the motor steering time, speed and stroke are calculated.

[0086] In step S3, the steering, speed and stroke of the motor are dynamically controlled based on the calculated parameters, including:

[0087] The turning trigger signal calculated by the artificial intelligence algorithm is received, and a forward or reverse rotation instruction is sent to the motor controller to make the motor complete the turning switching; the switching signal is received and analyzed by the PLC control unit, and the instruction is transmitted to the motor through the driver interface to realize the up-down reciprocating motion switching of the plunger;

[0088] The frequency modulation capability of the servo driver is utilized, and a dynamic frequency modulation control architecture is constructed based on the displacement-load double feedback to realize the regulation of the rotating speed; the control system includes an inner loop and an outer loop, the inner loop is a load index fused by the current, torque and vibration signal, and the outer loop is a real-time displacement; the frequency modulation control adopts a fuzzy self-adaptive PID algorithm, dynamically adjusts the output frequency according to the load change and the displacement demand, the system has a resonance dynamic detection function, identifies the potential resonance working condition by analyzing the frequency response and load fluctuation, and adopts a hierarchical response strategy, adjusts the frequency when it is slight, and enters a protection mode to reduce the rotating speed when it is serious;

[0089] The displacement sensor feeds back the real-time position of the plunger to the PLC, and the control execution module compares it with the historical stroke range, if it is determined that the stroke is incomplete or deviates from the preset upper and lower limits of the stroke, the reverse rotation logic is adjusted and the position trigger threshold is reset, so that the actual stroke is consistent with the target stroke, and finally the closed-loop control of the motor forward and reverse rotation, operating speed and stroke length is realized.

[0090] In step S4, the system operation state is continuously monitored in real time and feedback and optimization are performed, including:

[0091] After detecting the abnormal state of the system, the control strategy of the submersible electric pump is dynamically optimized based on the parameter abnormality feedback of the real-time monitoring module; the fault level of the submersible electric pump is divided into three levels:

[0092] First-level fault: motor overload, high current harmonic, stroke out-of-limit causing equipment damage;

[0093] Second-level fault: unstable displacement, abnormal vibration, stroke deviation with recoverability;

[0094] Third-level fault: vibration change or deviation fault;

[0095] The alarm module judges the fault level according to the abnormal type classification, and transmits the result to the control logic module, re-evaluates the turning, rotating speed and stroke according to the historical trend and current fluctuation range of the abnormal parameter, and adjusts the established control cycle;

[0096] The optimization module feeds back the corrected control parameters to the control execution module to execute new scheduling instructions, realizes the closed-loop correction and recovery control of the system operation strategy.

[0097] Embodiment 2, a ground control system of a submersible electric pump, the system comprises:

[0098] A sensor module is configured to collect real-time operation data of the electric submersible pump; the sensor module comprises a voltage sensor, a current sensor, a temperature sensor, and a stroke sensor; the sampling frequency of the current sensor and the stroke sensor is not less than 1 kHz, the sampling frequency of the voltage sensor is not less than 500 Hz, and the sampling frequency of the temperature sensor is not less than 0.1 Hz; each sensor is packaged with corrosion-resistant materials, and the signal transmission adopts a shielded cable and a differential transmission mode.

[0099] An intelligent computing module comprises an LSTM network for fault prediction and a reinforcement learning algorithm for control optimization, and is configured to analyze and calculate the collected real-time data of the electric submersible pump based on historical operation data and simulation data, to obtain optimized control parameters, and to train and periodically adjust the model based on the operation data of the electric submersible pump;

[0100] A control execution module is configured to generate instructions according to the optimized parameters, to adjust the steering, speed, and stroke parameters of the servo motor, and to adopt a redundant control logic, so that the PID algorithm is automatically switched when the artificial intelligence control is abnormal;

[0101] A real-time monitoring and feedback optimization module is connected to the control execution module, and is configured to monitor the system state after the execution of the control execution module in real time, and to determine whether the system operation is abnormal based on the real-time monitoring result;

[0102] An alarm optimization adjustment module is configured to dynamically optimize the control strategy of the electric submersible pump based on the parameter abnormality feedback of the real-time monitoring module after detecting an abnormal state of the system; wherein the abnormal alarm comprises a first-level alarm and a second-level alarm.

[0103] In one embodiment, the intelligent computing module comprises a data preprocessing unit, a feature extraction unit, and an intelligent analysis and calculation unit;

[0104] The data preprocessing unit is configured to standardize the data collected by the sensors; the data collected by the sensors is complemented by linear interpolation, and the abnormal values are corrected by the 3σ principle, and the expression is:

[0105]

[0106] In the formula, μ is the mean, σ is the variance, and x is the original data;

[0107] The current and voltage data are standardized by Z-score, and the expression is:

[0108]

[0109] In the formula, z is the standardized data;

[0110] The temperature is normalized by Min-Max standardization and mapped to the interval [0, 1], and the stroke is sine-coded, with the expression being:

[0111]

[0112] In the formula, x' is the normalized data, max(x) is the maximum value in the data, and min(x) is the minimum value in the data;

[0113] The feature extraction unit is used to extract key features from the preprocessed data; statistical features are extracted based on a sliding window, and time-frequency features are extracted based on wavelet changes;

[0114] The intelligent analysis and calculation unit includes an LSTM network for fault prediction and a reinforcement learning algorithm for control optimization, which analyzes and calculates real-time running data collected by the sensor based on historical running data and simulation data, and outputs optimized servo motor control parameters; wherein the LSTM network is used to model the running trend of the electric submersible pump, predict potential fault risks, and the reinforcement learning algorithm is used to generate the optimal steering timing, running speed and stroke control strategy according to the current working condition.

[0115] In one embodiment, the control execution module includes a steering control unit, a speed control unit and a stroke control unit for receiving the analysis results of the intelligent calculation module in real time and automatically performing adjustment operations on the steering, speed and stroke of the electric submersible pump motor. Among them, the steering control unit sends a forward or reverse instruction to the servo motor through the PLC controller according to the steering trigger signal output by the intelligent analysis module, to realize the up-down stroke switching of the plunger; the speed control unit adjusts the output frequency of the servo driver based on the target flow and real-time load through fuzzy logic and adaptive PID algorithm, to realize dynamic speed control under displacement-load double feedback; the stroke control unit compares the plunger displacement sensor data in real time to determine whether the stroke reaches the effective displacement standard, and when there is a deviation, adjusts the reverse timing or corrects the stroke target value according to the severity, to realize closed-loop adjustment of the stroke length; In one embodiment, the real-time monitoring and feedback optimization module includes a state monitoring unit and a feedback optimization unit;

[0116] The state monitoring unit is used to continuously monitor the running state after the execution of the control execution module, and based on the dynamic threshold model constructed based on the running history data, the mean and variance indicators in the window are used to determine whether the system running state deviates from the normal range;

[0117] The feedback optimization unit is used to feed back the monitoring results to the intelligent calculation module in real time to form a closed-loop optimization control.

[0118] In one embodiment, the alarm optimization adjustment module includes an abnormality diagnosis unit and an alarm response unit;

[0119] When the system monitors the abnormal state, the abnormal diagnosis unit analyzes the fault cause based on the parameters of current harmonic characteristics, axial vibration spectrum and well temperature gradient, generates an optimized adjustment strategy; when the efficiency optimization strategy and the protective load reduction strategy conflict, the efficiency optimization strategy is executed first; the alarm response unit issues an alarm signal in time and executes automatic optimization adjustment.

[0120] The optimized adjustment strategy has different strategy weights under different working conditions, and in high water cut wells, the gas lock risk is preferentially inhibited, and in low production wells, the sand sticking failure is preferentially inhibited.

[0121] To further prove the positive effect of the above embodiment, the present application based on the above technical solution carries out the following experiment.

[0122] The present application adopts LSTM network and reinforcement learning algorithm, both of which have been proved to have good prediction and control optimization ability in a large number of practical industrial scenes and literature. LSTM network is good at processing time series data, can capture long-term dependence in data, and accurately predict the running trend of electric submersible pump; reinforcement learning algorithm can adjust the control strategy according to real-time feedback data, realize dynamic optimization of steering, speed and stroke; the control system realizes qualitative improvement in data acquisition frequency, the sampling frequency of current and stroke is as high as 1kHz, and the sampling frequency of voltage is 500Hz, which significantly exceeds the sampling frequency of less than 200Hz of traditional control system. This high-frequency acquisition method can theoretically capture the small abnormal changes in system operation more accurately, and provides a strong data basis for subsequent rapid response. Through the dynamic frequency control architecture, based on the load-liquid discharge double feedback mechanism, the electric pump operating point can be adjusted in real time to better meet the actual demand, and theoretically the overall energy efficiency can be improved by 12%-15%.

[0123] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can make any modification, equivalent replacement and improvement within the technical range disclosed in the present application, which shall be covered within the protection scope of the present application.

Claims

1. A ground control system for a submersible electric pump, characterized in that: The system includes: Sensor module, used to collect operating data of submersible oil pumps in real time; The intelligent computing module, which includes an LSTM network for fault prediction and a reinforcement learning algorithm for control optimization, analyzes and calculates collected real-time data from submersible pumps based on historical operating data and simulation data to obtain optimized control parameters. It then trains and periodically adjusts the model based on the submersible pump operating data. The control execution module is used to generate instructions based on the optimized parameters to adjust the servo motor's direction, speed, and stroke parameters. The control execution module uses redundant control logic and automatically switches to the PID algorithm when artificial intelligence control fails. The real-time monitoring and feedback optimization module is connected to the control execution module and is used to monitor the system status after the control execution module is executed in real time, and determine whether there is any abnormality in the system operation based on the real-time monitoring results; The alarm optimization and adjustment module is used to dynamically optimize the control strategy of the submersible electric pump based on the parameter abnormalities fed back by the real-time monitoring module after detecting an abnormal state of the system; among them, the abnormal alarm includes level one alarm and level two alarm.

2. The ground control system of the submersible electric pump according to claim 1, characterized in that: The sensor module includes: voltage sensor, current sensor, temperature sensor and stroke sensor; The sampling frequency of the current sensor and stroke sensor shall not be less than 1kHz, the sampling frequency of the voltage sensor shall not be less than 500Hz, and the sampling frequency of the temperature sensor shall not be less than 0.1Hz; each sensor shall be encapsulated with corrosion-resistant materials, and signal transmission shall adopt shielded cable and differential transmission method.

3. The ground control system of the submersible electric pump according to claim 1, characterized in that: The intelligent computing module includes: data preprocessing unit, feature extraction unit and intelligent analysis and computing unit; The data preprocessing unit is used to standardize the data collected by the sensor. The data collected by the sensor is supplemented by linear interpolation, and the outliers are corrected using the 3σ principle. The expression is: In the formula, μ is the mean, σ is the variance, and x is the data collected by the sensor; The current and voltage data are standardized using the Z-score expression: In the formula, z is the standardized value, which indicates the degree of deviation of the data from the mean; The temperature is normalized using Min-Max and mapped to the interval [0,1], and the stroke is sine and cosine encoded. The expression is: In the formula, x' is the standardized data, max(x) is the maximum value in the data, and min(x) is the minimum value in the data; The feature extraction unit is used to extract key features from the preprocessed data; extract statistical features based on the sliding window and extract time-frequency features based on wavelet changes; The intelligent analysis and calculation unit includes an LSTM network for fault prediction and a reinforcement learning algorithm for control optimization. It analyzes and calculates the real-time operating data collected by sensors based on historical operating data and simulation data, and outputs optimized servo motor control parameters. Among them, the LSTM network is used to model the operating trends of the submersible electric pump and predict potential fault risks, and the reinforcement learning algorithm is used to generate the optimal steering timing, operating speed and stroke control strategy based on the current operating conditions.

4. The ground control system of the submersible electric pump according to claim 1, characterized in that: The control execution module includes: a steering control unit, a speed control unit and a stroke control unit, which are used to receive the analysis results of the intelligent calculation module in real time and automatically execute the adjustment operations on the steering, speed and stroke of the submersible oil pump motor; among them, the steering control unit sends a forward or reverse instruction to the servo motor through the PLC controller according to the steering trigger signal output by the intelligent analysis module to realize the switching of the plunger up and down stroke; the speed control unit adjusts the output frequency of the servo driver based on the target flow and real-time load through fuzzy logic and adaptive PID algorithm to realize dynamic speed control under displacement-load dual feedback; the stroke control unit compares the plunger displacement sensor data in real time to determine whether the stroke meets the effective displacement standard. When deviation occurs, it adjusts the reversal timing or corrects the stroke target value according to its severity to realize closed-loop adjustment of the stroke length.

5. The ground control system of the submersible electric pump according to claim 1, characterized in that: The real-time monitoring and feedback optimization module includes a status monitoring unit and a feedback optimization unit; Among them, the state monitoring unit is used to continuously monitor the operating status of the control execution module after execution. Based on the dynamic threshold model constructed based on the historical operation data, it judges whether the system operating status deviates from the normal range through the mean and variance indicators within the window; The feedback optimization unit is used to feed back the monitoring results to the intelligent computing module in real time to form a closed-loop optimization control.

6. The ground control system of the submersible electric pump according to claim 1, characterized in that: The alarm optimization and adjustment module includes an abnormality diagnosis unit and an alarm response unit; When the system detects an abnormal state, the abnormal diagnosis unit analyzes the cause of the fault based on the parameters of current harmonic characteristics, axial vibration spectrum and well temperature gradient, and generates an optimization adjustment strategy. When the efficiency optimization strategy conflicts with the protective load reduction strategy, the efficiency optimization strategy takes precedence. The alarm response unit sends out alarm signals in a timely manner and performs automatic optimization and adjustment; The optimization adjustment strategy has different strategy weights under different working conditions, giving priority to suppressing gas lock risks in high water content wells and suppressing sand stuck faults in low production wells.

7. A ground control method for a submersible electric pump, characterized in that: The method is used to control the ground control system of the submersible electric pump according to any one of claims 1 to 6, and the method comprises the following steps: S1, collects the voltage, current, temperature and stroke operation data of the submersible pump in real time through sensors; S2 pre-processes the collected submersible pump operating data and analyzes and calculates it through artificial intelligence algorithms to predict the optimal servo motor steering time, speed, and stroke parameters; S3, dynamically controlling the direction, speed and stroke of the motor based on the calculated parameters; S4, continuously monitors the system operating status in real time and provides feedback and optimization; if an abnormal situation is detected, an alarm signal is issued and the control strategy is optimized and adjusted until the system returns to normal operation.

8. The ground control method of a submersible electric pump according to claim 7, characterized in that: In step S2, an artificial intelligence algorithm is used to analyze and calculate the optimal servo motor steering time, speed, and stroke parameters, including: Based on the real-time data collected by the sensor, the motor current fluctuation is used to determine whether the plunger has reached its maximum stroke. The reversal moment of the previous cycle is compared with the current situation to determine the turning moment. According to the well fluid flow demand and stroke length and frequency requirements, combined with the real-time load conditions, the speed that meets the requirements under minimum energy consumption is calculated, and the fuzzy logic control algorithm is introduced to optimize the speed, setting the flow deviation ΔQ=Q target -Q t , Q target is the target flow calculated by the system, Q t is the flow rate value at the current moment, with a domain of [-20%, +20%], which is divided into seven fuzzy sets: NB, NM, NS, ZO, PS, PM, and PB. Each fuzzy set corresponds to a set of speed adjustment policies. The control system can adaptively adjust the weights. When the water content of the well fluid exceeds the set threshold, the energy efficiency weight is automatically adjusted to correct the control target. Real-time analysis of stroke displacement sensor data to identify effective up and down strokes. The effective stroke determination standard is: Δx ≥ 85% x theory , x theory is the theoretical displacement; the acceleration conforms to the pumping characteristics; when the system detects insufficient stroke, the control logic processes it according to the degree of deviation: in case of mild deviation, it is corrected by fine-tuning the steering time; in case of moderate and severe deviation, the stroke target value is corrected in combination with the displacement trends of the previous cycles; when the flow demand suddenly changes, the motor speed is adjusted quickly first, and when the motor speed cannot meet the demand, the stroke and steering logic are adjusted; finally, the calculation of the motor steering time, speed and stroke is realized.

9. The ground control method of a submersible electric pump according to claim 7, characterized in that: In step S3, the direction, speed, and stroke of the motor are dynamically controlled based on the calculated parameters, including: It receives the steering trigger signal calculated by the artificial intelligence algorithm and sends a forward or reverse rotation instruction to the motor controller to make the motor complete the steering switch. The switching signal is received and analyzed by the PLC control unit, and the instruction is transmitted to the motor through the driver interface to realize the up and down reciprocating motion switch of the plunger. Leveraging the servo drive's frequency modulation capabilities, a dynamic frequency modulation control architecture is constructed based on displacement-load dual feedback to achieve speed regulation. The control system comprises an inner loop and an outer loop. The inner loop is a load indicator formed by integrating current, torque, and vibration signals, while the outer loop is real-time displacement. Frequency modulation control uses a fuzzy adaptive PID algorithm to dynamically adjust the output frequency based on load changes and displacement requirements. The system also features a dynamic resonance detection function, identifying potential resonant operating conditions by analyzing frequency response and load fluctuations. A graded response strategy is employed, adjusting the frequency when the condition is mild and entering protection mode to reduce the speed when the condition is severe. The displacement sensor feeds back the real-time position of the plunger to the PLC, and the control execution module compares it with the historical stroke range. If it is determined that the stroke is incomplete or deviates from the preset upper and lower limits, the reversal logic is adjusted and the position trigger threshold is reset to ensure that the actual stroke is consistent with the target stroke, ultimately achieving closed-loop control of the motor's forward and reverse rotation, operating speed, and stroke length.

10. The surface control method of a submersible electric pump according to claim 7, characterized in that: In step S4, the system operation status is continuously monitored in real time and feedback and optimization are performed, including: After detecting an abnormal system state, the control strategy of the submersible pump is dynamically optimized based on the parameter abnormality feedback from the real-time monitoring module. The submersible pump fault level is divided into three levels: Level 1 fault: Motor overload, excessive current harmonics, or stroke exceeding the limit causing equipment damage; Secondary fault: unstable discharge, abnormal vibration, and stroke deviation are recoverable faults; Level 3 fault: vibration change or deviation fault; The alarm module determines the fault level based on the abnormality type and passes the result to the control logic module. Based on the historical trend and current fluctuation range of the abnormal parameters, the steering, speed and stroke are re-evaluated and the established control cycle is adjusted; The optimization module feeds back the corrected control parameters to the control execution module, executes new scheduling instructions, and realizes closed-loop correction and recovery control of the system operation strategy.

Images