Intelligent efficient energy-saving control method and system special for oil and gas field electric drive pump injection equipment

Abstract

The invention relates to the technical field of oil and gas field pumping equipment control. The intelligent efficient energy-saving control method and system special for the electric drive pumping equipment of the oil and gas field comprises the steps that real-time operation data of the electric drive pumping equipment are collected, and the collected real-time operation data are analyzed through the big data analysis technology; on the basis of the analysis result, according to the output pressure and flow requirements of the electric drive pump injection equipment, dynamically adjusting the operation parameters of the electric drive pump injection equipment; the historical operation data and the real-time operation data of the electric drive pump injection equipment are combined, and fault prediction is conducted on the electric drive pump injection equipment through a machine learning algorithm; and obtaining energy consumption data of the electrically-driven pump injection equipment, adjusting a control strategy according to a fault prediction result, and sending fault early warning information. By collecting the real-time operation data of the electric drive pump injection equipment and analyzing the data through the big data analysis technology, the problem that an existing oil and gas field electric drive pump injection equipment control method is insufficient in efficiency, energy consumption and fault processing is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of control of oil and gas field pumping equipment, and more specifically, to a special intelligent and efficient energy-saving control method and system for electric-driven pumping equipment in oil and gas fields. Background Art

[0002] In the field of oil and gas field exploitation, electric-driven pumping equipment plays a crucial role, and the quality of its control method directly affects the efficiency, stability, and intelligent level of oil and gas field production.

[0003] There are many problems to be solved urgently in the existing control methods of electric-driven pumping equipment in oil and gas fields. For example, in terms of efficiency and energy consumption, some traditional drive systems consist of asynchronous motors, belt drives, etc. The transmission efficiency of the asynchronous motor driving the pulley is low, and the supporting belt is easily damaged and needs to be replaced frequently. Although energy-saving measures such as variable frequency speed regulation technology have been adopted, there are still deficiencies in energy consumption data collection, analysis, and precise energy-saving control according to the analysis results, and the efficient utilization of energy cannot be fully realized, resulting in energy waste. Summary of the Invention

[0004] The purpose of the present invention is to provide a special intelligent and efficient energy-saving control method and system for electric-driven pumping equipment in oil and gas fields, aiming to solve the problems of deficiencies in efficiency, energy consumption, and fault handling in the existing control methods of electric-driven pumping equipment in oil and gas fields.

[0005] The present invention is achieved through the following technical solutions:

[0006] A special intelligent and efficient energy-saving control method for electric-driven pumping equipment in oil and gas fields includes the steps of:

[0007] Collect the real-time operation data of the electric-driven pumping equipment, and analyze the collected real-time operation data by using big data analysis technology;

[0008] Based on the analysis results, dynamically adjust the operation parameters of the electric-driven pumping equipment according to the output pressure and flow requirements of the electric-driven pumping equipment;

[0009] Combine the historical operation data and real-time operation data of the electric-driven pumping equipment, and perform fault prediction on the electric-driven pumping equipment through machine learning algorithms;

[0010] Obtain the energy consumption data of the electric-driven pumping equipment, adjust the control strategy according to the fault prediction results, and send fault warning information.

[0011] Optionally, the specific process of collecting the real-time operation data of the electric-driven pumping equipment and analyzing the collected real-time operation data by using big data analysis technology is as follows:

[0012] Utilize sensors distributed at key nodes of the electric drive pumping equipment to collect real-time operation data of the equipment at a preset frequency;

[0013] The collected data are aggregated to a data collection terminal through wired or wireless transmission, and the collected data are preprocessed to obtain a preprocessed data set;

[0014] The features of the preprocessed data set are extracted through data mining algorithms, and the internal connections and potential laws between various operating parameters are identified using big data analysis technology to obtain analysis results.

[0015] Optionally, the specific process of extracting features from the preprocessed data set by a data mining algorithm and identifying the inherent connections and potential rules between different operating parameters by using big data analysis technology to obtain the analysis results is as follows:

[0016] Based on the historical operation data of the electric drive pumping equipment, a machine learning algorithm is used to build an equipment status analysis model, and the equipment status analysis model is trained;

[0017] Clean and normalize the collected real-time operation data to obtain the initial data set;

[0018] Use dimensionality reduction algorithms to extract key features from the initial dataset;

[0019] According to the key features, the intrinsic relationship between the operating parameters is identified through clustering algorithm or classification algorithm to obtain the associated data set;

[0020] The associated data set is input into the trained equipment status analysis model to obtain the analysis results.

[0021] Optionally, the specific process of dynamically adjusting the operating parameters of the electric-driven pumping device based on the analysis result and according to the output pressure and flow rate requirements of the electric-driven pumping device is:

[0022] Extracting parameter information corresponding to the output pressure and flow rate of the electric drive pumping device from the analysis results obtained by the big data analysis;

[0023] Obtain the output pressure and flow rate values ​​that the current electric drive pumping equipment actually needs to achieve;

[0024] Based on the relationship between each parameter and pressure and flow in the analysis results, a mapping relationship is established and an adjustment strategy is determined;

[0025] According to the adjustment strategy, the operating parameters of the electric drive pumping equipment are dynamically adjusted.

[0026] Optionally, the specific process of combining the historical operation data and the real-time operation data of the electric drive pumping device and predicting the fault of the electric drive pumping device through the machine learning algorithm is as follows:

[0027] Integrate the historical operation data and real-time operation data of the electric drive pumping equipment to form a comprehensive data set;

[0028] Perform feature engineering processing on the comprehensive data set to obtain key features that have a significant impact on fault prediction;

[0029] Based on the processed comprehensive data set, select a suitable machine learning algorithm to build a fault prediction model;

[0030] Use a part of the historical operation data as the training set to train the constructed fault prediction model, optimize the model parameters, and improve the accuracy and generalization ability of the model;

[0031] Input the real-time operation data into the trained fault prediction model, and calculate the current fault risk probability of the equipment through the trained fault prediction model;

[0032] According to the preset fault risk threshold, judge whether the equipment has a fault risk. If the fault risk probability exceeds the threshold, it is predicted that the equipment has a fault.

[0033] Optionally, based on the processed comprehensive data set, the specific process of selecting a suitable machine learning algorithm to build a fault prediction model is as follows:

[0034] Build a fault prediction model through the logistic regression algorithm. Assume that the processed comprehensive data set is D = {(x 1 , y 1 ), (x 2 , y 2 ), …, (x n , y n )}, x i ∈R m , y i ∈{0, 1}; where, x i represents the feature vector of the i-th sample; y i represents the label of the i-th sample, 0 indicates no fault, and 1 indicates a fault;

[0035] The expression of the fault prediction model is shown in the following formula (1):

[0036]

[0037] Among them, P(y = 1∣x; θ) represents the probability that x i has a fault; θ represents the parameter vector of the fault prediction model; x represents the feature vector with a bias term added.

[0038] Optionally, a part of the historical operation data is used as a training set to train the constructed fault prediction model, and the specific process of optimizing the model parameters and improving the accuracy and generalization ability of the model is as follows:

[0039] The parameter vector θ of the fault prediction model is optimized through the logarithmic loss function, and the expression of the logarithmic loss function is shown in the following formula (2):

[0040]

[0041] where P(y = 1|x; θ) represents the probability that x i has a fault;

[0042] The gradient descent algorithm is used to minimize the loss function J(θ), and the update formula of the gradient descent algorithm is shown in the following formula (3):

[0043]

[0044] where θ j represents the i-th element of the parameter vector θ; α represents the learning rate; x ij represents the sample x i 's j-th feature.

[0045] Optionally, the specific process of obtaining the energy consumption data of the electric drive pump injection equipment, adjusting the control strategy according to the fault prediction result, and sending a fault warning message is as follows:

[0046] The energy consumption data of the equipment is collected in real time through the energy consumption monitoring sensors installed on the electric drive pump injection equipment;

[0047] The energy consumption data is processed and analyzed by using big data analysis technology to identify the operation modes with abnormal energy consumption or high energy consumption;

[0048] The correlation analysis is carried out between the energy consumption data analysis result and the fault risk probability output by the fault prediction model to judge whether there is abnormal energy consumption or energy waste phenomenon caused by faults in the equipment;

[0049] According to the fault prediction result and the energy consumption analysis result, the operation parameters of the electric drive pump injection equipment are dynamically adjusted;

[0050] When the fault risk probability output by the fault prediction model exceeds the preset threshold, or the energy consumption data analysis result shows that the equipment has abnormal energy consumption, a fault warning message is sent to the operator through the remote monitoring system, indicating that there is a fault risk or energy waste phenomenon in the equipment, and providing corresponding adjustment suggestions.

[0051] Optionally, after adjusting the control strategy, continuously monitor the operating status and energy consumption data of the device to ensure that the device operates stably under the optimized parameters, and further optimize the control strategy based on the real-time data to achieve the energy-saving goal.

[0052] Based on the same inventive concept, the present invention also provides a special intelligent and efficient energy-saving control system for oil and gas field electric drive pump injection equipment, which is used to implement the special intelligent and efficient energy-saving control method for oil and gas field electric drive pump injection equipment, including:

[0053] A data acquisition module, which is used to regularly collect the real-time operating data of the equipment at a preset frequency through sensors distributed at each key node of the electric drive pump injection equipment, and summarize the collected data to a data acquisition terminal through wired or wireless transmission methods;

[0054] A data preprocessing module, which is used to preprocess the collected real-time operating data to obtain a preprocessed data set;

[0055] A big data analysis module, which is used to extract features from the preprocessed data set through data mining algorithms, and use big data analysis technology to identify the internal relationships and potential laws between different operating parameters to obtain an analysis result;

[0056] A dynamic adjustment module, which is used to dynamically adjust the operating parameters of the electric drive pump injection equipment based on the analysis result obtained by the big data analysis module according to the output pressure and flow demand of the electric drive pump injection equipment;

[0057] A fault prediction module, which is used to combine the historical operating data and real-time operating data of the electric drive pump injection equipment, and perform fault prediction on the electric drive pump injection equipment through machine learning algorithms, and output a fault risk probability;

[0058] An energy consumption monitoring module, which is used to collect the energy consumption data of the equipment in real time through energy consumption monitoring sensors installed on the electric drive pump injection equipment, and use big data analysis technology to process and analyze the energy consumption data to identify abnormal energy consumption or high-energy consumption operating modes;

[0059] A control strategy adjustment module, which is used to dynamically adjust the operating parameters of the electric drive pump injection equipment according to the fault risk probability output by the fault prediction module and the energy consumption data analysis result of the energy consumption monitoring module, and send a fault warning message;

[0060] A remote monitoring and warning module, which is used to send a fault warning message to the operator through a remote monitoring system when the fault risk probability output by the fault prediction module exceeds a preset threshold, or the energy consumption data analysis result shows that the equipment has abnormal energy consumption, indicating that there is a fault risk or energy waste phenomenon in the equipment, and providing corresponding adjustment suggestions;

[0061] Continuous optimization module, which is used to continuously monitor the operating status and energy consumption data of the device after adjusting the control strategy, ensure the stable operation of the device under the optimized parameters, and further optimize the control strategy according to the real-time data to achieve the energy-saving goal.

[0062] The technical solution of the present invention has at least the following advantages and beneficial effects:

[0063] Improve energy utilization efficiency: By collecting the real-time operating data of the electric drive pump injection equipment and analyzing these data using big data analysis technology, it is possible to dynamically adjust the operating parameters according to the output pressure and flow requirements of the equipment, effectively reducing energy waste and significantly improving energy utilization efficiency.

[0064] Enhance the operating stability of the equipment: By combining historical operating data with real-time operating data and predicting equipment failures through machine learning algorithms, potential problems can be discovered in advance and corresponding control strategy adjustments can be made, which not only reduces the risk of sudden equipment failures, but also extends the service life of the equipment and improves the overall operating stability.

[0065] Reduce maintenance costs: Through the fault prediction and warning mechanism, equipment failures can be discovered and processed in a timely manner, avoiding production stoppages and maintenance costs caused by equipment damage; at the same time, the maintenance costs brought about by frequent replacement of vulnerable parts such as belts are reduced, further reducing the operating costs.

[0066] Improve the level of intelligence: By adopting big data analysis and machine learning algorithms, intelligent management of the electric drive pump injection equipment is realized, which not only improves the level of automatic control of the equipment, but also provides technical support for the intelligent upgrading of oil and gas field production, promoting the technological progress of the industry.

[0067] Optimize production management: Through real-time data collection and analysis, it is possible to provide a scientific basis for production management, optimize the production process and resource allocation, thereby improving the overall efficiency and economic benefits of oil and gas field production. Description of the Drawings

[0068] Figure 1 It is a schematic flow chart of a special intelligent and efficient energy-saving control method for oil and gas field electric drive pump injection equipment according to an embodiment of the present invention;

[0069] Figure 2 It is a schematic structural diagram of a fan adaptive pitch control system based on environmental data according to an embodiment of the present invention. Detailed Embodiments

[0070] The following are the detailed embodiments in conjunction with the drawings.

[0071] Referring to Figure 1 , a special intelligent and efficient energy-saving control method for oil and gas field electric drive pump injection equipment includes the steps:

[0072] Step 1: Collect the real-time operation data of the electric drive pump injection equipment, and analyze the collected real-time operation data by using big data analysis technology.

[0073] In some embodiments, the specific process of collecting the real-time operation data of the electric drive pump injection equipment and analyzing the collected real-time operation data by using big data analysis technology is as follows:

[0074] Use sensors distributed at each key node of the electric drive pump injection equipment to regularly collect the real-time operation data of the equipment at a preset frequency; at each key node such as the motor, pump body, and pipeline connection part of the electric drive pump injection equipment, reasonably install various sensors, such as pressure sensors, flow sensors, temperature sensors, vibration sensors, and current and voltage sensors. Set a preset frequency, for example, collect data every 10 seconds, to ensure that the changes in the operation state of the equipment can be captured in real time. Obtain the real-time operation data such as pressure, flow, temperature, vibration amplitude, current and voltage of the motor during the operation of the equipment through the sensors.

[0075] Summarize the collected data to the data acquisition terminal through wired or wireless transmission methods, and preprocess the collected data to obtain a preprocessed data set; use wired transmission methods (such as industrial Ethernet cables) or wireless transmission technologies (such as Wi-Fi, Bluetooth, ZigBee, LoRa, etc., select the appropriate wireless technology according to the equipment's environment and distance requirements) to send the data collected by the sensors to the data acquisition terminal. The data acquisition terminal has strong data reception and preliminary management capabilities, and can summarize and organize the data transmitted by different sensors to ensure the integrity and accuracy of the data. Preprocess the collected data at the data acquisition terminal. First, perform data cleaning to remove the error data and outliers caused by sensor failures, signal interference, etc., such as data that significantly exceeds the normal operation parameter range of the equipment. Then perform data denoising, and use filtering algorithms (such as mean filtering, median filtering, etc.) to remove the noise in the data and improve the data quality. Then perform data normalization processing to map different types and different dimensions of data to a specific interval (such as [0,1]) to facilitate subsequent data processing and analysis.

[0076] Extract features from the preprocessed data set through data mining algorithms, and use big data analysis technology to identify the internal relationships and potential laws among the operating parameters to obtain an analysis result.

[0077] In some embodiments, the specific process of extracting features from the preprocessed data set through data mining algorithms and using big data analysis technology to identify the internal relationships and potential laws among different operating parameters to obtain an analysis result is as follows:

[0078] Based on the historical operation data of the electric drive pumping equipment, a machine learning algorithm is used to construct an equipment status analysis model and train the equipment status analysis model; widely collect the historical operation data of the electric drive pumping equipment under different working conditions, including the data during normal operation, abnormal operation and different production stages of the equipment. These data cover parameters such as the rotational speed, current, voltage of the motor, the flow rate, pressure, temperature of the pump, and the vibration, displacement of each component of the equipment. According to the data characteristics and analysis objectives, a suitable machine learning algorithm, such as a neural network (such as a multi-layer perceptron MLP), is selected to construct the equipment status analysis model. The neural network has a powerful non-linear mapping ability and can capture complex relationships of operation parameters. Divide the collected historical operation data into a training set and a test set, usually in a ratio of 70%-30% or 80%-20%. Use the training set data to train the constructed neural network model, and continuously adjust the weights and biases of the model through the backpropagation algorithm, so that the model can accurately perform feature learning and pattern recognition on the input data. During the training process, set appropriate training parameters, such as the learning rate, the number of iterations, the number of hidden layer nodes, etc., to optimize the performance of the model.

[0079] Clean and normalize the collected real-time operation data to obtain an initial data set; clean the collected real-time operation data, remove duplicate data, error data and outliers caused by sensor failures or interference. For example, by setting reasonable parameter ranges to screen out data points that deviate significantly from the normal range and correct or eliminate them. Adopt a normalization method, such as min-max normalization or Z-score normalization, to convert real-time operation data with different dimensions into a unified numerical interval, making the data comparable. For example, map data with different ranges such as motor current and voltage to the [0,1] interval.

[0080] Use a dimensionality reduction algorithm to extract key features from the initial data set; use a dimensionality reduction algorithm such as principal component analysis (PCA) to process the normalized initial data set. The PCA algorithm converts high-dimensional data into low-dimensional data through linear transformation, while removing redundant information in the data while retaining the main features of the data. For example, convert high-dimensional data containing multiple operation parameters into several principal components, which can reflect the variance of the original data to the greatest extent, that is, the main change characteristics of the data.

[0081] Based on the key features, identify the internal relationships between operating parameters through clustering algorithms or classification algorithms to obtain an associated dataset; according to the extracted key features, use a clustering algorithm (such as the K-Means clustering algorithm) to perform clustering analysis on the data. The K-Means algorithm divides the data points into K clusters, such that the data points within the same cluster have a high degree of similarity, and the data points between different clusters have a low degree of similarity. Through clustering analysis, the operating state patterns of the device under different combinations of operating parameters can be discovered, and the internal relationships between parameters under similar operating states can be identified. For example, cluster the device operating data under different working conditions to find the associations between parameters such as motor speed and pump efficiency under high-flow and low-pressure conditions. Or use a classification algorithm (such as the decision tree classification algorithm) to classify the data. The decision tree algorithm constructs a decision tree model based on the features and labels of the data, and classifies the operating state of the device by judging the key features. For example, according to features such as motor current and pump pressure, classify the device operating state into categories such as normal operation, mild abnormality, and severe abnormality, so as to clarify the relationships and laws of different operating parameters under different states.

[0082] Input the associated dataset into the trained device state analysis model to obtain the analysis result. Input the associated dataset obtained through clustering or classification into the previously trained device state analysis model. The model analyzes and predicts the associated dataset based on the learned features and patterns, and finally obtains the analysis result regarding the operating state of the electric drive pump injection device, including the quantitative relationships between various operating parameters, the optimal operating parameter combinations under different working conditions, etc. By accurately identifying the internal relationships between operating parameters, the operating parameters of the electric drive pump injection device can be adjusted more precisely according to actual production requirements. For example, during the oil and gas field exploitation process, according to different oil reservoir pressures and flow rate requirements, quickly find the optimal combination of motor speed and pump operating frequency to ensure that the device is always in an efficient operating state, avoid low production efficiency caused by unreasonable parameter settings, and thus improve the exploitation efficiency of the oil and gas field.

[0083] Step 2: Based on the analysis result, dynamically adjust the operating parameters of the electric drive pump injection device according to the output pressure and flow rate requirements of the electric drive pump injection device.

[0084] In some embodiments, the specific process of dynamically adjusting the operating parameters of the electric drive pump injection device based on the analysis result according to the output pressure and flow rate requirements of the electric drive pump injection device is as follows:

[0085] Extract the parameter information corresponding to the output pressure and flow rate of the electric drive pump injection device from the analysis result obtained from big data analysis; this parameter information includes the speed of the motor, the stroke length of the pump, the opening degree of the valve, etc. For example, through data analysis, determine the specific associated values between the motor speed and the output pressure and flow rate under different working conditions, and obtain the actual output pressure value and flow rate value that the current electric drive pump injection device needs to reach;

[0086] Based on the relationships between the parameters in the analysis results and pressure and flow rate, establish a mapping relationship and determine an adjustment strategy; use a pressure sensor and a flow sensor to obtain in real time the actual output pressure value and flow rate value that the current electric drive pump injection equipment needs to achieve. These sensors convert the measured physical quantities into electrical signals and transmit them to the signal processing unit of the control system for processing and analysis. Based on the relationships between the parameters in the analysis results and pressure and flow rate, establish an accurate mapping relationship. For example, through a large amount of experimental data and data analysis, determine how much the output pressure and flow rate will increase correspondingly when the motor speed increases by a certain value. According to these mapping relationships and combined with the actual requirements, determine the adjustment strategy. If the current actual required output pressure value is higher than the pressure value generated by the current equipment operation, and the analysis results indicate that increasing the motor speed can effectively increase the pressure, then the adjustment strategy may be to appropriately increase the motor speed.

[0087] According to the adjustment strategy, dynamically adjust the operating parameters of the electric drive pump injection equipment. According to the determined adjustment strategy, the control system sends control instructions to the actuator of the electric drive pump injection equipment to dynamically adjust the operating parameters of the equipment. For example, adjust the motor speed through a frequency converter and adjust the valve opening through an electric actuator, etc., to make the output pressure and flow rate meet the actual requirements. By dynamically adjusting the operating parameters of the electric drive pump injection equipment in real time, its output pressure and flow rate can accurately match the actual requirements during the oil and gas field exploitation process, avoiding the problem of low production efficiency caused by unreasonable parameters. For example, in different stages of oil and gas field exploitation, the requirements for pressure and flow rate are different, and this method can timely adjust the equipment parameters to ensure the efficient progress of the production process.

[0088] Step 3: Combine the historical operation data and real-time operation data of the electric drive pump injection equipment, and perform fault prediction on the electric drive pump injection equipment through machine learning algorithms.

[0089] In some embodiments, the specific process of combining the historical operation data and real-time operation data of the electric drive pump injection equipment and performing fault prediction on the electric drive pump injection equipment through machine learning algorithms is as follows:

[0090] Integrate the historical operation data and real-time operation data of the electric drive pumping equipment to form a comprehensive data set; establish a data storage structure, for example, use a relational database (such as MySQL) or a non-relational database (such as MongoDB) to store the historical operation data and real-time operation data. The historical operation data usually includes various operation parameter records of the equipment over a past period of time, such as motor speed, current, voltage, pump flow rate, pressure, etc., and these data are stored in chronological order. The real-time operation data is collected in real time by sensors distributed at various key nodes of the electric drive pumping equipment, obtained regularly at a preset frequency (such as once per second or once per minute), and aggregated to the data acquisition terminal through wired or wireless transmission methods (such as Ethernet, Wi-Fi, 4G / 5G, etc.). Use data fusion technology to integrate the historical operation data and real-time operation data from different data sources and different formats. For example, for data collected by different sensors, there may be different timestamps and data accuracies, and time alignment and data standardization processing are required to ensure that all data is fused on the same time scale and data format to form a comprehensive data set. Integrating the historical operation data and real-time operation data can comprehensively reflect the operation status of the electric drive pumping equipment and avoid the limitations of a single data source. Through feature engineering processing, extract key features and remove redundant and noise information, so that the model can more accurately capture the relationship between equipment operation parameters and faults, thereby improving the accuracy of fault prediction. For example, through the analysis of a large amount of historical data and real-time data, it is found that features such as the fluctuation of motor current and the change rate of pump flow rate are highly correlated with equipment faults. Inputting these key features into the fault prediction model can more accurately judge whether there is a fault risk in the equipment.

[0091] Perform feature engineering processing on the comprehensive data set to obtain key features that have a significant impact on fault prediction; data cleaning: remove noise data and outliers from the comprehensive data set. By setting reasonable data thresholds, such as when the motor current exceeds the maximum or minimum value of the normal operating range, it can be determined as an outlier and removed; for missing values, methods such as mean filling, median filling, or prediction filling based on machine learning algorithms are used for processing. Data normalization: Normalize the feature data with different dimensions and value ranges so that they are on the same scale. For example, for two parameters with different dimensions, such as motor speed and pump pressure, the Min-Max normalization method can be used to map the data to the [0,1] interval.

[0092] Based on the processed comprehensive data set, select a suitable machine learning algorithm to construct a fault prediction model;

[0093] In some embodiments, based on the processed comprehensive data set, the specific process of selecting a suitable machine learning algorithm to construct a fault prediction model is as follows:

[0094] Construct a fault prediction model through the logistic regression algorithm. Assume that the processed comprehensive dataset is D = {(x 1 , y 1 ), (x 2 , y 2 ), …, (x n , y n ), x i ∈ R m , y i ∈ {0, 1}; where x i represents the feature vector of the i-th sample; y i represents the label of the i-th sample, 0 indicates no fault, and 1 indicates a fault;

[0095] The expression of the fault prediction model is shown in Equation (1) below:

[0096]

[0097] Among them, P(y = 1|x; θ) represents the probability that x i has a fault; θ represents the parameter vector of the fault prediction model; x represents the feature vector with a bias term added.

[0098] Use a part of the historical operation data as the training set to train the constructed fault prediction model, optimize the model parameters, and improve the accuracy and generalization ability of the model;

[0099] In some embodiments, the specific process of using a part of the historical operation data as the training set to train the constructed fault prediction model, optimize the model parameters, and improve the accuracy and generalization ability of the model is as follows:

[0100] A part can be divided from the historical operation data as the training set (e.g., 70% - 80%), and another part as the test set (e.g., 20% - 30%).

[0101] Through the logarithmic loss function, optimize the parameter vector θ of the fault prediction model. The expression of the logarithmic loss function is shown in Equation (2) below:

[0102]

[0103] Among them, P(y = 1|x; θ) represents the probability that x i has a fault;

[0104] Use the gradient descent algorithm to minimize the loss function J(θ). The update formula of the gradient descent algorithm is shown in Equation (3) below:

[0105]

[0106] Among them, θj represents the \(i\)-th element of the parameter vector \(\theta\); \(\alpha\) represents the learning rate; \(x\) ij represents the sample \(x\) i 's \(j\)-th feature.

[0107] Metrics such as accuracy, recall, and F1 value can be used to evaluate the performance of the model. Accuracy represents the proportion of samples correctly predicted by the model among the total samples; Recall represents the proportion of positive samples (faulty samples) correctly predicted by the model among the actual positive samples; The F1 value is the harmonic mean of accuracy and recall, comprehensively considering the accuracy and recall ability of the model. If the performance of the model is not ideal, the performance of the model can be improved by adjusting the hyperparameters of the model (such as the learning rate \(\alpha\), regularization parameter, etc. in logistic regression), increasing the training data, performing feature selection or feature engineering, etc.

[0108] Input the real-time operation data into the trained fault prediction model, and calculate the current fault risk probability of the device through the trained fault prediction model;

[0109] According to the preset fault risk threshold, determine whether the device has a fault risk. If the fault risk probability exceeds the threshold, it is predicted that the device will fail. By selecting an appropriate machine learning algorithm and training and optimizing the model, the relationship between the characteristics and faults in the operation data of the electric drive pump injection device can be better captured, thereby improving the accuracy of fault prediction. For example, the logistic regression algorithm can accurately calculate the probability of the device failing based on the operation parameters of the device, providing a reliable basis for the maintenance and management of the device.

[0110] Step Four: Obtain the energy consumption data of the electric drive pump injection device, adjust the control strategy according to the fault prediction result, and send a fault warning message.

[0111] In some embodiments, the specific process of obtaining the energy consumption data of the electric drive pump injection device, adjusting the control strategy according to the fault prediction result, and sending a fault warning message is as follows:

[0112] Through the energy consumption monitoring sensors installed on the electric drive pump injection device, the energy consumption data of the device is collected in real time; High-precision energy consumption monitoring sensors are installed at each key energy consumption node of the electric drive pump injection device, such as the motor, oil pump, etc. These sensors collect the energy consumption data such as voltage, current, and power of the device in real time at a set time interval (for example, every 5 minutes). The collected data is summarized to the data acquisition terminal through wired Ethernet or wireless Wi-Fi, Bluetooth and other transmission methods. For example, current sensors and voltage sensors are installed at the input end of the motor to monitor the power consumption of the motor in real time.

[0113] Use big data analysis technology to process and analyze energy consumption data, and identify abnormal or high-energy-consuming operation modes; transmit the collected energy consumption data to the big data analysis platform. On the platform, use data cleaning algorithms to remove noise and outliers in the data, and then use statistical analysis methods, such as calculating the average value, standard deviation, and change trend of energy consumption, to identify abnormal or high-energy-consuming operation modes. For example, by comparing historical data, if the power consumption of a device is significantly higher than the normal range within a certain period of time and the duration exceeds a certain threshold (such as 30 minutes), it is determined as a high-energy-consuming operation mode. At the same time, use the clustering algorithm in machine learning algorithms to perform clustering analysis on the energy consumption data under different working conditions, find operation modes with similar energy consumption, and further analyze their characteristics and causes.

[0114] Perform correlation analysis on the energy consumption data analysis results and the failure risk probability output by the failure prediction model to determine whether there is abnormal energy consumption or energy waste caused by faults in the equipment; correlate the energy consumption data analysis results with the failure risk probability output by the failure prediction model. Establish a correlation analysis model, such as using a Bayesian network model, to analyze the causal relationship between the failure risk and abnormal energy consumption. If the failure prediction model shows that a certain key component (such as a motor bearing) has a high failure risk probability, and the energy consumption data shows that the energy consumption of the motor has increased abnormally, it can be judged through correlation analysis that there may be abnormal energy consumption caused by faults in the equipment.

[0115] Dynamically adjust the operation parameters of the electric drive pump injection equipment according to the failure prediction results and energy consumption analysis results; dynamically adjust the operation parameters of the electric drive pump injection equipment according to the failure prediction results and energy consumption analysis results. If it is found that there is abnormal energy consumption caused by faults in the equipment and the failure risk probability exceeds the preset threshold, immediately reduce the operation load of the equipment, such as reducing the speed of the motor and reducing the displacement of the oil pump, etc., to avoid further damage to the equipment and energy waste. At the same time, formulate corresponding maintenance plans according to the type of failure and abnormal energy consumption. For example, when it is found that the failure risk of the motor bearing is high, arrange to replace the bearing during the next shutdown maintenance of the equipment.

[0116] When the failure risk probability output by the failure prediction model exceeds the preset threshold, or the energy consumption data analysis result shows that the device has abnormal energy consumption, a failure warning message is sent to the operator through the remote monitoring system, indicating that there is a failure risk or energy waste phenomenon in the device, and providing corresponding adjustment suggestions. The warning message can be sent to the operator in various ways such as text messages, emails, and APP push. For example, when it is detected that the energy consumption of a certain part of the device has increased abnormally and there may be a failure risk, the system automatically sends a text message to the operator's mobile phone, including the device name, failure location, failure type, abnormal energy consumption situation, and preliminary adjustment suggestions, such as "[Device name] has abnormal energy consumption at [specific location], there may be a [failure type], it is recommended to immediately check and reduce the device operation load.

[0117] In some embodiments, after adjusting the control strategy, continuously monitor the operating status and energy consumption data of the device to ensure that the device operates stably under the optimized parameters, and further optimize the control strategy based on the real-time data to achieve the energy-saving goal. With the help of the data acquisition module and the energy consumption monitoring module, continuously collect the real-time operating data and energy consumption data of the electric drive pump injection equipment at a high frequency (such as once per second). For example, use a pressure sensor to monitor the output pressure of the device, use a flow sensor to monitor the medium flow rate, and use current and voltage sensors to obtain the energy consumption data. Transmit the collected data to the data storage center quickly through wired or wireless transmission methods. A distributed database such as the Hadoop Distributed File System (HDFS) can be used to ensure the secure storage and efficient reading of data, providing data support for subsequent analysis. Use data visualization technology to display the device operating status and energy consumption data in an intuitive chart form on the monitoring interface. For example, use a line chart to show the change trend of energy consumption over time, and present the real-time values of key operating parameters in the form of a dashboard, facilitating the operator to grasp the device status in real time. Continuously monitor the working conditions of the device under the optimized operating parameters, and compare the actual operating parameters with the preset optimized parameters. If it is found that parameters such as the actual motor speed and pump displacement deviate from the optimized values, immediately analyze the reasons. Once abnormal device operation is detected, use the fault prediction module and big data analysis module for rapid diagnosis. For example, if the device vibration increases abnormally, analyze the historical data and real-time monitoring data to determine whether it is caused by faults such as bearing wear, and arrange maintenance personnel to repair it in time to ensure the stable operation of the device. Consider the impact of environmental factors such as temperature and humidity on the device performance during operation. Install environmental monitoring sensors to collect environmental data in real time. When the environmental parameters exceed the suitable operating range of the device, take corresponding adjustment measures, such as starting the cooling system or dehumidification equipment, to ensure that the device operates in a stable environment. Regularly conduct in-depth analysis and mining of the large amount of real-time data collected. Use machine learning algorithms, such as reinforcement learning algorithms, to find better control strategies based on the real-time operating status and energy consumption data of the device. For example, by continuously trying different combinations of motor speed and pump displacement, combined with the energy consumption data, determine the most energy-saving operating parameter combination under the current working conditions. Update and adjust the control strategy in a timely manner according to the data analysis results. Apply the new control strategy to the dynamic adjustment module and the control strategy adjustment module to achieve the re-optimization of the device operating parameters. For example, when it is found that the original pressure-flow control strategy is no longer energy-saving under a certain new working condition, determine the new control parameters through an optimization algorithm and apply them to the actual device control. Establish a feedback mechanism to feedback the optimized device operation effect to the data analysis link. Continuously evaluate the implementation effect of the new control strategy, continuously iterate and optimize the control strategy, and gradually achieve the energy-saving goal. For example, compare the energy consumption data before and after optimization. If the energy-saving effect is not obvious, re-analyze the data and further optimize the control strategy.

[0118] Based on the same inventive concept, corresponding to the positioning tooling of any of the above embodiments, referring to Figure 2 , the present invention provides a special intelligent and energy-efficient control system for oil and gas field electric drive pump injection equipment, which is used to implement the above-mentioned special intelligent and energy-efficient control method for oil and gas field electric drive pump injection equipment, and includes:

[0119] A data acquisition module, which is used to regularly collect the real-time operation data of the equipment at a preset frequency through sensors distributed at each key node of the electric drive pump injection equipment, and summarize the collected data to a data acquisition terminal through wired or wireless transmission methods;

[0120] A data preprocessing module, which is used to preprocess the collected real-time operation data to obtain a preprocessed data set;

[0121] A big data analysis module, which is used to extract features from the preprocessed data set through data mining algorithms, and use big data analysis technology to identify the internal relationships and potential laws between different operation parameters to obtain analysis results;

[0122] A dynamic adjustment module, which is used to dynamically adjust the operation parameters of the electric drive pump injection equipment based on the analysis results obtained by the big data analysis module according to the output pressure and flow requirements of the electric drive pump injection equipment;

[0123] A fault prediction module, which is used to combine the historical operation data and real-time operation data of the electric drive pump injection equipment, and perform fault prediction on the electric drive pump injection equipment through machine learning algorithms, and output a fault risk probability;

[0124] An energy consumption monitoring module, which is used to collect the energy consumption data of the equipment in real time through energy consumption monitoring sensors installed on the electric drive pump injection equipment, and use big data analysis technology to process and analyze the energy consumption data to identify abnormal energy consumption or high-energy consumption operation modes;

[0125] A control strategy adjustment module, which is used to dynamically adjust the operation parameters of the electric drive pump injection equipment according to the fault risk probability output by the fault prediction module and the energy consumption data analysis results of the energy consumption monitoring module, and send a fault warning message;

[0126] A remote monitoring and warning module, which is used to send a fault warning message to the operator through a remote monitoring system when the fault risk probability output by the fault prediction module exceeds a preset threshold, or the energy consumption data analysis results show that the equipment has abnormal energy consumption, to prompt that there is a fault risk or energy waste phenomenon in the equipment, and provide corresponding adjustment suggestions;

[0127] A continuous optimization module, which is used to continuously monitor the operation status and energy consumption data of the equipment after adjusting the control strategy, ensure that the equipment operates stably under the optimized parameters, and further optimize the control strategy according to the real-time data to achieve the energy-saving goal.

Claims

1. An intelligent and efficient energy-saving control method for electric drive pumping equipment in oil and gas fields, characterized in that: Includes steps: Collect the real-time operation data of the electric drive pumping equipment, and use big data analysis technology to analyze the collected real-time operation data; Based on the analysis results, the operating parameters of the electric-driven pumping equipment are dynamically adjusted according to the output pressure and flow requirements of the electric-driven pumping equipment; Combine the historical and real-time operation data of the electric-driven pumping equipment to predict faults of the electric-driven pumping equipment through machine learning algorithms; Obtain energy consumption data of electric drive pumping equipment, adjust control strategy according to fault prediction results, and send fault warning information.

2. The intelligent, efficient and energy-saving control method for oil and gas field electric drive pumping equipment according to claim 1 is characterized in that: The specific process of collecting the real-time operation data of the electric drive pump injection equipment and analyzing the collected real-time operation data using big data analysis technology is as follows: Utilize sensors distributed at key nodes of the electric drive pumping equipment to collect real-time operation data of the equipment at a preset frequency; The collected data are aggregated to a data collection terminal through wired or wireless transmission, and the collected data are preprocessed to obtain a preprocessed data set; The features of the preprocessed data set are extracted through data mining algorithms, and the internal connections and potential laws between various operating parameters are identified using big data analysis technology to obtain analysis results.

3. The intelligent, efficient and energy-saving control method for oil and gas field electric drive pumping equipment according to claim 2 is characterized in that: The specific process of extracting features from the preprocessed data set by using a data mining algorithm and identifying the inherent connections and potential rules between different operating parameters using big data analysis technology to obtain the analysis results is as follows: Based on the historical operation data of the electric drive pumping equipment, a machine learning algorithm is used to build an equipment status analysis model, and the equipment status analysis model is trained; Clean and normalize the collected real-time operation data to obtain the initial data set; Use dimensionality reduction algorithms to extract key features from the initial dataset; According to the key features, the intrinsic relationship between the operating parameters is identified through clustering algorithm or classification algorithm to obtain the associated data set; The associated data set is input into the trained equipment status analysis model to obtain the analysis results.

4. The intelligent, efficient and energy-saving control method for oil and gas field electric drive pumping equipment according to claim 1, characterized in that: The specific process of dynamically adjusting the operating parameters of the electric drive pumping equipment based on the analysis results and the output pressure and flow requirements of the electric drive pumping equipment is as follows: Extracting parameter information corresponding to the output pressure and flow rate of the electric drive pumping device from the analysis results obtained by the big data analysis; Obtain the output pressure and flow rate values ​​that the current electric drive pumping equipment actually needs to achieve; Based on the relationship between each parameter and pressure and flow in the analysis results, a mapping relationship is established and an adjustment strategy is determined; According to the adjustment strategy, the operating parameters of the electric drive pumping equipment are dynamically adjusted.

5. The intelligent, efficient and energy-saving control method for electric drive pumping equipment in oil and gas fields as claimed in claim 1, characterized in that: The specific process of combining the historical operation data and real-time operation data of the electric drive pumping equipment and predicting the fault of the electric drive pumping equipment through the machine learning algorithm is as follows: Integrate historical and real-time operation data of electric drive pumping equipment to form a comprehensive data set; Perform feature engineering on the comprehensive data set to obtain key features that have a significant impact on fault prediction; Based on the processed comprehensive data set, select the appropriate machine learning algorithm to build a fault prediction model; Use part of the historical operation data as a training set to train the constructed fault prediction model, optimize the model parameters, and improve the accuracy and generalization ability of the model; Input the real-time operation data into the trained fault prediction model, and calculate the current fault risk probability of the equipment through the trained fault prediction model; Based on the preset fault risk threshold, determine whether the equipment has a fault risk. If the fault risk probability exceeds the threshold, predict that the equipment will fail.

6. The intelligent, efficient and energy-saving control method for oil and gas field electric drive pumping equipment according to claim 5, characterized in that: Based on the processed comprehensive data set, the specific process of selecting a suitable machine learning algorithm and building a fault prediction model is as follows: The fault prediction model is constructed by the logistic regression algorithm. Assuming that the processed comprehensive data set is D = {(x1, y1), (x2, y2), …, (x n ,y n )},x i ∈R m ,y i ∈{0,1}; where x o represents the feature vector of the i-th sample; y i represents the label of the i-th sample, 0 means no fault, 1 means fault; The expression of the fault prediction model is shown in the following formula (1): Where P(y=1|x;θ) represents x i is the probability of failure; θ represents the parameter vector of the fault prediction model; x represents the feature vector with the bias term added.

7. The intelligent, efficient and energy-saving control method for oil and gas field electric drive pumping equipment according to claim 6 is characterized in that: The specific process of using a part of the historical operation data as a training set to train the constructed fault prediction model, optimize the model parameters, and improve the accuracy and generalization ability of the model is as follows: The parameter vector θ of the fault prediction model is optimized through the logarithmic loss function. The expression of the logarithmic loss function is shown in the following formula (2): Where P(y=1|x;θ) represents x i is the probability of failure; Use the gradient descent algorithm to minimize the loss function J(θ). The update company of the gradient descent algorithm is shown in the following formula (3): Among them, θ j represents the i-th element of the parameter vector θ; α represents the learning rate; x ij Represents sample x i The jth feature of .

8. The intelligent, efficient and energy-saving control method for electric drive pumping equipment in oil and gas fields as claimed in claim 1, characterized in that: The specific process of obtaining the energy consumption data of the electric drive pumping equipment, adjusting the control strategy according to the fault prediction result, and sending the fault warning information is as follows: Energy consumption data of the equipment can be collected in real time by installing energy consumption monitoring sensors on the electric drive pumping equipment; Use big data analysis technology to process and analyze energy consumption data and identify abnormal or high energy consumption operation modes; Correlate the energy consumption data analysis results with the fault risk probability output by the fault prediction model to determine whether the equipment has abnormal energy consumption or energy waste caused by the fault; Dynamically adjust the operating parameters of the electric drive pumping equipment according to the fault prediction results and energy consumption analysis results; When the failure risk probability output by the fault prediction model exceeds the preset threshold, or the energy consumption data analysis results show that the equipment has abnormal energy consumption, a fault warning message is sent to the operator through the remote monitoring system to prompt the operator that the equipment has a failure risk or energy waste, and provide corresponding adjustment suggestions.

9. The intelligent, efficient and energy-saving control method for oil and gas field electric drive pumping equipment according to claim 8, characterized in that: After adjusting the control strategy, continuously monitor the operating status and energy consumption data of the equipment to ensure that the equipment operates stably under the optimized parameters, and further optimize the control strategy based on real-time data to achieve energy-saving goals.

10. An intelligent and efficient energy-saving control system for oil and gas field electric drive pumping equipment, used to implement the intelligent and efficient energy-saving control method for oil and gas field electric drive pumping equipment as described in any one of claims 1 to 9, characterized in that: include: The data acquisition module is used to collect the real-time operation data of the equipment at a preset frequency through sensors distributed at various key nodes of the electric drive pumping equipment, and summarize the collected data to the data acquisition terminal through wired or wireless transmission; A data preprocessing module is used to preprocess the collected real-time operation data to obtain a preprocessed data set; The big data analysis module is used to extract features from the preprocessed data set through data mining algorithms, and use big data analysis technology to identify the inherent connections and potential laws between different operating parameters to obtain analysis results; A dynamic adjustment module is used to dynamically adjust the operating parameters of the electric drive pumping equipment according to the output pressure and flow requirements of the electric drive pumping equipment based on the analysis results obtained by the big data analysis module; The fault prediction module is used to combine the historical operation data and real-time operation data of the electric drive pumping equipment, predict the faults of the electric drive pumping equipment through machine learning algorithms, and output the fault risk probability; Energy consumption monitoring module, which is used to collect energy consumption data of the equipment in real time through energy consumption monitoring sensors installed on the electric drive pumping equipment, and process and analyze the energy consumption data using big data analysis technology to identify abnormal or high energy consumption operation modes; A control strategy adjustment module is used to dynamically adjust the operating parameters of the electric drive pumping equipment and send fault warning information according to the fault risk probability output by the fault prediction module and the energy consumption data analysis results of the energy consumption monitoring module; The remote monitoring and early warning module is used to send fault warning information to operators through the remote monitoring system when the fault risk probability output by the fault prediction module exceeds the preset threshold, or the energy consumption data analysis results show that the equipment has abnormal energy consumption, prompting the equipment to have a fault risk or energy waste phenomenon, and providing corresponding adjustment suggestions; The continuous optimization module is used to continuously monitor the operating status and energy consumption data of the equipment after adjusting the control strategy, to ensure that the equipment operates stably under the optimized parameters, and to further optimize the control strategy based on real-time data to achieve energy-saving goals.

Images