Oil well dynamic data early warning method and system

Abstract

The invention relates to an oil well dynamic data early warning method and system. The method comprises the following steps that multiple pieces of variable data about oil well production dynamics are obtained; one variable is selected from the multiple variable data to serve as a dependent variable, and other variables serve as independent variables; performing correlation calculation on the different independent variable data according to the dependent variable data, obtaining mutual information values of the different independent variable data, and screening out the independent variable data related to the dependent variable data according to the mutual information values; performing principal component analysis on the related independent variable data and dependent variables to obtain principal component data; dividing the principal component data and the dependent variable data into a preset number of training sets and test sets, and predicting the dependent variable data for preset time according to a tested model to obtain predicted dependent variable data; and the dependent variable data and the predicted dependent variable data are compared with a preset dependent variable early warning range, so that early discovery and early warning of dynamic abnormal changes of oil well production are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of oil well data processing, and in particular to a method and system for early warning of oil well dynamic data. Background Art

[0002] Traditionally, production early warning of oil wells is conducted by production management or researchers by checking the daily production reports or production dynamic curves of oil wells. Wells that contribute greatly to production are included in the list of key wells for management if their dynamic data changes greatly. This early warning method mainly relies on manpower, which reduces the timeliness of early warning.

[0003] At the same time, the dynamic data of oil wells covers a wide range, including production data, pump working conditions, pump inlet temperature and pressure, manifold nozzle, wellhead temperature and pressure, etc. These data are affected by many factors such as geology, reservoir, and working system. Because the relationship between data is complex, it is difficult for traditional research methods to obtain high-precision prediction results. Taking water cut as an example to illustrate the difficulty of prediction, first of all, the water cut is on the rise, and it changes complexly over time, making it difficult to accurately predict; secondly, the increase in water cut is affected by many factors, such as geological factors (interlayers, interlayers), reservoir factors (edge ​​and bottom water drive types), working system (liquid lifting, liquid reduction), well interference, wellbore factors, etc., making modeling difficult; thirdly, there are many factors affecting water cut, the data is high-dimensional, the amount of information is large, and feature selection and model design are difficult; fourthly, some oil wells are about to break through the water cut stage, the water cut changes rapidly, and the water cut is difficult to predict. This limits the research methods of conventional reservoir engineering methods and reduces the prediction accuracy. In the process of production dynamic tracking and management, such oil wells require prediction models with high accuracy and real-time performance. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for early warning of oil well dynamic data in view of the problems existing in the prior art.

[0005] The technical invention adopted by the present invention to solve the technical problem is: a method for early warning of oil well dynamic data, comprising the following steps:

[0006] S1: Obtain multiple variable data on oil well production performance;

[0007] S2: Select one variable from the plurality of variable data as the dependent variable, and the other variables as independent variables; wherein the data corresponding to the dependent variable is the dependent variable data, and the data corresponding to the independent variable is the independent variable data;

[0008] S3: performing correlation calculation on different independent variable data according to the dependent variable data, and obtaining mutual information values ​​of different independent variable data, and screening out independent variable data related to the dependent variable data according to the mutual information values;

[0009] S4: performing principal component analysis on the relevant independent variable data and the dependent variable to obtain principal component data;

[0010] S5: dividing the principal component data and the dependent variable data into a preset number of training sets and test sets, inputting the training sets into a preset model to perform optimization training on model hyperparameters to obtain a trained model, applying the test set to the trained model to verify the reliability and generalization ability of the trained preset model, and obtaining a tested model;

[0011] S6: predicting the dependent variable data for a preset time according to the tested model to obtain predicted dependent variable data;

[0012] S7: Comparing the dependent variable data and the predicted dependent variable data with a preset dependent variable warning range respectively to obtain a comparison result;

[0013] S8: issuing an early warning of the oil well production status based on the comparison result.

[0014] In one embodiment, S2 includes: preprocessing the multiple variable data to obtain multiple cleaned variable data, and selecting a cleaned variable from the multiple cleaned variable data as a dependent variable, and the other cleaned variables as independent variables.

[0015] In one embodiment, the preprocessing includes at least one of missing value processing, outlier processing, and data standardization processing.

[0016] In one embodiment, the step of calculating the correlation between different independent variable data and obtaining mutual information values ​​of different independent variable data according to the dependent variable data includes:

[0017] The mutual information value between the dependent variable data and different independent variable data is calculated by using the mutual information calculation formula.

[0018] In one embodiment, the calculation formula of the mutual information is:

[0019] I(X;Y)=∑∑p(x,y)*log(p(x,,y) / (p(x)*p(y)))

[0020] Where X represents the independent variable, Y represents the dependent variable, p(x,y) represents the joint probability of the independent variable X and the dependent variable Y, p(x) and p(y) represent the probabilities of the independent variable X and the dependent variable Y respectively, and I(x;y) is the mutual information value.

[0021] In one embodiment, the preset model includes: an LSTM neural network.

[0022] In one embodiment, the step of inputting the training set into a preset model for training to obtain a trained model comprises:

[0023] Genetic algorithm and particle swarm optimization algorithm are used to optimize the neural network hyperparameters of LSTM through the training set to obtain a trained model.

[0024] In one embodiment, the S7 includes:

[0025] The change rates of the dependent variable data and the predicted dependent variable data within a preset time are calculated respectively, and the change rates are compared with the preset early warning change rates to obtain early warning results.

[0026] In one embodiment, the plurality of variable data on oil well production performance include: oil production, water production, gas production, water content, wellhead pressure and wellhead temperature.

[0027] The present application also provides a method for early warning of oil well dynamic data, comprising a processor and a memory storing a computer program, wherein the memory stores a program, and the program is loaded and executed by the processor to implement any of the above-mentioned methods for early warning of oil well dynamic data.

[0028] The implementation of the present invention has the following beneficial effects: by screening variables directly related to the oil well production dynamics as independent variables and calculating their mutual information values ​​with the dependent variables, this method effectively reduces data noise and redundancy, and significantly improves the accuracy and reliability of the early warning model. Subsequently, the principal component analysis is used to reduce the dimension of the screened data, which not only reduces the computational complexity, but also enhances the generalization ability of the model, so that it can flexibly respond to data fluctuations under different working conditions. This early warning system can predict future dependent variables based on a preset model and compare them with the preset early warning range to achieve early detection and early warning of abnormal changes in oil well production dynamics, providing a basis for timely intervention for oilfield managers, effectively preventing production accidents, and ensuring production safety. At the same time, the early warning results help oilfield managers scientifically adjust production plans, optimize resource allocation, improve production efficiency and economic benefits, and promote the transformation of oilfields to intelligent and refined management. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0030] Figure 1 It is a flow chart of an embodiment of a method for early warning of oil well dynamic data provided by the present invention;

[0031] Figure 2 It is a flow chart of another embodiment of the method for early warning of oil well dynamic data provided by the present invention;

[0032] Figure 3The method for early warning of oil well dynamic data provided by the present invention is the main parameters of the daily production of oil wells;

[0033] Figure 4 It is a diagram of the variable analysis result of the oil well water content in the method for early warning of oil well dynamic data provided by the present invention;

[0034] Figure 5 It is a result diagram of principal component analysis of water content in oil wells in the method for early warning of oil well dynamic data provided by the present invention;

[0035] Figure 6 It is a structural diagram of the LSTM time series neural network in the oil well dynamic data early warning method provided by the present invention;

[0036] Figure 7 It is a diagram of the prediction result of the water content of an oil well in the method for early warning of the oil well dynamic data provided by the present invention;

[0037] Figure 8 It is a forecast graph of the water content of the oil well in the next month in the method for early warning of the oil well dynamic data provided by the present invention;

[0038] Fig. 9 It is a probability distribution diagram of the water content of the oil well in the method for early warning of the oil well dynamic data provided by the present invention;

[0039] Fig.10 It is an oil well water content early warning diagram in the oil well dynamic data early warning method provided by the present invention. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical invention in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] like Figure 1 and Figure 2 As shown, the method for early warning of oil well dynamic data provided by the present invention comprises the following steps:

[0042] S1: Obtain multiple variable data on oil well production performance;

[0043] Specifically, Figure 3 As shown, in one embodiment, the variables include but are not limited to daily oil production, daily water production, daily gas production, tested water cut, oil nozzle, wellhead pressure, wellhead temperature, sand content, frequency, bottom hole pressure, pump inlet pressure, pump outlet pressure, pump inlet temperature, daily liquid production, water cut, etc.

[0044] S2: Select one variable from multiple variable data as the dependent variable, and the other variables as independent variables; wherein the data corresponding to the dependent variable is the dependent variable data, and the data corresponding to the independent variable is the independent variable data.

[0045] S21: Preprocessing the plurality of variable data to obtain cleaned plurality of variable data, and selecting one cleaned variable from the cleaned plurality of variable data as the dependent variable, and the other cleaned variables as the independent variables. The preprocessing includes at least one of missing value processing, outlier processing, and data standardization processing.

[0046] Specifically, preprocessing of multiple variable data aims to effectively solve problems such as missing, abnormal, repeated, inconsistent and unreasonable data, thereby improving the explanatory power of the data, reducing the complexity of model training, and significantly enhancing the prediction and analysis efficiency of the model. In view of the uniqueness of oil well data, data preprocessing includes accurately filling missing values, carefully identifying and adjusting outliers, and standardizing the data using the range standardization method to ensure the numerical consistency of each variable. Subsequently, one variable is selected from the cleaned data as the research object-dependent variable. As an example, water cut is selected as the dependent variable, and the remaining variables naturally become independent variables (such as oil production, water production, bottom hole pressure, etc.) to deeply analyze or predict the dynamic changes of water cut.

[0047] like Figure 4 As shown, S3: performing correlation calculation on different independent variable data according to dependent variable data, and obtaining mutual information values ​​of different independent variable data, and screening out independent variable data related to dependent variable data according to the mutual information values.

[0048] S31: Calculate the mutual information value between the dependent variable data and different independent variable data by using the calculation formula of mutual information. The calculation formula of mutual information is:

[0049] I(X;Y)=∑∑p(x,y)*log(p(x,y) / (p(x)*p(y)))

[0050] Where X represents the independent variable, Y represents the dependent variable, p(x,y) represents the joint probability of the independent variable X and the dependent variable Y, p(x) and p(y) represent the probabilities of the independent variable X and the dependent variable Y respectively, and I(x;y) is the mutual information value.

[0051] Specifically, in order to deeply analyze the complex relationship between the variables in the oil well dynamic data, the mutual information analysis tool is used. Mutual information can not only transcend the constraints of linear relationships and capture the potential nonlinear correlation between variables, but also directly reflect the strength of the independent variable's predictive ability on the dependent variable. Specifically, the degree of correlation between the dependent variable (water cut) and multiple independent variables (such as production time, daily oil production, nozzle, wellhead temperature, etc.) is quantified by calculating the mutual information value between them. In this process, the mutual information value I (x; y) is calculated based on the calculation formula of mutual information, that is, considering the joint probability p (x, y) of the independent variable X and the dependent variable Y and their respective probabilities p (x) and p (y). In the analysis of oil well dynamic data, the mutual information method is used to effectively screen out 13 parameters that are highly correlated with water cut (such as daily oil production, nozzle, pump inlet pressure, cumulative oil production, etc.), which will be used as key input variables in the subsequent principal component analysis. By eliminating irrelevant parameters, not only the complexity of data analysis is simplified, but also the accuracy and reliability of the analysis results are improved.

[0052] S4: Perform principal component analysis on the relevant independent variable data and dependent variables to obtain principal component data.

[0053] like Figure 5 As shown in the figure, specifically, the principal component analysis (PCA) data dimensionality reduction technology is used. PCA converts the original n-dimensional variables (such as the 13 parameters closely related to water content in this example: daily oil production, oil nozzle, pump inlet pressure, cumulative oil production, etc.) into a set of new, mutually unrelated variables, namely, principal components, through linear transformation. This process not only significantly reduces the amount of data, but also retains the main information in the data, thereby greatly simplifying the complexity of the analysis and improving the efficiency of the analysis.

[0054] Specifically, in the implementation of PCA, for n-dimensional variables (X1, X2, ..., X n ) The covariance matrix C is obtained by calculation, where C ij is the covariance of the i-th feature and the j-th feature. The covariance matrix C is:

[0055]

[0056] Among them, C ij is the covariance of the i-th feature and the j-th feature. Then, the covariance matrix is ​​decomposed by eigenvalue. The eigenvalue decomposition is to decompose the covariance matrix C into:

[0057] C=VΛV T

[0058] in,

[0059] V T V=I, where I is the unit matrix, V is an orthogonal matrix, and is the eigenvalue matrix. The main diagonal elements of the eigenvalue matrix are eigenvalues, which are arranged from large to small. The size of the eigenvalue represents the amount of variance contributed by each principal component, that is, the amount of data information. Based on the size of the eigenvalue, the first p (4 in this case, i.e. PC1 to PC4) eigenvectors with the largest eigenvalues ​​are selected, and these vectors construct a low-dimensional space. By projecting the original data into this low-dimensional space, a new feature matrix, i.e., the principal component, is obtained.

[0060] The analysis results show that the two principal components PC1 and PC2 contain nearly 80% of the data information, while the four principal components (PC1 to PC4) together contain more than 90% of the data information. This result not only verifies the effectiveness of PCA in data dimensionality reduction, but also shows that the selected four principal components can well represent the core characteristics of the original data. Through principal component analysis, the originally complex 13 variables are simplified into 4 easy-to-handle variables, which not only reduces the difficulty of data processing, but also provides a clearer and more accurate data basis for subsequent in-depth analysis and prediction. Figure 5 The results of this principal component analysis are presented visually.

[0061] S5: Divide the principal component data and dependent variable data into a preset number of training sets and test sets, input the training set into the preset model to perform optimization training on the model hyperparameters, obtain a trained model, use the test set on the trained model to verify the reliability and generalization ability of the trained model, and obtain a tested model;

[0062] In one embodiment, the preset model includes: LSTM neural network. Wherein, inputting the training set into the preset model for training to obtain the trained model includes: using genetic algorithm and particle swarm optimization algorithm to optimize the LSTM neural network hyperparameters through the training set to obtain the trained model.

[0063] This application uses a genetic algorithm and particle swarm optimization algorithm to implement a set of optimization strategies for the hyperparameters of the LSTM (Long Short-Term Memory Network) model, namely the learning rate and the number of neurons in the hidden layer. The specific implementation steps first involve randomly initializing a particle swarm, where each particle represents a set of potential optimization parameter combinations. Subsequently, during the training of the LSTM model, the root mean square error (RMSE) is used as the error calculation formula, and is also used as a loss function to evaluate the model performance, with the goal of minimizing the loss function. In order to quantify the pros and cons of the particles, a fitness function is designed, whose value is the inverse of the RMSE, to measure the degree to which the parameter combination represented by each particle improves the performance of the LSTM model.

[0064] like Figure 6As shown in the figure, during the optimization process, if the best fitness in the current particle swarm meets the preset conditions or the number of iterations reaches the upper limit, the optimization process is terminated and the global optimal particle is output. Otherwise, the algorithm will apply the selection, hybridization (crossover) and mutation mechanisms in the genetic algorithm to update the particle swarm, generate a new particle swarm, and re-evaluate the fitness of each particle. This process is iterated until the stopping condition is met. Finally, the parameter combination represented by the global optimal particle is determined as the optimal hyperparameter of the LSTM model, thereby constructing a LSTM neural network model with optimized performance. In terms of prediction accuracy, the structural design of the LSTM time series neural network also fully considers the characteristics of time series data, such as Figure 6 As shown in the figure, C1 is the cell state at time t1, h1 is the hidden layer state at time t1, and at time t, f t For the forget gate, i t is the input gate, C t is the cell state, o t is the output gate, h t is the current unit output, h t-1 is the unit output at the previous moment, is a candidate memory cell, σ is a sigmoid activation function, which determines which values ​​will be updated, and tanh is a hyperbolic cosine activation function, which is used to activate candidate memory cells. This ensures the stability and accuracy of the model when processing complex time series data. The root mean square error, as an evaluation indicator of model accuracy, provides an intuitive and quantitative evaluation method to help continuously optimize and improve model performance.

[0065] The LSTM neural network update formula is:

[0066] f t =σ(W f ·[h t-1 , Xt]+b f )

[0067] i t =σ(W i ·[h t-1 , Xt]+b i )

[0068]

[0069] o t =σ(W o ·[h t-1 , Xt]+b o )

[0070] h t =o t *tanh(C t )

[0071]

[0072] Where W f , W i , W o , W c and b f 、b i 、b o 、b c are the weights and biases of the forget gate, input gate, output gate, and candidate memory cells respectively, and Xt is the independent variable at time t.

[0073] The LSTM model accuracy evaluation index uses the root mean square error, and the specific formula is:

[0074]

[0075] Where n is the number of prediction samples; y i is the actual value; ci is the predicted value.

[0076] Specifically, in the dynamic prediction of oil well production, in view of the complexity, high dimension and time series characteristics of its data, an optimized LSTM neural network model is used to improve the prediction accuracy and real-time performance. The proportion of the research object data set can be divided according to actual needs. In this embodiment, 70% is used as a training set and 30% is used as a test set, that is, the principal component data and the dependent variable data are divided according to the ratio of 70% training set and 30% test set to ensure the independence and reliability of model training and testing. In order to further improve the performance of the LSTM neural network, the key parameters of the LSTM model (such as learning rate and number of hidden layer neurons) are optimized by integrating genetic algorithm and particle swarm optimization algorithm. With its advantages of strong global search capability and high population diversity, the genetic algorithm is combined with the characteristics of fast convergence and information sharing of the particle swarm algorithm to effectively avoid local optimal solutions and accelerate the optimization process.

[0077] In the specific optimization process, the inverse of the root mean square error (RMSE) is used as the fitness function, and the particle swarm is continuously updated iteratively to find the LSTM model parameter combination that minimizes the RMSE. After multiple iterations, the global optimal particle is obtained, and the parameters represented by it are set as the optimal parameters of the LSTM neural network. Subsequently, the LSTM model is trained using these optimized parameters. The mean square error is used as the loss function during the training process, and 100 iterations of training are performed with the support of the Adam optimizer to ensure the stability and accuracy of the model. Finally, an LSTM neural network model that is dual-optimized by the genetic algorithm and the particle swarm optimization algorithm is obtained. This model can not only effectively handle the complex time series relationship in the dynamic data of oil well production, but also improve the real-time performance of the prediction while ensuring the prediction accuracy. By applying this model to the test set data, its effectiveness and reliability in practical applications are verified, providing strong support for oil well production management and decision-making.

[0078] S6: Predict the dependent variable data for a preset time period based on the tested model to obtain predicted dependent variable data.

[0079] like Figure 7 and Figure 8 Specifically, after completing the optimization and training of the LSTM neural network model and verifying its effectiveness and reliability on the test set, this carefully constructed model is then used to predict the water cut in oil well production for the next 30 days. This step is based on the trained LSTM model, taking historical data and current production status as input, and capturing and simulating the law of water cut changes over time through the complex calculation process within the model, especially using its excellent time series processing capabilities. Finally, the predicted data for water cut in the next 30 days is obtained.

[0080] S7: Compare the dependent variable data and the predicted dependent variable data with the preset dependent variable warning range to obtain a comparison result.

[0081] Further, S7 includes: respectively calculating the change rate of the dependent variable data and the predicted dependent variable data within a preset time, and comparing the change rate with a preset warning change rate.

[0082] S8: Issue early warning on the oil well production status based on the comparison results.

[0083] Specifically, compared with the traditional 3σ criterion, such as Fig. 9As shown, some variables of oil well dynamic data do not obey the normal distribution. In order to reduce the early warning judgment error rate caused by the non-normal distribution of dynamic data, this technology uses oil well planned production indicators as the judgment standard for quantitative early warning. Specifically, first, the dependent variable (such as the preset range of water content and the threshold of its change rate are set. These thresholds are dynamically adjusted according to the production plan indicators of the oil well. For example, the increase rate of water content is considered normal if it does not exceed 3%. By calculating the change rate of measured data and predicted data within a preset time (such as one day, two days, etc.), and comparing them in detail with these thresholds, the system can accurately identify abnormal fluctuations and generate intuitive early warning diagrams based on this. Fig.10 , clearly showing the changing trends and early warning status of key indicators such as water content. By adopting the planned production indicators of oil wells as the criteria for quantitative early warning, and adjusting these criteria in real time to match changes in production plans, the present invention provides a more scientific and accurate basis for management decisions. When the early warning system sends out an abnormal signal, management personnel can respond quickly, deeply analyze the cause of the abnormality based on the actual situation of the oil well, and take effective measures to deal with it, thereby significantly improving the management efficiency and safety of oil well production. This intelligent early warning and rapid response mechanism provides a strong guarantee for the sustained and stable operation of oil well production.

[0084] In order to more precisely evaluate the changing trend of moisture content, the change rates of measured data and predicted data within a preset time are calculated, where the preset time can be one day, two days, one month, etc., and these change rates are compared in detail with the preset moisture content change rate threshold (such as no more than 3% for a normal day). This quantitative analysis method not only improves the accuracy of early warning, but also provides a scientific basis for management decisions.

[0085] Based on the above comparison results, an intuitive early warning map was generated using the intelligent early warning system. Fig.10 It clearly displays the changing trend of water content and the early warning status, allowing managers to understand the oil well production situation at a glance. When the early warning system sends out an abnormal signal, it can quickly combine the actual situation of the oil well, deeply analyze the cause of the abnormality, and take corresponding treatment measures, thereby effectively improving the management efficiency and safety of oil well production. Through this intelligent early warning and rapid response mechanism, it can provide a strong guarantee for the stable operation of oil well production.

[0086] The present application also provides a system for early warning of oil well dynamic data, including a processor and a memory storing a computer program, wherein the memory stores a program, and the program is loaded and executed by the processor to implement the method for early warning of oil well dynamic data as described in any one of the above items.

[0087] Beneficial effects: By selecting variables directly related to oil well production dynamics as independent variables and calculating their mutual information values ​​with dependent variables, this method effectively reduces data noise and redundancy, and significantly improves the accuracy and reliability of the early warning model. Subsequently, principal component analysis is used to reduce the dimension of the screened data, which not only reduces the computational complexity, but also enhances the generalization ability of the model, enabling it to flexibly respond to data fluctuations under different working conditions. This early warning system can predict future dependent variables based on the tested model and compare them with the preset early warning range to achieve early detection and early warning of abnormal changes in oil well production dynamics, providing a basis for timely intervention for oilfield managers, effectively preventing production accidents and ensuring production safety. At the same time, the early warning results help oilfield managers scientifically adjust production plans, optimize resource allocation, improve production efficiency and economic benefits, and promote the transformation of oilfields to intelligent and refined management. Ultimately, this early warning mechanism significantly reduces the overall operating costs of oilfields and improves overall operating efficiency by reducing economic losses caused by equipment failures and production interruptions and reducing unnecessary maintenance costs.

[0088] It can be understood that the above embodiments only express the preferred implementation modes of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the patent scope of the present invention. It should be pointed out that, for ordinary technicians in this field, the above technical features can be freely combined without departing from the concept of the present invention, and several deformations and improvements can be made, which all belong to the protection scope of the present invention. Therefore, all equivalent changes and modifications made to the scope of the claims of the present invention should belong to the coverage of the claims of the present invention.

Claims

1. A method for early warning of oil well dynamic data, characterized in that: include: S1: Obtain multiple variable data on oil well production performance; S2: Select one variable from the plurality of variable data as the dependent variable, and the other variables as independent variables; wherein the data corresponding to the dependent variable is the dependent variable data, and the data corresponding to the independent variable is the independent variable data; S3: performing correlation calculation on different independent variable data according to the dependent variable data, and obtaining mutual information values ​​of different independent variable data, and screening out independent variable data related to the dependent variable data according to the mutual information values; S4: performing principal component analysis on the relevant independent variable data and the dependent variable to obtain principal component data; S5: dividing the principal component data and the dependent variable data into a preset number of training sets and test sets, inputting the training sets into a preset model to perform optimization training on model hyperparameters to obtain a trained model, applying the test set to the trained model to verify the reliability and generalization ability of the trained model, and obtaining a tested model; S6: predicting the dependent variable data for a preset time according to the tested model to obtain predicted dependent variable data; S7: Comparing the dependent variable data and the predicted dependent variable data with a preset dependent variable warning range respectively to obtain a comparison result; S8: issuing an early warning of the oil well production status based on the comparison result.

2. The method for early warning of oil well dynamic data according to claim 1, characterized in that: The S2 includes: The multiple variable data are preprocessed to obtain cleaned multiple variable data, and one cleaned variable is selected from the cleaned multiple variable data as a dependent variable, and the other cleaned variables are independent variables.

3. The method for early warning of oil well dynamic data according to claim 1, characterized in that: The preprocessing includes at least one of missing value processing, outlier processing, and data standardization processing.

4. The method for early warning of oil well dynamic data according to claim 1, characterized in that: The performing correlation calculation on different independent variable data according to the dependent variable data and obtaining mutual information values ​​of different independent variable data includes: The mutual information value between the dependent variable data and different independent variable data is calculated by using the mutual information calculation formula.

5. The method for early warning of oil well dynamic data according to claim 4, characterized in that: The calculation formula of the mutual information is: I(X;Y)=∑∑p(x,y)*log(p(x,y) / (p(x)*p(y)) Where X represents the independent variable, Y represents the dependent variable, p(x,y) represents the joint probability of the independent variable X and the dependent variable Y, p(x) and p(y) represent the probabilities of the independent variable X and the dependent variable Y respectively, and I(x;y) is the mutual information value.

6. The method for early warning of oil well dynamic data according to claim 1, characterized in that: The preset model includes: LSTM neural network.

7. The method for early warning of oil well dynamic data according to claim 6, characterized in that: The training set is input into a preset model for training to obtain a trained preset model, comprising: Genetic algorithm and particle swarm optimization algorithm are used to optimize the neural network hyperparameters of LSTM through the training set to obtain a trained model.

8. The method for early warning of oil well dynamic data according to claim 1, characterized in that: The S7 includes: The change rates of the dependent variable data and the predicted dependent variable data within a preset time are calculated respectively, and the change rates are compared with the preset early warning change rates.

9. The method for early warning of oil well dynamic data according to claim 1, characterized in that: The multiple variable data on oil well production dynamics include: oil production, water production, gas production, water content, wellhead pressure and wellhead temperature.

10. A system for early warning of oil well dynamic data, comprising a processor and a memory storing a computer program, characterized in that: The memory stores a program, which is loaded and executed by the processor to implement the method for early warning of oil well dynamic data as described in any one of claims 1 to 9.

Images