CN114398828A - Drilling rate intelligent prediction and optimization method, system, equipment and medium - Google Patents

Abstract

The invention relates to a drilling rate intelligent prediction and optimization method, a system, equipment and a medium, which comprises the following steps: preprocessing the acquired original drilling data to obtain a sample set; training the built drilling speed prediction model based on the sample set to obtain a trained drilling speed prediction model; and inputting the real-time drilling data of the block to be predicted into the drilling speed prediction model for prediction to obtain the mechanical drilling speed. According to the invention, on one hand, real-time data and results of well logging and well logging are selected from the data, on the other hand, a BP neural network with strong nonlinear fitting capability and a genetic algorithm with strong nonlinear optimization capability are selected from the data processing method, so that the data and algorithm processes are reduced. Therefore, the invention can be widely applied to the field of petroleum exploration and development.

Description

A kind of drilling speed intelligent prediction and optimization method, system, equipment and medium

technical field

The invention relates to the field of petroleum exploration and development, in particular to a method, system, equipment and medium for intelligent prediction and optimization of drilling speed based on real-time drilling data in drilling engineering.

Background technique

In recent years, with the increasing demand for oil and gas resources and increasing efforts in exploration and development, my country's oil and gas development field is expanding from conventional oil and gas resources to low-permeability, deep, ultra-deep, deep water, shale oil and gas and other unconventional oil and gas resources. This also brings new challenges to drilling: the geological conditions are more complex, the environment is more severe, and the operating conditions are more severe, making the mining cost even higher. Therefore, oil exploration technology urgently needs further development.

In order to better carry out exploration and development, improve drilling efficiency, reduce drilling costs, and reduce drilling risks, one of the most effective methods is to increase the ROP. Optimizing the ROP is one of the urgent problems to be solved in drilling engineering, and it is also a key project in drilling engineering. During the drilling process, the ROP of drilling can be accurately predicted. On the one hand, it can monitor and prevent the occurrence of drilling accidents in advance, reduce drilling risks, and improve drilling safety; on the other hand, it can optimize drilling based on real-time prediction. The method provides strong support, so that the drilling cycle is reduced and the cost is reduced.

From different perspectives, many scholars have explored the method of combining multiple means to improve the ROP. However, there are still many problems in the current research:

(1) With the continuous improvement of the informatization level of major oil companies and the rapid development of downhole measurement tools over the years, a large amount of historical and real-time data has been accumulated in the drilling technology system. These data not only have dozens of types, but also Moreover, the volume of data is huge, and there is currently a lack of processes, specifications and standardized methods for effectively processing and utilizing real-time data;

(2) The underground situation in drilling is complex, and there are too many factors affecting ROP (rate of penetration), and various data are not completely independent, and there is a complex internal connection between them, which makes it difficult to build a model, and lacks the basis of Drilling optimization methods for real-time data prediction.

SUMMARY OF THE INVENTION

In view of the above-mentioned problems, the purpose of the present invention is to provide a method, system, equipment and medium for intelligent prediction and optimization of ROP, analyze the influence of drilling parameters on ROP based on real drilling data, use BP neural network to construct a ROP model, and use A large amount of relevant data is used for training and verification, and then the optimal solution is globally obtained through the genetic algorithm, that is, the maximum average ROP value is found, and the corresponding parameter value is obtained.

To achieve the above object, the present invention adopts the following technical solutions:

In a first aspect, the present invention provides a method for intelligent prediction and optimization of drilling speed, which includes the following steps:

Preprocess the acquired raw drilling data to obtain a sample set;

The established drilling rate prediction model is trained based on the sample set, and the trained drilling rate prediction model is obtained;

Input the real-time drilling data of the block to be predicted into the ROP prediction model for prediction, and obtain the ROP.

Further, the method for preprocessing the acquired original drilling data to obtain a sample set includes:

Obtain the original drilling data; perform data cleaning on the original drilling data; standardize the cleaned drilling data to obtain a sample set.

Further, the method for data cleaning on the original drilling data includes: missing value processing, abnormal value processing, repeated value processing and sampling value processing.

Further, the method for training the established drilling rate prediction model based on the sample set to obtain the trained drilling rate prediction model, including:

Establish a drilling rate prediction model; obtain training set, validation set and test set based on the sample set, use the training set to train the drilling rate prediction model, and use the validation set and test set to verify the training results.

Further, the drilling rate prediction model is a three-layer BP neural network, including an input layer, an intermediate layer and an output layer;

The input layer is used for receiving drilling data in the training set;

The middle layer is used for processing drilling data to obtain the optimized ROP;

The output layer is used to output the ROP.

Further, the method for training the drilling rate prediction model includes: using the training set to train the drilling rate prediction model, and using the verification set and the test set to test and verify the trained model until the model verification requirements are met.

Further, the model verification requirement means that the correlation R2 between the ^ROP predicted by the model and the actual ROP in the training set, the test set and the validation set satisfies the set minimum ^R2 requirement of the model.

In a second aspect, the present invention provides an intelligent prediction and optimization system for drilling speed, characterized in that the system includes:

The sample set acquisition module is used to preprocess the acquired original drilling data to obtain a sample set;

The model training module is used to train the established drilling rate prediction model based on the sample set to obtain the trained drilling rate prediction model;

The ROP prediction module is used to input the real-time drilling data of the block to be predicted into the ROP prediction model for prediction to obtain the ROP.

In a third aspect, the present invention provides a processing device, the processing device includes at least a processor and a memory, and a computer program is stored on the memory, and the processor executes the computer program to implement the drill. The steps of the fast intelligent prediction and optimization method.

In a fourth aspect, the present invention provides a computer storage medium on which computer-readable instructions are stored, and the computer-readable instructions can be executed by a processor to implement the steps of the method for intelligent prediction and optimization of penetration rate.

The present invention has the following advantages due to taking the above technical solutions:

(1) Aiming at the problems of a wide variety of data, huge data volume and complex internal relationships, the present invention adopts a machine learning method (such as a deep neural network model), and does not independently study the influence of one factor, but conducts all data analysis. Comprehensive learning analysis to reduce the impact of human experience;

(2) In order to ensure the real-time performance of the model in practical application, on the one hand, the present invention selects the real-time data and results of logging and logging in the data, and on the other hand, selects the BP with strong nonlinear fitting ability in the data processing method. The neural network and algorithm are easy to parallelize and the genetic algorithm with strong nonlinear optimization ability reduces the data and algorithm process;

(3) In order to ensure the practicability of the machine learning model, the purpose of the present invention is to find out the maximum average drilling speed for a certain formation, and obtain its corresponding parameters to guide drilling;

Therefore, the present invention can be widely used in the field of petroleum exploration and development.

Description of drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are for the purpose of illustrating preferred embodiments only and are not to be considered limiting of the invention. The same reference numerals are used to refer to the same parts throughout the drawings. In the attached image:

Fig. 1 is the flow chart of ROP intelligent prediction and optimization method in the embodiment of the present invention;

Fig. 2 is the training set correlation of ROP prediction model in the embodiment of the present invention;

Fig. 3 is the validation set correlation of the ROP prediction model in the embodiment of the present invention;

Fig. 4 is the test set correlation of the ROP prediction model in the embodiment of the present invention;

Fig. 5 is the correlation R ² of all data of predicted ROP and actual ROP in the embodiment of the present invention;

6 is a comparison of depth-based predicted ROP and actual ROP in an embodiment of the present invention;

FIG. 7 is a comparison between an actual ROP based on depth and an optimized ROP in an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, embodiments of the present invention. Based on the described embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art fall within the protection scope of the present invention.

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

In some embodiments of the present invention, a method for intelligent prediction and optimization of ROP is provided, which includes the following steps: preprocessing the acquired original drilling data to obtain a sample set; training the established ROP prediction model based on the sample set, The trained ROP prediction model is obtained; the real-time drilling data of the block to be predicted is input into the ROP prediction model for prediction, and the ROP is obtained. Aiming at the problems of various types of data, huge data volume and complex internal relationships, the present invention adopts a machine learning method (such as a deep neural network model) to conduct comprehensive learning and analysis on all data to reduce the influence of human experience. at the same time,

Correspondingly, in other embodiments of the present invention, there is also provided an intelligent prediction and optimization system, equipment and medium for drilling speed.

Example 1

As shown in FIG. 1 , the present embodiment provides a method for intelligent prediction and optimization of penetration rate, including the following steps:

Step 1: Preprocess the acquired raw drilling data to obtain a sample set.

Specifically, the above step 1 can be implemented by the following steps: acquiring original drilling data; performing data cleaning on the original drilling data; and standardizing the cleaned drilling data to obtain a sample set.

In a preferred embodiment, the method for obtaining the original drilling data is as follows: select a block, and organize all the drilling data, including the whole logging data and logging data, etc., into an Excel table as the original drilling data. .

In a preferred embodiment, the data cleaning method for raw drilling data includes: missing value processing, outlier processing, repeated value processing and sampling value processing.

Among them, missing value processing refers to: compensating or deleting the missing part of the data generated in the collection process according to the actual situation.

Outlier processing refers to re-compensating or deleting abnormal data with errors due to various reasons according to the actual situation.

Duplicate value processing refers to: performing duplicate detection and deletion on the repeatedly recorded data.

Sampling value processing refers to: processing different sampling frequencies of each parameter to make it reach the same sampling depth to obtain a processed data set.

In a preferred embodiment, the method for standardizing the cleaned drilling data includes: scaling the drilling data according to a preset ratio, so as to prevent the magnitude of the parameters or the magnitude of the data from being too different to cause "large" The phenomenon of "counting and eating decimals" causes "small" data to be ineffective or less effective.

Step 2: Train the established drilling rate prediction model based on the sample set to obtain a trained drilling rate prediction model.

Specifically, the above step 2 can be realized by the following steps: establishing a drilling rate prediction model; obtaining a training set, a verification set and a test set based on the sample set, which are used to train, verify and test the drilling rate prediction model respectively.

In a preferred embodiment, the drilling rate prediction model adopts a three-layer BP neural network, which includes an input layer, an intermediate layer and an output layer. Among them, the input layer is used to receive the drilling data in the training set; the middle layer is used to process the drilling data to obtain the optimized ROP. The middle layer is provided with appropriate layers and nodes, which can increase the depth of the neural network. Improves the ability to characterize nonlinear models; the output layer is used to output the resulting ROP.

In a preferred embodiment, the method for obtaining a training set, a validation set and a test set based on a sample set includes: sampling from the sample set by random sampling according to a preset ratio to obtain a training set, a validation set and a test set. Preferably, the preset proportions of the training set, the validation set and the test set are 70%, 15% and 15%, respectively.

In a preferred embodiment, a method for training a drilling rate prediction model includes: using a training set to train the drilling rate prediction model, and using a verification set and a test set to test and verify the trained model, until it satisfies the Model validation requirements.

Preferably, the model verification requirement means that the correlation R2 between the ^ROP predicted by the model and the actual ROP in the training set, the test set and the validation set satisfies the set minimum ^R2 requirement of the model.

Optimizing drilling parameters, using the characteristics of self-organization, self-adaptation and intelligence of genetic algorithm, by adjusting and setting controllable drilling parameters, namely WOB and rotational speed RPM, using genetic algorithm to update parameters, find out the BP neural network training results. The maximum value of the complex function is to find the maximum average ROP value, and obtain the corresponding WOB and rotational speed RPM.

Step 3: Input the real-time drilling data of the block to be predicted into the ROP prediction model for prediction, and obtain the ROP.

Example 2

This embodiment takes the real-time data in the drilling process of a certain block as an example, and introduces the method of Embodiment 1 in detail, including the following steps:

Step 1: Select a block, and organize all the available drilling data, that is, the whole logging data and logging data, into an Excel table as the original drilling data;

Step 2: Data preprocessing on the original drilling data;

Step 3: Standardize the preprocessed drilling data, and scale the data to the range of (0, 1) according to the proportion to prevent the magnitude of the parameters or the magnitude of the data from being too different to cause "big numbers eat decimals". ” phenomenon, causing “small” data to be ineffective or less effective. Part of the processed data is shown in Table 1 below.

Part of the processed data in Table 1

Step 4: Establish a ROP prediction model, determine that the input layer vector of the three-layer BP neural network is the drilling data after standard processing, the number of layers in the middle layer is set to one layer and 30 neurons, and the output layer is the actual ROP;

Step 5: The training data set uses 70% of the randomly sampled preprocessed data to train the model, and uses the remaining 30% of the data as the test and verification set to test and verify the trained model;

As shown in Figures 2 to 4, the correlation R2 of the training set (Figure ² ), the correlation of the verification set (Figure 3), and the correlation of the test set, respectively, are the predicted ROP and the actual ROP. (Fig. 4); The lowest R ² of the model validation set is selected to be 85%, so the model meets the accuracy requirements;

Step 6: Use all the data of training and validation tests to finally establish a prediction model, and the correlation R ² between the predicted ROP and the actual ROP is shown in Figure 5, and the depth-based comparison of its ROP is shown in Figure 6;

Step 7: Optimizing drilling parameters, using the characteristics of self-organization, self-adaptation and intelligence of genetic algorithm, by adjusting and setting controllable drilling parameters (weight-on-bit WOB and rotational speed RPM), the BP neural network can be found based on simple binary genetic coding. The maximum value of the complex function trained by the network is to find the maximum average ROP value and obtain the corresponding parameter value. The depth-based comparison between the actual ROP and the optimized ROP is shown in Figure 7. Average ROP improved by 35.7%.

The method for intelligent prediction and optimization of drilling rate based on real-time drilling data provided by the embodiment of the present invention uses the entire real-time logging data of wells in the same block to train the model, so that the model has strong versatility in this block; The embodiment of the invention also performs prediction optimization based on real-time drilling data, which enables the model to be effectively used in actual drilling work and to guide the work. In addition, the strong nonlinearity and regularity of ROP and related parameters are difficult to characterize by conventional methods. Machine learning models are good at dealing with nonlinear problems, and can further optimize parameters after real-time modeling, which is of great significance for guiding field operations. Using the neural network model in machine learning, through the training and verification of side wells, logging, and logging data, the genetic algorithm is used to solve the global optimal ROP, and the highest average ROP and its corresponding control parameter values are obtained.

Example 3

The above embodiment 1 provides a method for intelligent prediction and optimization of the drilling rate, and correspondingly, this embodiment provides an intelligent prediction and optimization system for the drilling rate. The identification system provided in this embodiment can implement the method for intelligent prediction and optimization of drilling speed in Embodiment 1, and the system can be implemented by software, hardware, or a combination of software and hardware. For example, the system may include integrated or separate functional modules or functional units to perform corresponding steps in each method of Embodiment 1. Since the system of this embodiment is basically similar to the method embodiment, the description process of this embodiment is relatively simple, and reference may be made to part of the description of Embodiment 1 for related parts, and the embodiment of the system of this embodiment is only illustrative.

A drilling rate intelligent prediction and optimization system provided by this embodiment includes:

The sample set acquisition module is used to preprocess the acquired original drilling data to obtain a sample set;

The model training module is used to train the established drilling rate prediction model based on the sample set to obtain the trained drilling rate prediction model;

The ROP prediction module is used to input the real-time drilling data of the block to be predicted into the ROP prediction model for prediction to obtain the ROP.

Example 4

This embodiment provides a processing device corresponding to the method for intelligent prediction and optimization of drilling rate provided in Embodiment 1. The processing device may be a processing device used for a client, such as a mobile phone, a notebook computer, a tablet computer, a desktop computer, etc. , to perform the method of Example 1.

The processing device includes a processor, a memory, a communication interface and a bus, and the processor, the memory and the communication interface are connected through the bus to complete mutual communication. A computer program that can be run on the processor is stored in the memory, and when the processor runs the computer program, the method for intelligent prediction and optimization of drilling speed provided in Embodiment 1 is executed.

In some embodiments, the memory may be a high-speed random access memory (RAM: Random Access Memory), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

In other embodiments, the processor may be various types of general-purpose processors such as a central processing unit (CPU), a digital signal processor (DSP), etc., which are not limited herein.

Example 5

The method for intelligent prediction and optimization of drilling rate in this embodiment 1 may be embodied as a computer program product, and the computer program product may include a computer-readable storage medium on which a method for executing the method described in this embodiment 1 is uploaded. Computer readable program instructions for an intelligent prediction and optimization method for drilling rate.

A computer-readable storage medium may be a tangible device that retains and stores instructions for use by the instruction execution device. The computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination of the above.

It should be noted that the flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. Each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing the specified logical function(s).

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: the present invention can still be Modifications or equivalent replacements are made to the specific embodiments of the present invention, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention shall be included within the protection scope of the claims of the present invention.

The above-mentioned embodiments are only used to illustrate the present invention, and the structure, connection method and manufacturing process of each component can be changed to some extent. Any equivalent transformation and improvement based on the technical solution of the present invention should not be used. Excluded from the scope of protection of the present invention.

Claims (10)

1. a drilling speed intelligent prediction and optimization method, is characterized in that comprising the following steps: Preprocess the acquired raw drilling data to obtain a sample set; The established drilling rate prediction model is trained based on the sample set, and the trained drilling rate prediction model is obtained; Input the real-time drilling data of the block to be predicted into the ROP prediction model for prediction, and obtain the ROP. 2. The method for intelligent prediction and optimization of ROP as claimed in claim 1, wherein the method for preprocessing the acquired original drilling data to obtain a sample set comprises: Obtain the original drilling data; perform data cleaning on the original drilling data; standardize the cleaned drilling data to obtain a sample set. 3. The method for intelligent prediction and optimization of ROP as claimed in claim 2, characterized in that: the method for performing data cleaning on the original drilling data comprises: missing value processing, abnormal value processing, repeated value processing and sampling value processing deal with. 4. a kind of ROP intelligent prediction and optimization method as claimed in claim 1 is characterized in that: described based on the sample set, the established ROP prediction model is trained, and the method for obtaining the trained ROP prediction model, comprising: : Establish a drilling rate prediction model; obtain training set, validation set and test set based on the sample set, use the training set to train the drilling rate prediction model, and use the validation set and test set to verify the training results. 5. A kind of intelligent prediction and optimization method of penetration rate as claimed in claim 4, is characterized in that: described penetration rate prediction model is a three-layer BP neural network, comprising input layer, middle layer and output layer; The input layer is used for receiving drilling data in the training set; The middle layer is used for processing drilling data to obtain the optimized ROP; The output layer is used to output the ROP. 6. The intelligent prediction and optimization method for ROP as claimed in claim 4, wherein the method for training the ROP prediction model comprises: using a training set to train the ROP prediction model, and using The validation set and test set are used to test and validate the trained model until the model validation requirements are met. 7. A kind of ROP intelligent prediction and optimization method as claimed in claim 6, it is characterized in that: described model verification requirement refers to: the ROP and training set, test set and verification set that the model predicts obtains the actual mechanical The ROP correlation R ² meets the set model minimum R ² requirement. 8. An intelligent prediction and optimization system for drilling speed, characterized in that the system comprises: The sample set acquisition module is used to preprocess the acquired original drilling data to obtain a sample set; The model training module is used to train the established drilling rate prediction model based on the sample set to obtain the trained drilling rate prediction model; The ROP prediction module is used to input the real-time drilling data of the block to be predicted into the ROP prediction model for prediction to obtain the ROP. 9. A processing device, the processing device comprising at least a processor and a memory, and a computer program is stored on the memory, wherein the processor executes the computer program to implement any of claims 1 to 7 when the processor runs the computer program. One of the steps of a method for intelligent prediction and optimization of drilling speed. 10. A computer storage medium, wherein computer-readable instructions are stored thereon, and the computer-readable instructions can be executed by a processor to realize a drilling speed intelligence according to any one of claims 1 to 7 Steps in prediction and optimization methods.

Images