CN109902360A - Method, device and machine equipment for optimizing engineering parameters in drilling site operations - Google Patents

Abstract

The present disclosure discloses a method for optimizing engineering parameters in drilling site operations, including: acquiring geological data, tool data, mud data and engineering data ranges set for new well operations; In the constructed optimization space, the positions of a preset number of particles are randomly initialized; according to the position of each particle in the optimization space, the values of geological data, tool data, mud data and corresponding engineering data are used as the Build the input of the big data model, and predict the rock-breaking efficiency corresponding to the location of the particle; according to the predicted rock-breaking efficiency, adjust the optimization direction of the corresponding particle in the optimization space to determine the optimal rock-breaking efficiency. position; obtain the engineering data corresponding to the optimal rock breaking efficiency according to the determined position. This ensures the high efficiency and real-time performance of engineering data corresponding to the optimal rock breaking efficiency.

Description

Method, device and machine equipment for optimizing engineering parameters in drilling site operations

This application claims the priority of Chinese patent application CN2018109336177, which was filed on August 16, 2018 and is entitled "A GPU Parallel Fast Optimization Method for Big Data Intelligent Drilling Field Operations", the entire contents of which are hereby incorporated by reference. here.

technical field

The present disclosure relates to the field of oil exploitation, and in particular, to a method, device and machine equipment for optimizing engineering parameters in drilling site operations.

Background technique

During drilling operations, drilling conditions such as geological structure, tools, and mud directly affect the rock-breaking efficiency, and the level of rock-breaking efficiency directly affects the drilling cost and drilling time.

Specifically, during the drilling process, different drilling operating conditions correspond to different rock breaking efficiencies, such as changes in geological structure, such as selected engineering parameters. In order to ensure the rock-breaking efficiency of drilling operations and achieve high-efficiency rock-breaking, operators need to optimize the engineering parameters that can be adjusted during drilling operations to ensure high rock-breaking efficiency.

In the existing technology, the engineering parameters are generally optimized by experienced engineers according to their work experience to maximize the rock breaking efficiency. On the one hand, the optimization of engineering parameters is based on experienced engineers, which has high requirements and dependence on engineers; On the one hand, during the drilling process, the drilling conditions change in real time (eg, the drilling depth changes), so it is necessary to optimize the engineering parameters in real time according to the changes of the drilling conditions. However, optimizing engineering parameters by engineers based on work experience cannot meet the real-time requirements of optimization.

It can be seen from the above that the engineering parameter optimization based on the engineer's work experience in the prior art has low efficiency and cannot be applied to the engineering parameter optimization in the drilling process that changes in real time.

Therefore, there is an urgent need for an efficient and real-time optimization method of engineering parameters in drilling operation sites.

SUMMARY OF THE INVENTION

In order to solve the problems existing in the related art, the present disclosure provides a method, device and machine equipment for optimizing engineering parameters in drilling site operations.

In a first aspect, a method of optimizing engineering parameters in drilling site operations, comprising:

Obtain geological data, tool data, mud data and engineering data ranges for new well operations;

In the optimization space constructed according to the engineering data range, a preset number of particles are randomly initialized positions, and each point in the optimization space corresponds to an engineering data value within the engineering data range;

According to the position of each particle in the optimization space, the geological data, tool data, mud data and corresponding engineering data are taken as the input of the constructed big data model, and it is predicted that the position of the particle will be Corresponding rock breaking efficiency;

According to the predicted rock-breaking efficiency, adjust the optimization direction of the corresponding particle in the optimization space to determine the position where the optimal rock-breaking efficiency is located;

The engineering data corresponding to the optimal rock breaking efficiency is obtained according to the determined position.

In a second aspect, a device for optimizing engineering parameters in drilling site operations, the device comprising:

The data acquisition module is used to acquire the geological data, tool data, mud data of the new well operation and the range of engineering data set for the operation of the new well;

The initialization module is used to randomly initialize positions for a preset number of particles in the optimization space constructed according to the engineering data range, and each point in the optimization space corresponds to a point within the engineering data range. Engineering data value;

The prediction module is used for taking the geological data, tool data, mud data and corresponding engineering data values as the input of the constructed big data model according to the position of each particle in the optimization space, and predicting the obtained data. The rock breaking efficiency corresponding to the position of the particle;

The adjustment module is used to adjust the optimization direction of the corresponding particles in the optimization space according to the predicted rock-breaking efficiency, so as to determine the position where the optimal rock-breaking efficiency is located;

The engineering data acquisition module is used to obtain the engineering data corresponding to the optimal rock breaking efficiency according to the determined position.

In one embodiment, the apparatus further includes:

The drilling operation data acquisition module is used to acquire a number of drilling operation data collected during the drilling operation and the actual rock breaking efficiency corresponding to the drilling operation data. The drilling operation data includes geological data and tools during the drilling operation. data, mud data and engineering data;

a preprocessing module for preprocessing the drilling operation data and the corresponding actual rock breaking efficiency;

The iterative training module is configured to perform iterative training with the corresponding actual rock breaking efficiency as the target through the preprocessed drilling operation data to construct the big data model.

In one embodiment, the preprocessing module includes:

a normalization unit for normalizing the geological data, tool data, mud data and engineering data in each of the drilling operation data;

a correlation coefficient calculation unit, configured to calculate the correlation coefficients of two different types of parameters according to the normalized data of each drilling operation and the corresponding actual rock breaking efficiency;

A data removal unit, configured to remove data with low correlation with rock breaking efficiency in the drilling operation data according to the calculated correlation coefficient, and process pairs of data with high correlation in the drilling operation data to obtain data for constructing Drilling operation data of the big data model.

In one embodiment, the iterative training module includes:

a prediction unit, configured to predict the rock-breaking efficiency in the pre-built model according to the pre-processed drilling operation data to obtain the corresponding predicted rock-breaking efficiency;

A weight parameter adjustment unit, configured to adjust the weight parameters of the pre-built model according to the predicted rock-breaking efficiency and the corresponding actual rock-breaking efficiency, so that the corresponding predicted rock-breaking efficiency and the corresponding actual rock-breaking efficiency the smallest distance;

A big data model building unit, configured to use the pre-built model after adjusting the weight parameters as the big data model.

In one embodiment, the apparatus further includes:

The GPU allocation module is configured to allocate corresponding GPUs to the preset number of particles according to the number of configured GPUs, and perform prediction of rock-breaking efficiency and/or adjustment of optimization directions of each particle in parallel through the configured GPUs.

In one embodiment, the initialization module includes:

The optimization space construction unit is configured to construct the optimization space according to the range of the top pressure, the range of displacement and the range of the rotational speed of the top drive in the engineering data range, and the coordinates of each point in the optimization space indicate the The corresponding top pressure, displacement and top drive speed;

The initialization position determination unit is used for randomly arranging a preset number of particles in the optimization space to obtain the position determined by the initialization of each particle.

In one embodiment, the adjustment module includes:

an update unit, configured to update the position of the individual optimal rock-breaking efficiency of the particles and the position of the group's optimal rock-breaking efficiency according to the predicted rock-breaking efficiency;

an optimization direction adjustment unit, configured to adjust the optimization direction of the particle in the optimization space according to the individual historical optimal position and the group historical optimal position;

a moving unit, configured to adjust the position of the particle in the optimization space according to the adjusted optimization direction until the iteration end condition is reached;

The location determination unit is used to determine the location where the optimal rock-breaking efficiency of the group is located as the location where the optimal rock-breaking efficiency is located when the iteration end condition is reached.

In one embodiment, the update unit includes:

an obtaining unit for obtaining the individual optimal rock breaking efficiency and the group optimal rock breaking efficiency stored by each particle;

The individual optimal rock-breaking efficiency adjustment unit is used to adjust the position of the individual optimal rock-breaking efficiency of the particles if the predicted rock-breaking efficiency is greater than the individual optimal rock-breaking efficiency of the particles is the location of the predicted rock breaking efficiency;

The group optimal rock-breaking efficiency adjustment unit is used to adjust the group if the largest individual optimal rock-breaking efficiency of the individual optimal rock-breaking efficiencies of each particle after the update is greater than the group optimal rock-breaking efficiency. The position of the optimal rock-breaking efficiency is adjusted to the position of the largest individual optimal rock-breaking efficiency.

In a third aspect, a machine device, the device includes:

processor; and

A memory having computer readable instructions stored thereon, the computer readable instructions implementing the method as described above when executed by the processor.

The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects:

By using a preset number of particles for optimization in the optimization space constructed according to the range of engineering data, the position of the optimal rock-breaking efficiency is determined, and the engineering data corresponding to the optimal rock-breaking efficiency is obtained. Therefore, engineering data corresponding to the optimal rock breaking efficiency can be obtained in real time at the drilling site. It solves the dependence of engineering parameter optimization on engineers in engineering data in the prior art, and can optimize engineering parameters according to real-time drilling conditions, with high efficiency and real-time performance.

It is to be understood that the foregoing general description and the following detailed description are exemplary only and do not limit the present disclosure.

Description of drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description serve to explain the principles of the invention.

1 is a block diagram of a server according to an exemplary embodiment;

Fig. 2 is a flow chart of a method for optimizing engineering parameters in drilling site operations according to an exemplary embodiment;

Fig. 3 is the flow chart of the steps before step S150 of the corresponding embodiment of Fig. 2;

FIG. 4 is a flowchart of step S230 in the embodiment corresponding to FIG. 3 in an embodiment;

FIG. 5 is a flowchart of step S250 in the embodiment corresponding to FIG. 3 in an embodiment;

FIG. 6 is a flowchart of step S130 in the embodiment corresponding to FIG. 2 in an embodiment;

FIG. 7 is a flowchart of step S170 in the embodiment corresponding to FIG. 2 in an embodiment;

FIG. 8 is a flowchart of step S171 in the embodiment corresponding to FIG. 7 in an embodiment;

FIG. 9 is a block diagram of an apparatus for optimizing engineering parameters in drilling site operations according to an exemplary embodiment;

Fig. 10 is a block diagram of a machine device according to an exemplary embodiment.

By means of the above-mentioned drawings, which have shown specific embodiments of the present invention, a more detailed description will follow, these drawings and written descriptions are not intended to limit the scope of the inventive concept in any way, but by reference to specific embodiments. The concept of the present invention is explained to those skilled in the art.

Detailed ways

The description will now be made in detail of exemplary embodiments, examples of which are illustrated in the accompanying drawings. Where the following description refers to the drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the illustrative examples below are not intended to represent all implementations consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with some aspects of the invention as recited in the appended claims.

Fig. 1 is a block diagram of a server according to an exemplary embodiment. The server 200 can act as the executive body of the present disclosure to execute the technical solutions of the present disclosure.

It should be noted that the server 200 is only an example suitable for the present invention, and should not be considered as providing any limitation on the scope of use of the present invention. The server 200 should also not be interpreted as requiring or necessarily having one or more components of the exemplary server 200 shown in FIG. 2 .

The hardware structure of the server 200 may vary greatly due to different configurations or performance. As shown in FIG. 2, the server 200 includes: a power supply 210, an interface 230, at least one memory 250, and at least one central processing unit (CPU, Central Processing Units) 270.

The power supply 210 is used to provide working voltages for each hardware device on the server 200 .

The interface 230 includes at least one wired or wireless network interface 231, at least one serial-parallel conversion interface 233, at least one input/output interface 235, and at least one USB interface 237, etc., for communicating with external devices.

The memory 250 is a carrier for resource storage, which can be a read-only memory, a random access memory, a magnetic disk or an optical disk, etc. The resources stored on it include the operating system 251, the application program 253 or the data 255, etc., and the storage method can be short-term storage or permanent storage. . Wherein, the operating system 251 is used to manage and control each hardware device and application program 253 on the server 200, so as to realize the calculation and processing of the massive data 255 by the central processing unit 270, which can be WindowsServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, FreeRTOS, etc. The application program 253 is a computer program based on the operating system 251 to complete at least one specific task, which may include at least one module (not shown in FIG. 2 ), and each module may respectively include a series of computers for the server 200 Readable instructions. The data 255 may be drilling operation data collected during the drilling operation, or the like.

The central processing unit 270 may include one or more processors, and is configured to communicate with the memory 250 through a bus, so as to operate and process the mass data 255 in the memory 250 .

As described in detail above, the server 200 to which the present invention is applied will implement the method of optimizing engineering parameters in drilling site operations through the central processing unit 270 reading a series of computer readable instructions stored in the memory 250 .

In addition, the present invention can also be implemented by hardware circuits or hardware circuits combined with software instructions. Therefore, implementation of the present invention is not limited to any specific hardware circuits, software, and combinations of the two.

FIG. 2 is a flow chart of a method for optimizing engineering parameters in drilling site operations, according to an exemplary embodiment. The method for optimizing engineering parameters during drilling site operations, which may be executed by the server 200, may include the following steps.

In step S110, the geological data, tool data, mud data of the operation of the new well and the range of engineering data set for the operation of the new well are acquired.

Among them, drilling refers to the process of using certain tools to drill required holes in underground rock formations, such as drilling for oil and gas extraction.

A new well is for a well that is currently about to be drilled, and is obtained by drilling at a corresponding geographic location. Correspondingly, the mentioned range of geological data, tool data, mud data and engineering data is also for new wells. In other words, different drilling sites have corresponding geological data, tool data, mud data, and engineering data ranges. Of course, for a well that has been drilled, the engineering data is determined, that is, the engineering parameter values used in the drilling process.

As with any drilling operation, the drilling operation is performed directly through the tool on the geological formation to be drilled. In other words, the tools used for drilling, the parameters of the tools and the address structure to be operated will all affect the efficiency of drilling operations, such as the rock breaking efficiency mentioned below.

In order to ensure the efficiency of drilling, before drilling, log the position to be drilled, and obtain the geological data of the corresponding geological structure. That is, the geological data is data for describing the geological structure in which the drilling operation is to be performed. Geological data such as geophysical data such as temperature, pressure, porosity, permeability, lithology, and mineral composition.

In a specific embodiment, the geological data includes neutron porosity, gamma density, gamma spectroscopy, resistivity, and the like. And these geological data can be obtained by logging tools. Among them, the neutron porosity is the formation porosity measured by the neutron logging tool calibrated in the neutron calibration well. Gamma density is the porosity of the formation measured by the use of gamma rays and various effects of the geological medium. Gamma spectroscopy is to measure the energy spectrum of naturally occurring gamma rays in the formation through tools, such as natural gamma spectroscopy instruments, and the obtained gamma energy spectrum can be used to divide the formation and the distribution of radioactive elements in the formation in detail. various geological problems.

Tool data refers to data describing the tools used to conduct new well operations. Among them, tool data such as casing, open hole size, number of water holes, water hole area, drill bit size, etc. Tool data is selected from existing tools prior to new well operations.

Among them, casing drilling refers to the use of casing instead of drill pipe to apply torque and WOB to the drill bit, so as to realize the rotation and drilling of the drill bit. The casing referred to in the tool data refers to drilling by casing drilling. In the drilling process, in order to keep the strata in some special formation sections in the original state, and some special layers are not casing during drilling, open hole refers to the way of drilling without casing.

In the drilling process, in order to ensure the service life of the drill bit, the drill bit is generally equipped with a nozzle, that is, the drill bit nozzle, also known as the drill bit water hole, which is a hole opened at the appropriate position of the drill bit body and communicated with the flow channel of the drill bit body cavity. It constitutes a channel for drilling fluid to enter the bottom of the well from the inside of the drill pipe. The number of water holes refers to the number of water holes to be used in new well operations. The water eye area refers to the cross-sectional area of the water eye. Drill size refers to drill parameters such as drill diameter.

Drilling mud, also known as drilling fluid and mud, must be used in the drilling process to assist drilling. Mud data is the data used to describe the mud parameters used in new well operations. Mud data such as mud density, funnel viscosity, solids content, dynamic shear, etc. The mud data can be obtained by laboratory measurements of the mud to be used.

Engineering data consists of engineering parameters that can be selected and adjusted during drilling. In a specific embodiment, the engineering parameters include weight on bit, displacement and top drive speed. These engineering data can be adjusted during real-time drilling. The range of engineering data is the range for selection and adjustment set by each engineering parameter.

Step S130: In the optimization space constructed according to the engineering data range, randomly initialize positions of a preset number of particles, and each point in the optimization space corresponds to a value of engineering data within the engineering data range.

The constructed optimization space is constructed according to the parameter range set for each engineering parameter in the engineering data range. The value space of each engineering parameter in the engineering data is limited by the constructed optimization space.

In one embodiment, as shown in FIG. 6 , step S130 includes:

In step S131, an optimization space is constructed according to the range of top pressure, the range of displacement and the range of top drive rotational speed in the engineering data range, and the coordinates of each point in the search space indicate the corresponding top pressure, displacement and top drive. Rotating speed.

In step S133, a preset number of particles are randomly arranged in the optimization space, and the position determined by the initialization of each particle is obtained.

That is, the values of WOB, displacement, and top drive rotational speed are taken as the coordinate value of each point in the optimization space, so that each point in the optimization space corresponds to the corresponding WOB value, displacement value and Top drive speed value.

Wherein, what is performed is the random initialization position of the particles, that is, a preset number of particles are randomly distributed into the optimization space, so that the coordinates of the particles are the initialization positions of the particles.

The number of particles can be set according to the actual calculation accuracy, that is, the accuracy of the engineering data corresponding to the optimal rock breaking efficiency. Of course, when the hardware capability of the equipment is satisfied, the more particles the more the number of particles, the higher the calculation accuracy. will also increase accordingly. In a specific embodiment, the number of particles can be preset, so that optimization is performed in the optimization space according to the set number of particles, and engineering data corresponding to the optimal rock breaking efficiency is obtained.

Step S150: According to the position of each particle in the optimization space, the geological data, tool data, mud data and corresponding engineering data values are used as the input of the constructed big data model, and the particle size corresponding to the position of the particle is predicted. Rock efficiency.

As described above, there is corresponding engineering data for each point in the optimization space, such as WOB, displacement and top drive speed. Correspondingly, for a particle that exists at a point in the optimization space, the corresponding engineering data can be determined according to the position of the particle.

The rock breaking efficiency in the drilling process is directly related to the values of various parameters in the geological data, tool data, mud data and engineering data. Therefore, in the technical solution of the present disclosure, the prediction of rock breaking efficiency is performed according to geological data, tool data, mud data and engineering data.

The predictions made are made by means of machine learning through the constructed big data model. Before step S150, a big data model is constructed by using the drilling operation data collected in the completed drilling operation and the corresponding actual rock breaking efficiency, as described in the following for details. Through the constructed big data model, the rock breaking efficiency can be predicted according to the geological data, tool data, mud data and engineering data in the operation.

The big data model is, for example, a model constructed based on Boosting or a neural network, which is not specifically limited here.

Step S170, according to the predicted rock-breaking efficiency, adjust the optimization direction of the corresponding particle in the optimization space to determine the position where the optimal rock-breaking efficiency is located.

In the technical solution of the present disclosure, the particle swarm algorithm is used to search and determine the location of the optimal rock-breaking efficiency in the optimization space, that is, the optimal rock-breaking efficiency is determined by the movement of a preset number of particles in the optimization space. s position.

In the particle swarm optimization algorithm, in order to search and determine the position of the optimal rock breaking efficiency in the optimization space, the particles are constantly moving in the optimization space, that is, the moving speed and position of the particles in the optimization space are constantly adjusted. .

In one embodiment, as shown in FIG. 7 , step S170 includes:

Step S171, according to the predicted rock-breaking efficiency, update the position of the individual optimal rock-breaking efficiency of the particles and the position of the group's optimal rock-breaking efficiency.

The position where the individual optimal rock-breaking efficiency of the particles is located is the position corresponding to the maximum rock-breaking efficiency in the positions where the particles pass.

The location where the optimal rock-breaking efficiency of the group is located refers to the position corresponding to the maximum rock-breaking efficiency in the particle swarm composed of a preset number of particles, among the positions passed by each particle.

As before, through step S150, the rock breaking efficiency corresponding to each particle at the randomly initialized position is predicted and obtained through the big data model. Therefore, according to the predicted rock-breaking efficiency corresponding to the location of each particle, the location of the individual optimal rock-breaking efficiency of the particle and the location of the group optimal rock-breaking efficiency are adjusted.

In one embodiment, as shown in FIG. 8 , step S171 includes:

In step S310, the individual optimal rock-breaking efficiency and the group optimal rock-breaking efficiency stored by each particle are obtained.

For each particle in the optimization space, the particle stores the individual optimal rock breaking efficiency and the group optimal rock breaking efficiency. Therefore, every time the particle passes through a position, the stored individual optimal rock-breaking efficiency and group optimal rock-breaking efficiency are updated correspondingly according to the rock-breaking efficiency corresponding to the location.

Step S320, if the predicted rock-breaking efficiency is greater than the individual optimal rock-breaking efficiency of the particles, adjust the location of the individual optimal rock-breaking efficiency of the particle to the location of the predicted rock-breaking efficiency.

Step S330, if the individual optimal rock-breaking efficiency of each particle after the update is the largest individual optimal rock-breaking efficiency is greater than the group optimal rock-breaking efficiency, adjust the position of the group optimal rock-breaking efficiency to the largest individual optimal rock-breaking efficiency. The location of the rock breaking efficiency.

After randomly initializing the position for the particle, since the position of the individual optimal rock-breaking efficiency stored by the particle and the position of the group's optimal rock-breaking efficiency are empty, the current position of the particle is directly used as the location of the individual optimal rock-breaking efficiency. The position of the particle with the largest rock-breaking efficiency in the particle swarm is taken as the location of the optimal rock-breaking efficiency of the group.

In the subsequent update, after the rock-breaking efficiency is predicted according to the position of the particle, the position of the individual optimal rock-breaking efficiency of the particle and the position of the group's optimal rock-breaking efficiency are continuously updated.

Step S172, adjust the optimization direction of the particle in the optimization space according to the individual historical optimal position and the group historical optimal position.

The optimization direction refers to the moving direction of the particle in the optimization space. Of course, the goal of particles in the optimization space is where the optimal rock breaking efficiency is located. In each movement of the particle, the optimization direction is determined based on the current position of the particle and the updated position of the optimal rock-breaking efficiency of the group. In other words, the particle constructs the optimal shortest path according to the current position and the position of the updated optimal rock-breaking efficiency. Therefore, the direction from the current position of the particle to the updated optimal rock-breaking efficiency is the optimal direction.

Step S173, adjust the position of the particle in the optimization space according to the adjusted optimization direction until the iteration end condition is reached.

Among them, the optimization direction of the adjusted particle is to update the speed of the particle. In particle swarm optimization, the particle velocity update formula is:

is the velocity component of the k-th iteration particle i in the d-th dimension in the optimization space; ω is the inertia weight, non-negative; c ₁ , c ₂ are the acceleration constants; r ₁ , r ₂ are two random numbers, the value range is [0,1]; pbest _i,d is the d-th dimension component of the location where the individual optimal rock-breaking efficiency is stored by particle i; gbest _d is the d-th dimension component of the location where the group optimal rock-breaking efficiency is stored by the particle i .

The particle's position update formula is:

is the d-dimensional component of the position vector of particle i in the k-th iteration.

Therefore, the particles in the optimization space are updated with the speed and position according to the above formula. After the position is updated, the rock-breaking efficiency corresponding to the updated position is repeatedly executed, and the process of adjusting the optimization direction and updating the position of the particle is repeated. until the preset iteration end condition is reached. Among them, each speed and position update performed by the particle is regarded as an iteration.

In step S174, when the iteration end condition is reached, the position where the group optimal rock breaking efficiency is located is determined as the position where the optimal rock breaking efficiency is located.

When the iteration end condition is reached, the location of the stored optimal rock-breaking efficiency of the group is determined as the location of the optimal rock-breaking efficiency in the optimization space. The iteration end condition is such as the preset maximum number of iterations, when the number of iterations reaches the maximum number of iterations, the iteration end condition is deemed to be reached; for another example, the iteration end condition is that the historical optimal position of the group remains unchanged in N consecutive iterations , then in the continuous N iterations, the group history optimal position of the particle swarm is the same point in the optimization space, which is regarded as reaching the iteration end condition. Of course, the above is only an exemplary example of the iteration end condition, and in a specific embodiment, the iteration end condition may also be other set conditions.

Step S190, obtaining engineering data corresponding to the optimal rock breaking efficiency according to the determined position.

As described above, in the optimization space, each point corresponds to an engineering data value within the engineering data range. Therefore, the engineering data corresponding to the optimal rock breaking efficiency can be obtained according to the determined position. Therefore, it is convenient for the new well operation site to carry out the new well operation according to the obtained engineering data, for example, applying the drilling pressure according to the corresponding WOB, and carrying out the drilling operation according to the corresponding displacement and corresponding top drive rotation speed.

In drilling operations, in order to ensure the rock-breaking efficiency, it is necessary to optimize the engineering parameters in the engineering data during the drilling process. In the prior art, the engineering parameters are generally optimized by experienced engineers based on their work experience to maximize the rock-breaking efficiency. On the one hand, the optimization of engineering parameters is based on experienced engineers, which has high requirements and dependence on engineers; on the other hand, during the drilling process, the drilling conditions are changing in real time, (for example, the drilling depth is different, the geological data of the drilling also The engineering data corresponding to the optimal rock breaking efficiency may also be different), and the optimization of engineering parameters based on the engineer’s work experience has low efficiency and is not suitable for real-time changes in the drilling process. Parameter optimization. With the technical solution of the present disclosure, the dependence of engineering parameter optimization on engineers can be solved, and engineering parameter optimization can be performed according to real-time drilling conditions, with high efficiency and real-time performance.

In one embodiment, as shown in FIG. 3, before step S150, the method further includes:

In step S210, several drilling operation data collected during the drilling operation and the actual rock breaking efficiency corresponding to the drilling operation data are obtained. The drilling operation data includes geological data, tool data, mud data and engineering data during the drilling operation.

Step S230, preprocessing the drilling operation data and the corresponding actual rock breaking efficiency.

In one embodiment, as shown in FIG. 4 , step S230 includes:

Step S231, normalize the geological data, tool data, mud data and engineering data in each drilling operation data.

Among them, the normalization process performed is to calculate the average value and standard deviation of each parameter according to the value of each parameter in the collected drilling operation data. Therefore, according to the formula

Transform each parameter value, where, represents the average value of the parameter X, and σ represents the standard deviation of the parameter X. Therefore, through the above transformation, the parameter X is transformed into X', and the normalization of the parameter X is realized.

Among them, parameters such as neutron porosity, resistivity, etc. in geological data. It is worth mentioning that in the normalization process, each parameter in the geological data, tool data, mud data and engineering data in the drilling operation data is normalized according to the above formula.

In one embodiment, before step S231, the collected drilling operation data is cleaned and mapped. Wherein, the cleaning refers to removing abnormal data or erroneous data in the drilling operation data, and/or filling up the missing data, so that the constructed big data model is constructed based on complete and real data, To ensure the validity of the big data model.

The mapping performed therein refers to mapping qualitative (text) parameters in drilling operation data to quantitative (numeric) parameters. For example, the casing method mentioned above is mapped to 0, and the naked-eye method is mapped to 1, so that through the mapping, the quantification of text parameters is realized and used to build a big data model.

Step S232, according to the normalized data of each drilling operation and the corresponding actual rock breaking efficiency, the correlation coefficients of the two different types of parameters are calculated and obtained.

The calculation of the correlation coefficient, that is, the calculation of the parameters of the geological data in the drilling operation data, the parameters of the mud data, the parameters of the tool data, the parameters of the engineering data and the parameter of the rock-breaking efficiency, are relative to each other. Correlation coefficient between two different parameters.

The formula for calculating the correlation coefficient is:

where Cov(X, Y) is the covariance of parameter X and parameter Y, σ _x is the variance of parameter X, and σ _y is the variance of parameter Y, so the correlation coefficient between parameter X and parameter Y is calculated according to the above formula. The range of the calculated correlation coefficient is [0,1], 0 means no correlation, 1 means perfect linear correlation.

Therefore, the magnitude of the correlation between the two parameters is determined according to the calculated correlation coefficient.

Step S231, according to the calculated correlation coefficient, remove the data with low correlation with the rock breaking efficiency in the drilling operation data, and process the paired data with high correlation in the drilling operation data, and obtain the drilling operation data for building the big data model. .

Wherein, the judgment of high or low correlation is performed based on the calculated correlation coefficient. Specifically, corresponding thresholds are respectively set for high correlation and low correlation. For example, the threshold set for high correlation is A: that is, in When the correlation coefficient exceeds the threshold A, the correlation between the two parameters is considered to be high; for another example, the threshold set for the low correlation is B: that is, when the correlation coefficient is lower than the threshold B, the correlation between the two parameters is considered to be low.

Furthermore, the data with low correlation with the rock-breaking efficiency in the drilling operation data is removed. For pairs of data with high correlation in drilling operation data, one data in the pair of data can be represented by the relational expression of the other data, so that one of the pair of data can be removed, and according to Another data for iterative training.

Step S250 , iterative training is performed with the corresponding actual rock breaking efficiency as the target through the preprocessed drilling operation data, and a big data model is constructed and obtained.

In the iterative training, the loss function of the model is constructed through the preprocessed drilling operation data, and the big data model is constructed by adjusting the parameters of the model by solving the corresponding loss function. The loss function is a function that describes the difference between the predicted value predicted by the model and the actual value corresponding to the drilling data.

In one embodiment, as shown in FIG. 5 , step S250 includes:

Step S251 , predicting the rock-breaking efficiency in the pre-built model according to the pre-processed drilling operation data, to obtain the corresponding predicted rock-breaking efficiency.

The pre-built model may be a model constructed by boosting, neural network, or the like. The pre-built model is used to predict the rock-breaking efficiency according to each and processed drilling operation data, so as to obtain the corresponding predicted rock-breaking efficiency, that is, the predicted value mentioned above. Correspondingly, the actual rock breaking efficiency corresponding to the drilling operation data is the actual value.

Step S252: Adjust the weight parameters of the pre-built model according to the predicted predicted rock breaking efficiency and the corresponding actual rock breaking efficiency, so as to minimize the distance between the corresponding predicted rock breaking efficiency and the corresponding actual rock breaking efficiency.

Step S253, the pre-built model after adjusting the weight parameters is used as the big data model.

The distance between the predicted rock-breaking efficiency and the actual rock-breaking efficiency can be represented by Euclidean distance, etc., which is not specifically limited here.

Therefore, a big data model for predicting rock-breaking efficiency is constructed through the collected drilling operation data and the corresponding actual rock-breaking efficiency.

In one embodiment, before step S130, the method further includes:

According to the number of configured GPUs, the corresponding GPUs are allocated to the preset number of particles, and the prediction of the rock-breaking efficiency of each particle and/or the adjustment of the optimization direction is performed in parallel by the configured GPUs.

Therefore, for the particles in the optimization space, the allocated GPU opens threads to predict the rock-breaking efficiency corresponding to the particles and adjust the optimization direction, thereby realizing multi-GPU parallelism, shortening the running time, and further improving the performance of the particle. The rate at which the engineering data corresponding to the optimal rock breaking efficiency is obtained at the drilling site, which meets the real-time requirements of the drilling site. The allocation performed therein is, for example, based on the computing capability of the GPU, which is not specifically limited herein.

The following is the embodiment of the apparatus of the present disclosure, which can be used to execute the embodiment of the method for optimizing engineering parameters in drilling site operations performed by the server 200 of the present disclosure. For details not disclosed in the device embodiments of the present disclosure, please refer to the method embodiments of the present disclosure for optimizing engineering parameters during drilling site operations.

FIG. 9 is a block diagram of an apparatus for optimizing engineering parameters in drilling site operations according to an exemplary embodiment. The apparatus for optimizing engineering parameters in drilling site operations may be configured in the server 200 shown in FIG. 2 , and any of the above All or part of the steps of an illustrated method for optimizing engineering parameters during drilling site operations. As shown in FIG. 9 , the apparatus includes, but is not limited to, a data acquisition module 110 , an initialization module 130 , a prediction module 150 , an adjustment module 170 and an engineering data acquisition module 190 . in,

The data acquisition module 110 is used for acquiring geological data, tool data, mud data of the new well operation and the range of engineering data set for the new well operation.

The initialization module 130, which is connected to the data acquisition module 110, is used to randomly initialize positions for a preset number of particles in the optimization space constructed according to the engineering data range, and each point in the optimization space corresponds to engineering data A project data value within the range.

The prediction module 150, which is connected with the initialization module 130, is used for taking geological data, tool data, mud data and corresponding engineering data values as the values of the constructed big data model according to the position of each particle in the optimization space. Input, predict the rock breaking efficiency corresponding to the particle position.

The adjustment module 170, which is connected to the prediction module 150, is used for adjusting the optimization direction of the corresponding particle in the optimization space according to the predicted rock-breaking efficiency, so as to determine the position where the optimal rock-breaking efficiency is located.

The engineering data acquisition module 190, which is connected to the adjustment module 170, is used for obtaining engineering data corresponding to the optimal rock breaking efficiency according to the determined position.

For details of the realization process of the functions and functions of each module in the above-mentioned device, please refer to the realization process of the corresponding steps in the above-mentioned method for optimizing engineering parameters during drilling site operations, and will not be repeated here.

It can be understood that these modules can be implemented by hardware, software, or a combination of the two. When implemented in hardware, these modules may be implemented as one or more hardware modules, such as one or more application specific integrated circuits. When implemented in software, these modules may be implemented as one or more computer programs executed on one or more processors, such as programs stored in memory 250 executed by central processing unit 270 of FIG. 2 .

In one embodiment, the device for optimizing engineering parameters during drilling site operations further includes:

The drilling operation data acquisition module is used to obtain several drilling operation data collected during the drilling operation and the actual rock breaking efficiency corresponding to the drilling operation data. The drilling operation data includes the geological data, tool data and mud data during the drilling operation. and engineering data.

The preprocessing module is used to preprocess the drilling operation data and the corresponding actual rock breaking efficiency.

The iterative training module is used to perform iterative training with the corresponding actual rock breaking efficiency as the target through the preprocessed drilling operation data to construct a big data model.

In one embodiment, the preprocessing module includes:

The normalization unit is used for normalizing the geological data, tool data, mud data and engineering data in each drilling operation data.

The correlation coefficient calculation unit is used for calculating the correlation coefficients of two different types of parameters according to the normalized data of each drilling operation and the corresponding actual rock breaking efficiency.

The data removal unit is used to remove the data with low correlation with the rock breaking efficiency in the drilling operation data according to the calculated correlation coefficient, and process the paired data with high correlation in the drilling operation data, and obtain the data used for building the big data model. Drilling operation data.

In one embodiment, the iterative training module includes:

The prediction unit is used to predict the rock-breaking efficiency in the pre-built model according to the pre-processed drilling operation data, and obtain the corresponding predicted rock-breaking efficiency.

The weight parameter adjustment unit is used to adjust the weight parameters of the pre-built model according to the predicted predicted rock breaking efficiency and the corresponding actual rock breaking efficiency, so as to minimize the distance between the corresponding predicted rock breaking efficiency and the corresponding actual rock breaking efficiency.

The big data model building unit is used to use the pre-built model after adjusting the weight parameters as the big data model.

In one embodiment, the device for optimizing engineering parameters during drilling site operations further includes:

The GPU allocation module is configured to allocate a corresponding GPU to a preset number of particles according to the number of configured GPUs, and perform prediction of rock-breaking efficiency of each particle and/or adjustment of the optimization direction in parallel through the configured GPU.

In one embodiment, the initialization module includes:

The optimization space construction unit is used to construct the optimization space according to the range of the top pressure, the range of displacement and the range of the top drive speed in the engineering data range. The coordinates of each point in the optimization space indicate the corresponding top pressure, displacement and top drive speed.

The initialization position determination unit is used for randomly arranging a preset number of particles in the optimization space to obtain the position determined by the initialization of each particle.

In one embodiment, the adjustment module includes:

The updating unit is used to update the location of the individual optimal rock-breaking efficiency of the particles and the location of the optimal rock-breaking efficiency of the group according to the predicted rock-breaking efficiency.

The optimization direction adjustment unit is used to adjust the optimization direction of the particles in the optimization space according to the historical optimal position of the individual and the historical optimal position of the group.

The moving unit is used to adjust the position of the particle in the optimization space according to the adjusted optimization direction until the iteration end condition is reached.

The location determination unit is used to determine the location where the optimal rock-breaking efficiency of the group is located as the location where the optimal rock-breaking efficiency is located when the iteration end condition is reached.

In one embodiment, the update unit includes:

The obtaining unit is used to obtain the individual optimal rock-breaking efficiency and the group optimal rock-breaking efficiency stored by each particle.

The individual optimal rock-breaking efficiency adjustment unit is used to adjust the position of the particle's individual optimal rock-breaking efficiency to the predicted rock-breaking efficiency if the predicted rock-breaking efficiency is greater than the particle's individual optimal rock-breaking efficiency. location.

The group optimal rock-breaking efficiency adjustment unit is used to adjust the location of the group optimal rock-breaking efficiency if the largest individual optimal rock-breaking efficiency of each particle after the update is greater than the group optimal rock-breaking efficiency The position is adjusted to the position where the maximum individual optimal rock breaking efficiency is located.

For details of the realization process of the functions and functions of each module in the above-mentioned device, please refer to the realization process of the corresponding steps in the above-mentioned method for optimizing engineering parameters during drilling site operations, and will not be repeated here.

Optionally, the present disclosure further provides a machine device, which can be used to perform all or part of the steps of the method for optimizing engineering parameters in drilling site operations shown in any of the above method embodiments. As shown in Figure 10, the machine equipment includes:

processor 1001; and

The memory 1002 has computer-readable instructions stored on the memory 1002, and when the computer-readable instructions are executed by the processor 1001, any one of the above method implementations is implemented.

The method in any of the above embodiments is implemented when the executable instruction is executed by the processor 1001 . The executable instructions are, for example, computer-readable instructions. When the processor 1001 executes the instructions, the processor reads the computer-readable instructions stored in the memory through a communication line/bus 1003 connected to the memory.

The specific manner in which the processor in this embodiment performs operations has been described in detail in the embodiment related to the method for optimizing engineering parameters in drilling site operations, and will not be described in detail here.

In an exemplary embodiment, a computer-readable storage medium is also provided, on which a computer program is stored, and when the computer program is executed by a processor, implements the method in any of the above method embodiments. The computer-readable storage medium includes, for example, the memory 250 of the computer program, and the above-mentioned instructions can be executed by the central processing unit 270 of the server 200 to implement the above-mentioned method.

The specific manner in which the processor in this embodiment performs operations has been described in detail in the embodiment related to the method for optimizing engineering parameters in drilling site operations, and will not be described in detail here.

It should be understood that the present invention is not limited to the precise structures described above and illustrated in the accompanying drawings and that various modifications and changes may be made without departing from its scope. The scope of the present invention is limited only by the appended claims.

Claims (10)

1. A method of optimizing engineering parameters in a drilling site operation, comprising:

acquiring geological data, tool data and slurry data of new well operation and an engineering data range set for performing the new well operation;

in an optimization space constructed according to the engineering data range, randomly initializing positions for a preset number of particles, wherein each point in the optimization space corresponds to an engineering data value in the engineering data range;

according to the position of each particle in the optimizing space, the geological data, the tool data, the mud data and the corresponding engineering data value are used as the input of a constructed big data model, and the rock breaking efficiency corresponding to the position of the particle is obtained through prediction;

according to the predicted rock breaking efficiency, adjusting the optimizing direction of the corresponding particles in the optimizing space to determine the position of the optimal rock breaking efficiency;

and acquiring engineering data corresponding to the optimal rock breaking efficiency according to the determined position.

2. The method according to claim 1, wherein before the geological data, tool data, mud data and corresponding engineering data values are used as input parameters of the constructed big data model according to the position of each particle in the optimization space, and the rock breaking efficiency corresponding to the position of the particle is predicted, the method further comprises:

acquiring a plurality of drilling operation data acquired in drilling operation and actual rock breaking efficiency corresponding to the drilling operation data, wherein the drilling operation data comprises geological data, tool data, mud data and engineering data in the drilling operation process;

preprocessing the drilling operation data and the corresponding actual rock breaking efficiency;

and performing iterative training by taking the corresponding actual rock breaking efficiency as a target through the preprocessed drilling operation data to construct and obtain the big data model.

3. The method of claim 2, wherein the pre-processing the drilling operation data and the corresponding actual rock breaking efficiency comprises:

performing normalization processing on geological data, tool data, mud data and engineering data in each drilling operation data;

calculating to obtain correlation coefficients of two different types of parameters according to each drilling operation data after normalization processing and the corresponding actual rock breaking efficiency;

and removing data with low correlation with the rock breaking efficiency from the drilling operation data according to the calculated correlation coefficient, and processing pairwise data with high correlation from the drilling operation data to obtain the drilling operation data for constructing the big data model.

4. The method of claim 2, wherein the drilling operation data after preprocessing is subjected to iterative training with the corresponding actual rock breaking efficiency as a target, and the building of the big data model comprises:

predicting rock breaking efficiency in a pre-constructed model according to the preprocessed drilling operation data to obtain corresponding predicted rock breaking efficiency;

adjusting the weight parameters of the pre-constructed model according to the predicted rock breaking efficiency and the corresponding actual rock breaking efficiency so as to minimize the distance between the corresponding predicted rock breaking efficiency and the corresponding actual rock breaking efficiency;

and taking the pre-constructed model after the weight parameters are adjusted as the big data model.

5. The method of claim 1, wherein the randomly initializing positions for a predetermined number of particles in a search space constructed from the engineering data range, each point in the search space corresponding to a value of engineering data in the engineering data range, further comprises:

and distributing corresponding GPUs for the preset number of particles according to the number of the configured GPUs, and performing prediction and/or adjustment of optimizing directions of the rock breaking efficiency of the particles in parallel through the configured GPUs.

6. The method of claim 1, wherein the randomly initializing positions for a predetermined number of particles in the optimization space constructed from the engineering data range, each point in the optimization space corresponding to an engineering data value in the engineering data range comprises:

constructing and obtaining the optimization space according to the range of top pressure, the range of displacement and the range of top drive rotating speed in the engineering data range, wherein the coordinates of each point in the optimization space indicate the corresponding top pressure, displacement and top drive rotating speed;

randomly arranging a preset number of particles in the optimizing space, and obtaining the position determined by initialization of each particle.

7. The method of claim 1, wherein the adjusting the optimizing direction of the corresponding particle in the optimizing space according to the predicted rock breaking efficiency to determine the position of the optimal rock breaking efficiency comprises:

updating the position of the individual optimal rock breaking efficiency of the particles and the position of the group optimal rock breaking efficiency according to the predicted rock breaking efficiency;

adjusting the optimizing direction of the particles in the optimizing space according to the individual historical optimal position and the group historical optimal position;

adjusting the position of the particle in the optimizing space according to the adjusted optimizing direction until an iteration end condition is reached;

and when the iteration ending condition is reached, determining the position of the group optimal rock breaking efficiency as the position of the optimal rock breaking efficiency.

8. The method of claim 7, wherein updating the location of the individual optimal rock-breaking efficiency of the particles and updating the location of the population optimal rock-breaking efficiency according to the predicted rock-breaking efficiency comprises:

obtaining the individual optimal rock breaking efficiency and the group optimal rock breaking efficiency stored in each particle;

if the predicted rock breaking efficiency is greater than the individual optimal rock breaking efficiency of the particles, adjusting the position of the individual optimal rock breaking efficiency of the particles to the position of the predicted rock breaking efficiency;

if the maximum individual optimal rock breaking efficiency in the individual optimal rock breaking efficiencies of each particle after updating is larger than the group optimal rock breaking efficiency, adjusting the position of the group optimal rock breaking efficiency to the position of the maximum individual optimal rock breaking efficiency.

9. An apparatus for optimizing engineering parameters in a drilling site operation, the apparatus comprising:

the data acquisition module is used for acquiring geological data, tool data and mud data of new well operation and an engineering data range set for the new well operation;

the initialization module is used for randomly initializing positions for a preset number of particles in an optimization space constructed according to the engineering data range, wherein each point in the optimization space corresponds to an engineering data value in the engineering data range;

the prediction module is used for taking the geological data, the tool data, the mud data and the corresponding engineering data values as the input of the constructed big data model according to the position of each particle in the optimizing space, and predicting the rock breaking efficiency corresponding to the position of the particle;

the adjusting module is used for adjusting the optimizing direction of the corresponding particles in the optimizing space according to the predicted rock breaking efficiency so as to determine the position of the optimal rock breaking efficiency;

and the engineering data acquisition module is used for acquiring engineering data corresponding to the optimal rock breaking efficiency according to the determined position.

10. A machine device, characterized in that the device comprises:

a processor; and

a memory having computer readable instructions stored thereon which, when executed by the processor, implement the method of any of claims 1 to 8.