CN115855872A - Fluid type identification method and device based on downhole spectral measurement and sampling instrument - Google Patents
Fluid type identification method and device based on downhole spectral measurement and sampling instrument Download PDFInfo
- Publication number
- CN115855872A CN115855872A CN202211634147.7A CN202211634147A CN115855872A CN 115855872 A CN115855872 A CN 115855872A CN 202211634147 A CN202211634147 A CN 202211634147A CN 115855872 A CN115855872 A CN 115855872A
- Authority
- CN
- China
- Prior art keywords
- fluid
- data
- fluid type
- spectral
- type identification
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a fluid type identification method and device based on downhole spectral measurement and a sampling instrument. The method comprises the following steps: acquiring fluid sample training data, wherein the fluid sample training data comprises: fluid spectrum data and a labeled fluid type label corresponding to the fluid spectrum data; performing model training on the neural network according to the fluid sample training data to obtain a fluid type identification model; acquiring fluid spectral data of a fluid to be identified, which is measured by a downhole spectral measuring device; and inputting the fluid spectrum data into a fluid type identification model to identify the type of the downhole fluid, so as to obtain the fluid type corresponding to the fluid to be identified. The scheme provided by the invention can accurately identify the type of the fluid, and lays a solid foundation for further eliminating the influence of water components on oil and gas spectrums in the spectrum composition analysis and more accurately analyzing the oil and gas compositions, thereby reducing the operation risk and saving the operation time and cost.
Description
Technical Field
The invention relates to the technical field of exploration, in particular to a fluid type identification method and device based on downhole spectral measurement and a sampling instrument.
Background
Oil and gas reservoirs are not only complex in construction but also contain fluids of very different nature. Reservoir fluids can be generally classified as dry gas, wet gas, lean condensate, rich condensate, critical fluids, volatile oils, black oil, heavy oil, and oil sands (bitumen). During oil and gas production, changes in temperature and pressure may cause complex (gas, liquid and solid) phase changes (i.e., compositional changes) in the fluid. Especially the formation of solid matter (e.g. precipitation and precipitation of hydrates, paraffins, asphaltenes, inorganic salts, etc.) can lead to plugging of pipelines and production facilities, seriously affecting oil and gas production. Thus, reservoir undisturbed fluid properties are critical to overall field exploration, development and production. However, at present, the type of the fluid is judged only by the experience of a formation tester operation engineer and an interpretation engineer in China, the method is simple and rough in qualitative, the identification accuracy is low, the required time is long, and the field problem cannot be solved essentially.
Disclosure of Invention
In view of the above, the present invention has been developed to provide a fluid type identification method, apparatus and sampling tool based on downhole spectral measurements that overcome or at least partially address the above-mentioned problems.
According to one aspect of the invention, a fluid type identification method based on downhole spectral measurement is provided, comprising:
acquiring fluid sample training data, wherein the fluid sample training data comprises: fluid spectrum data and a labeled fluid type label corresponding to the fluid spectrum data;
performing model training on the neural network according to the fluid sample training data to obtain a fluid type recognition model;
acquiring fluid spectral data of a fluid to be identified, which is measured by a downhole spectral measuring device;
and inputting the fluid spectrum data into a fluid type identification model to identify the type of the downhole fluid to obtain the fluid type corresponding to the fluid to be identified.
According to another aspect of the present invention, there is provided a downhole spectral measurement-based type identification apparatus comprising:
a first acquisition module adapted to acquire fluid sample training data, wherein the fluid sample training data comprises: fluid spectrum data and a labeled fluid type label corresponding to the fluid spectrum data;
the model training module is suitable for performing model training on the neural network according to the fluid sample training data to obtain a fluid type identification model;
the second acquisition module is suitable for acquiring fluid spectral data of the fluid to be identified, which is measured by the downhole spectral measurement device;
and the identification module is suitable for inputting the fluid spectrum data into the fluid type identification model to carry out downhole fluid type identification so as to obtain the fluid type corresponding to the fluid to be identified.
According to another aspect of the present invention, there is provided a sampling apparatus comprising: the fluid type identification device based on the downhole spectral measurement and the downhole spectral measurement device are provided.
According to yet another aspect of the present invention, there is provided a computing device comprising: the processor, the memory and the communication interface complete mutual communication through the communication bus;
the memory is used for storing at least one executable instruction, and the executable instruction enables the processor to execute the operation corresponding to the fluid type identification method based on the downhole spectral measurement.
According to yet another aspect of the present invention, a computer storage medium is provided, wherein at least one executable instruction is stored in the storage medium, and the executable instruction causes a processor to execute the operation corresponding to the fluid type identification method based on downhole spectral measurement.
The scheme provided by the invention can accurately identify the type of the fluid, for example, identify that the fluid belongs to mud (emulsified fluid or invalid measurement), gas, oil, water, oil-water mixture and gas-water mixture, and lay a solid foundation for further eliminating the influence of water components on oil and gas spectrums in the spectrum composition analysis and more accurately analyzing the oil and gas compositions, thereby reducing the operation risk and saving the operation time and cost.
The foregoing description is only an overview of the technical solutions of the present invention, and the embodiments of the present invention are described below in order to make the technical means of the present invention more clearly understood and to make the above and other objects, features, and advantages of the present invention more clearly understandable.
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 only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
FIG. 1A shows a schematic flow diagram of a fluid type identification method based on downhole spectral measurements according to one embodiment of the present invention;
FIG. 1B is a schematic diagram of a downhole spectroscopy measurement principle;
FIG. 1C is a near infrared spectrum of various fluids;
FIG. 1D is a schematic diagram of a neural network;
FIG. 1E is a schematic view of a cable formation pressure measurement sampler (EFDT);
FIG. 1F is a schematic diagram of a sample while drilling Instrument (IFSA);
FIG. 2 shows a schematic structural diagram of a fluid type identification device based on downhole spectral measurements according to one embodiment of the present invention;
FIG. 3 shows a schematic structural diagram of a computing device according to one embodiment of the invention.
Detailed Description
Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.
FIG. 1A shows a schematic flow diagram of a fluid type identification method based on downhole spectral measurements according to one embodiment of the invention. As shown in fig. 1A, the method comprises the steps of:
step S101, obtaining fluid sample training data, wherein the fluid sample training data includes: and the fluid spectrum data and the labeled fluid type label corresponding to the fluid spectrum data.
Specifically, in the laboratory and under the high-temperature and high-pressure conditions in the well, the spectrograms of various fluids in each spectral channel are measured, the present embodiment takes 256 spectral channels as an example, and the spectrograms of various fluids in 256 spectral channels are measured, it should be noted that the present embodiment is not limited to 256 spectral channels, wherein the spectrograms are graphs formed by the change of absorbance along with the change of the spectral channels.
The types of fluids can be roughly divided into: gas, oil, water, gas-water mixture, oil-water mixture, emulsion (mud, null measurement); it can be further subdivided into: dry gas, wet gas, lean condensate, rich condensate, near critical oil, volatile oil, black oil, heavy oil, water, dry gas-water mixture, wet gas-water mixture, lean condensate-water mixture, rich condensate-water mixture, near critical oil-water mixture, volatile oil-water mixture, black oil-water mixture, heavy oil-water mixture, emulsion (slurry), etc. When the fluid type recognition model is trained, a proper fluid type can be selected for training according to actual needs.
Manually labeling different fluids to obtain labeled fluid type labels, so as to obtain fluid sample training data, wherein the fluid sample training data comprises: and the fluid spectrum data and the labeled fluid type label corresponding to the fluid spectrum data. The fluid spectral data includes: and the spectrum channel and the fluid absorbance data corresponding to the spectrum channel. The absorbance (Optical sensitivity) expresses the absorption capacity of a substance for light.
Wherein, the fluid spectrum data can be further measured by the following method: the underground spectral measurement device measures the incident light intensity corresponding to each spectral channel before the light beam generated by the light source passes through the fluid and the transmission light intensity corresponding to each spectral channel after the light beam passes through the fluid; and calculating fluid absorbance data corresponding to each spectral channel according to the incident light intensity and the transmitted light intensity.
The principle of downhole spectroscopy is shown in FIG. 1B. First, the light source generates light intensity I 0 The near infrared beam of light, passes through a first sapphire window on the fluid line (sapphire does not absorb any light) and then through a few millimeters thick fluid to be identified. Where part of the light beam is absorbed by the fluid, some of the light beam may be scattered (depending on the size of the fluid particles and the wavelength of the light beam). The remaining beam then exits the second sapphire window and passes through a spectral detector to measure the intensity I of the transmitted light. Wherein different fluids have different spectral characteristics (characteristic peaks), i.e. do notThe absorbance at the same wavelength is different, so that the fluid type can be identified through the fluid spectrum data.
Typically, the absorbance (Optical sensitivity) is used to express the absorption capacity of a fluid for light. The absorbance OD (λ) at a wavelength λ is defined as; intensity of transmitted light I (lambda) and intensity of incident light I 0 Negative logarithm of the ratio of (λ):
if all light is transmitted through the fluid, the light transmission is 100% and the absorbance is 0. If 1/10 of the light is transmitted through the fluid (i.e., 90% of the light is absorbed by the fluid), the absorbance is 1. Likewise, if 1/100 of the light is transmitted through the fluid (99% of the light is absorbed by the fluid), the OD is 2. Most downhole fluid analyzers measure absorbance in the range of 0 to 5.
FIG. 1C shows near infrared absorption spectra of water, diesel, methane, carbon dioxide, volatile oil, black oil, and heavy oil measured in the laboratory.
As can be seen in FIG. 1C, two of the water peaks (approximately between the 80-100 and 190-256 spectral channels) reach saturation, and thus the absorbance appears as a spike. The hydrocarbon peaks are between 130-180 spectral channels. The carbon dioxide peak is concentrated in the 215-248 spectral channel interval. Water and hydrocarbons are easily distinguished (commonly referred to as oil-water or gas-water fractions) due to the large difference between the water peak and the hydrocarbon peak. It can also be used to distinguish mobile phases between water and hydrocarbons. For example, with proper placement of the spectroscopy module, water and hydrocarbon slugs can be clearly identified. In addition, the hydrocarbon peak and the carbon dioxide peak are also greatly different, and thus can be easily distinguished. However, the carbon dioxide peak overlaps with the water peak, and the presence of water can seriously affect the determination of the composition of carbon dioxide, especially when the water-based drilling fluid is used for sampling and analyzing oil and gas.
Therefore, in order to be able to accurately identify the fluid type, model training is performed here using a neural network.
In an optional embodiment of the invention, before model training the neural network according to the fluid sample training data, the method further comprises:
preprocessing the fluid sample training data to obtain preprocessed fluid sample training data, and performing dimensionality reduction processing on the preprocessed fluid sample training data; and performing model training on the neural network according to the fluid sample training data subjected to the dimension reduction treatment to obtain a fluid type identification model.
In particular, preprocessing the fluid sample training data may specifically be interpolating and reasonably supplementing infinite, invalid points in the fluid sample training data, for example, a median average method may be used for supplementing. Then, a normalization (or "centering") process is performed. After the preprocessing or the normalization process is performed on the fluid sample training data, the preprocessed fluid sample training data is subjected to a dimensionality reduction process, for example, a Principal Component Analysis (PCA) Principal Component Analysis method (Principal Component Analysis) may be used for dimensionality reduction, and the PCA calculation may use a nonlinear iterative partial least squares (NIPALS) algorithm. The dimension reduction treatment process comprises the following steps:
there are m (n =256 spectral channels) dimensional fluid samples X = (X) 1 ,x 2 ,…,x m ) Here, the original data is reduced to k dimensions, and then the reduced data is used to complete model training.
1. The original data is firstly grouped into n rows and m columns of matrix X according to columns, and then the mean value of each dimension in X is subtracted to obtain X'. Wherein: m = number of samples, n = number of dimensions. (all fluid samples are "centered", i.e., centered
Then, forming the data into n rows and m columns of matrix X' equivalent by columns)
2. Computing a covariance matrix for a fluid sample
3. The eigenvalues of the covariance matrix and the corresponding eigenvectors are solved (eigenvalue decomposition or singular value decomposition).
4. And arranging the eigenvectors into a matrix from top to bottom according to the size of the corresponding eigenvalue, and taking the first k rows to form the matrix P. For example, the top 10 maximum eigenvalues are taken, which are only illustrative and not limiting.
And 5.Y = PX is the data after dimensionality reduction to the k dimension.
And S102, performing model training on the neural network according to the fluid sample training data to obtain a fluid type identification model.
After the fluid sample training data is acquired, model training can be performed on the neural network according to the fluid sample training data, a fluid type identification model can be obtained through the model training, the fluid type identification model has the input of fluid spectrum data and the output of fluid type.
Specifically, the following method can be used for model training: performing model training on the neural network according to the fluid spectrum data to obtain a fluid type result, wherein the fluid type result is an estimation result output by the neural network; calculating the loss between the fluid type result and the labeled fluid type label to obtain a model loss function, and updating the model parameters of the neural network according to the model loss function; and (3) iteratively executing the steps until the iteration times reach preset iteration times, and/or the output value of the model loss function is smaller than a preset threshold value, so as to obtain the fluid type identification model. The preset iteration number and the preset threshold value may be flexibly set according to actual experience, for example, the preset iteration number may be set to 10000.
Neural networks are mainly composed of artificial neurons (Neuron), which are the most basic units constituting neural networks, and are similar to biological neurons. The artificial neuron consists of five parts; (1) inputting: x is the number of 1 ，x 2 ，…，x m Are the m input variables of the artificial neuron. (2) weight and threshold parameters: w is a 1 ，w 2 ，…，w m M network weights of the artificial neuron reflecting the connection strength of the input variable to the neural network, b is a threshold of the artificial neuron, b enables transmissionThe function moves left and right, which is beneficial to the capability of the neural network. (3) linear combination: linearly combining the input values with weights and thresholds, z = ∑ x i w i + b. (4) transfer function: the transfer function is also called action function, transfer function, excitation function, etc. and is used to perform function operation on z to obtain the output of artificial neuron. Commonly used transfer functions are: threshold functions, linear functions, logarithmic Sigmoid functions, tangent Sigmoid functions, and the like. (5) outputting: get the output of the artificial neuron, o = f (Σ x) i w i +b)。
With the artificial neurons in place, the neural networks can be connected, as shown in FIG. 1D. The neural network consists of three layers: an input layer, a hidden layer, and an output layer. The data X is input from an input layer, is subjected to preprocessing (such as denoising, smoothing, standardization, normalization and the like), then is subjected to dimensionality reduction by using a PCA dimensionality reduction method to obtain a matrix T, and is input into the hidden layer together with the weight and the threshold value. The hidden layer may be a single layer or multiple layers, with one or more artificial neurons in each layer. And the hidden layer is transmitted to the output layer after the transfer function of the weight and the threshold is operated. The output layer gives the predicted (estimated) value y of the neural network and the label (expected) value
The Loss (Loss) or objective function is calculated by comparison, and the objective function can be minimized by gradient descent, M-L optimization or quasi-newton method to obtain the values of parameters w and b in the model, thereby obtaining y = f (X, w, b). For classification problems, the output layer can use the Softmax function to obtain the probabilities of various classes. And obtaining a fluid type recognition model through training.
And step S103, acquiring fluid spectrum data of the fluid to be identified, which is measured by the downhole spectrum measuring device.
In performing downhole operations, when there is a need for fluid type identification, fluid spectral data of a fluid to be identified may be measured by a downhole spectral measurement device, wherein the measurement of the fluid spectral data of the fluid to be identified by the downhole spectral measurement device may further be achieved by: the underground spectral measurement device measures the incident light intensity corresponding to each spectral channel before the light beam generated by the light source passes through the fluid to be identified and the transmitted light intensity corresponding to each spectral channel after the light beam passes through the fluid to be identified; and calculating fluid absorbance data corresponding to each spectral channel according to the incident light intensity and the transmitted light intensity. For specific implementation, reference may be made to a corresponding part in step S101, which is not described herein again.
And step S104, inputting the fluid spectrum data into the fluid type identification model for carrying out downhole fluid type identification to obtain the fluid type corresponding to the fluid to be identified.
After fluid spectrum data of the fluid to be identified are acquired, the fluid spectrum data are input into a fluid type identification model for fluid type identification, the identification process is carried out underground due to the fact that the transmission data rate is limited, and the fluid type identification model outputs the fluid type corresponding to the fluid to be identified through operation processing.
In the embodiment, the fluid type identification model is embedded into an EFDT or IFSA formation testing sampling instrument as a fluid type identification method for formation testing. The pumping parameters and actual fluid spectral data at multiple times may be collected by a wireline formation pressure sampler (EFDT) or a sample while drilling (IFSA) instrument.
An implementation manner of acquiring pumping parameters and actual fluid spectrum data values through an electrical cable formation pressure measurement sampler (EFDT) is shown in fig. 1E, and includes:
before logging, a cable formation pressure measurement sampling instrument (EFDT) is placed at a target depth in a well, then a probe is seated on a well wall, after the seating is successful, a piston pump is started, formation fluid enters a pipeline through a suction port, a spectrum signal value (fluid spectrum data measured in real time) is measured through an underground spectrum measuring device in the piston pump, identification processing is carried out underground through a fluid type identification model in the embodiment (operation must be carried out underground due to the fact that the transmission data rate is limited, and the result is obtained), and the fluid type is uploaded to a ground logging system in real time through remote transmission of a cable. It should be noted that fig. 1E is a schematic illustration of a cable formation pressure sampler, with some components not shown.
The method for acquiring the pumping parameters and the actual fluid spectrum data value by the sampling while drilling Instrument (IFSA) is shown in fig. 1F, and comprises the following steps:
before logging, a sampling while drilling Instrument (IFSA) is placed to a target depth underground, a ground system is communicated with the underground slurry transmission device through the slurry transmission device, and the underground slurry transmission device issues a ground command to the sampling while drilling instrument. Then, the sampling while drilling instrument seats the probe on the well wall, after the seating is successful, a piston pump is started, formation fluid enters a pipeline through a suction port, a spectral data signal value (fluid spectral data measured in real time) is measured through a downhole spectral measurement device in the sampling while drilling instrument, identification processing is performed underground through the fluid type identification model in the embodiment (operation must be performed underground due to the fact that the transmission data rate is limited, and the result is obtained), and then the calculated fluid type is uploaded to a ground logging system in real time through a mud transmission device. It should be noted that fig. 1F is a schematic illustration of the sample while drilling instrument, and some components are not shown.
The following describes, by way of different examples, how a fluid type can be accurately identified by the fluid type identification method provided in the present embodiment:
gas sample (filtrate oil-formation gas sample mixture, which are completely miscible and mixed into a single phase gas)
Example one is fluid type identification during a gas sampling process under offshore oil-based mud. When the probe is pushed against the borehole wall and pumping is initiated, the spectrum is measured as a slurry with many micro-sized solid particles, scattering the spectrum, and the measured spectrum oscillates back and forth up and down, hence the name null spectrum (Invalid). Then the pump out is a gas. Finally, the spectral measurement is again mud when the pump is stopped. The fluid type identification method can accurately predict the whole pumping sampling process.
Oil sample (filtrate oil-formation oil sample mixture, which are completely miscible, mixed into a single phase oil)
Example two is fluid type identification during a certain oil sampling process under offshore oil-based mud. The spectra were measured as mud, invalid spectra (Invalid) as the probe was pushed against the borehole wall and pumping was initiated. And then pumped out as oil. Finally, the mud is measured again by the optical spectrum when the pump is stopped. The fluid type identification method can accurately predict the whole pumping sampling process.
Water sample (filtrate water-stratum water sample, they are fully miscible, mixed into single phase water)
Example three is fluid type identification in a certain water sampling process under the offshore water-based mud. Initially, the water sample from the previous sampling point was measured. The probe is then pushed against the borehole wall and the slurry is measured spectrally as Invalid spectra (Invalid) as soon as pumping is initiated. Secondly, water is pumped out. Finally, the spectral measurement is again mud when the pump is stopped. The fluid type identification method can accurately predict the whole pumping sampling process.
Filtrate water-oil sample (water-oil immiscible, forming two phase slug flow)
Example four is fluid type identification during sampling of a sample of oil under an offshore water-based mud. Initially, the water sample from the previous sampling point was measured. The probe is then pushed against the borehole wall and the slurry is measured spectrally as Invalid spectra (Invalid) as soon as pumping is initiated. Then, a filtrate water-oil mixture is pumped followed by a filtrate water-oil slug. Finally, the mud is measured again by the optical spectrum when the pump is stopped. The fluid type identification method can accurately predict the whole pumping sampling process.
In an optional embodiment of the present invention, after obtaining the fluid type corresponding to the fluid to be identified, the method further comprises: and uploading the fluid type corresponding to the fluid to be identified to a surface logging system, so that the surface logging system can perform logging operation according to the fluid type, for example, the temperature, the pressure and the like of production can be adjusted according to the fluid type.
Of course, other models may be used for model training in this embodiment, for example, 1) Complex Tree; 2) Medium Tree; 3) Simple Tree; 4) Linear Discriminant; 5) A Quadratic Discriminant; 6) A Linear support vector machine (Linear SVM); 7) A Quadratic support vector machine (Quadratic SVM); 8) A Cubic support vector machine (Cubic SVM); 9) A Fine Gaussian support vector machine (Fine Gaussian SVM); 10 Medium Gaussian support vector machine Medium Gaussian SVM;11 Coarse Gaussian support vector machine Coarse Gaussian SVM;12 Fine K nearest neighbor method Fine KNN;13 Medium K nearest neighbor Medium KNN;14 Coarse K nearest neighbor Coarse KNN;15 Cosine K nearest neighbor Cosine KNN;16 Cubic K nearest neighbor Cubic KNN;17 Weight K nearest neighbor KNN;18 Set-enhanced tree Ensemble-Boosted Trees;19 ) the Ensemble-bag tree, ensemble-Bagged Trees;20 set-Subspace Discriminant Ensemble-Subspace discriminatant; 21 set-Subspace K nearest neighbor-Subspace KNN;22 set-RUS enhanced tree-RUS enhanced Trees.
Table 1 shows the recognition results of 23 fluid type recognition models based on downhole spectral measurement, and it can be seen from table 1 that the accuracy of the neural network is the highest, so that the fluid type prediction is performed by using the fluid type recognition model obtained by training the neural network.
TABLE 1
The scheme provided by the invention can accurately identify the type of the fluid, for example, the fluid is identified to belong to mud (emulsified fluid or invalid measurement), gas, oil, water, oil-water mixture and gas-water mixture, thereby improving the accuracy of fluid type identification, improving the identification efficiency, reducing the labor cost of fluid type identification, laying a solid foundation for further eliminating the influence of water components on oil and gas spectrums in spectrum composition analysis and more accurately analyzing oil and gas compositions, and further reducing the operation risk and saving the operation time and cost.
Fig. 2 shows a schematic structural diagram of a fluid type identification device based on downhole spectral measurement according to an embodiment of the invention. As shown in fig. 2, the apparatus includes:
a first acquiring module 201 adapted to acquire fluid sample training data, wherein the fluid sample training data comprises: fluid spectrum data and a labeled fluid type label corresponding to the fluid spectrum data;
the model training module 202 is suitable for performing model training on the neural network according to the fluid sample training data to obtain a fluid type identification model;
a second obtaining module 203, adapted to obtain fluid spectrum data of the fluid to be identified measured by the downhole spectrum measuring device;
the identification module 204 is adapted to input the fluid spectrum data into the fluid type identification model for downhole fluid type identification, so as to obtain a fluid type corresponding to the fluid to be identified.
Optionally, the apparatus further comprises: the processing module is suitable for preprocessing the fluid sample training data to obtain preprocessed fluid sample training data;
performing dimensionality reduction on the preprocessed fluid sample training data;
the model training module is further adapted to: and performing model training on the neural network according to the fluid sample training data subjected to the dimension reduction treatment to obtain a fluid type identification model.
Optionally, the model training module is further adapted to: performing model training on the neural network according to the fluid spectrum data to obtain a fluid type result;
calculating the loss between the fluid type result and the labeled fluid type label to obtain a model loss function, and updating the model parameters of the neural network according to the model loss function;
and (3) iteratively executing the steps until the iteration times reach preset iteration times, and/or the output value of the model loss function is smaller than a preset threshold value, so as to obtain the fluid type identification model.
Optionally, the fluid spectral data comprises: and the spectrum channel and the fluid absorbance data corresponding to the spectrum channel.
Optionally, the apparatus further comprises: and the transmission module is suitable for uploading the fluid type corresponding to the fluid to be identified to the ground logging system so that the ground logging system can carry out logging operation according to the fluid type.
The scheme provided by the invention can accurately identify the type of the fluid, for example, the fluid is identified to belong to mud (emulsified fluid or invalid measurement), gas, oil, water, oil-water mixture and gas-water mixture, thereby improving the accuracy of fluid type identification, improving the identification efficiency, reducing the labor cost of fluid type identification, laying a solid foundation for further eliminating the influence of water components on oil and gas spectrums in spectral composition analysis and more accurately analyzing oil and gas compositions, and further reducing the operation risk and saving the operation time and cost.
The embodiment of the invention also provides a sampling instrument, which comprises: the fluid type identification device based on the downhole spectral measurement and the downhole spectral measurement device are provided. The sampling instrument may be a cable formation pressure measurement sampling instrument (EFDT) or an while drilling sampling Instrument (IFSA), and specific working principles are not described herein with reference to corresponding parts of the embodiment shown in fig. 1A.
The embodiment of the application also provides a nonvolatile computer storage medium, and the computer storage medium stores at least one executable instruction, and the computer executable instruction can execute the fluid type identification method based on the downhole spectral measurement in any method embodiment.
Fig. 3 is a schematic structural diagram of a computing device according to an embodiment of the present invention, and the specific embodiment of the present invention does not limit the specific implementation of the computing device.
As shown in fig. 3, the computing device may include: a processor (processor) 302, a communication Interface 304, a memory 306, and a communication bus 308.
Wherein:
the processor 302, communication interface 304, and memory 306 communicate with each other via a communication bus 308.
A communication interface 304 for communicating with network elements of other devices, such as clients or other servers.
The processor 302, configured to execute the program 310, may specifically execute the relevant steps in the embodiment of the fluid type identification method based on the downhole spectral measurement.
In particular, program 310 may include program code comprising computer operating instructions.
The processor 302 may be a central processing unit CPU, or an Application Specific Integrated Circuit ASIC (Application Specific Integrated Circuit), or one or more Integrated circuits configured to implement embodiments of the present invention. The computing device includes one or more processors, which may be the same type of processor, such as one or more CPUs; or may be different types of processors such as one or more CPUs and one or more ASICs.
And a memory 306 for storing a program 310. Memory 306 may comprise high-speed RAM memory and may also include non-volatile memory (non-volatile memory), such as at least one disk memory.
The program 310 may be specifically configured to cause the processor 302 to perform a downhole spectroscopy measurement based fluid type identification method in any of the method embodiments described above. For specific implementation of each step in the procedure 310, reference may be made to corresponding steps and corresponding descriptions in units in the above embodiment of fluid type identification based on downhole spectral measurement, which are not described herein again. It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described devices and modules may refer to the corresponding process descriptions in the foregoing method embodiments, and are not described herein again.
The algorithms or displays presented herein are not inherently related to any particular computer, virtual system, or other apparatus. Various general purpose systems may also be used with the teachings herein. The required structure for constructing such a system will be apparent from the description above. In addition, embodiments of the present invention are not directed to any particular programming language. It is appreciated that a variety of programming languages may be used to implement the teachings of the present invention as described herein, and any descriptions of specific languages are provided above to disclose the best mode of the invention.
In the description provided herein, numerous specific details are set forth. It is understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the invention, various features of the embodiments of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the invention and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be interpreted as reflecting an intention that: that the invention as claimed requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.
Those skilled in the art will appreciate that the modules in the device in an embodiment may be adaptively changed and disposed in one or more devices different from the embodiment. The modules or units or components in the embodiments may be combined into one module or unit or component, and furthermore, may be divided into a plurality of sub-modules or sub-units or sub-components. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or elements of any method or apparatus so disclosed, may be combined in any combination, except combinations where at least some of such features and/or processes or elements are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.
Furthermore, those skilled in the art will appreciate that while some embodiments herein include some features included in other embodiments, rather than other features, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the following claims, any of the claimed embodiments may be used in any combination.
The various component embodiments of the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that a microprocessor or Digital Signal Processor (DSP) may be used in practice to implement some or all of the functionality of some or all of the components according to embodiments of the present invention. The present invention may also be embodied as apparatus or device programs (e.g., computer programs and computer program products) for performing a portion or all of the methods described herein. Such programs implementing the present invention may be stored on computer-readable media or may be in the form of one or more signals. Such a signal may be downloaded from an internet website or provided on a carrier signal or in any other form.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means can be embodied by one and the same item of hardware. The usage of the words first, second and third, etcetera do not indicate any ordering. These words may be interpreted as names. The steps in the above embodiments should not be construed as limiting the order of execution unless specified otherwise.
Claims (10)
1. A method for fluid type identification based on downhole spectral measurements, comprising:
acquiring fluid sample training data, wherein the fluid sample training data comprises: fluid spectral data and a labeled fluid type label corresponding to the fluid spectral data;
performing model training on a neural network according to the fluid sample training data to obtain a fluid type identification model;
acquiring fluid spectral data of a fluid to be identified, which is measured by a downhole spectral measuring device;
and inputting the fluid spectrum data into the fluid type identification model for carrying out downhole fluid type identification to obtain the fluid type corresponding to the fluid to be identified.
2. The method of claim 1, wherein prior to model training a neural network from the fluid sample training data, the method further comprises:
preprocessing the fluid sample training data to obtain preprocessed fluid sample training data;
performing dimensionality reduction on the preprocessed fluid sample training data;
training a neural network according to the fluid sample training data to obtain a fluid type recognition model, further comprising:
and performing model training on the neural network according to the fluid sample training data subjected to the dimension reduction treatment to obtain a fluid type identification model.
3. The method of claim 1 or 2, wherein the model training of the neural network according to the fluid sample training data, resulting in the fluid type recognition model further comprises:
performing model training on a neural network according to the fluid spectrum data to obtain a fluid type result;
calculating the loss between the fluid type result and the labeled fluid type label to obtain a model loss function, and updating the model parameters of the neural network according to the model loss function;
and (3) iteratively executing the steps until the iteration times reach preset iteration times, and/or the output value of the model loss function is smaller than a preset threshold value, so as to obtain the fluid type identification model.
4. The method of claim 1 or 2, wherein the fluid spectral data comprises: and the spectrum channel and the fluid absorbance data corresponding to the spectrum channel.
5. The method of claim 4, wherein the fluid spectral data of the fluid to be identified measured by the downhole spectral measuring device further comprises:
the underground spectral measurement device measures the incident light intensity corresponding to each spectral channel before the light beam generated by the light source passes through the fluid to be identified and the transmitted light intensity corresponding to each spectral channel after the light beam passes through the fluid to be identified;
and calculating fluid absorbance data corresponding to each spectral channel according to the incident light intensity and the transmitted light intensity.
6. The method according to claim 1 or 2, wherein after obtaining the fluid type corresponding to the fluid to be identified, the method further comprises:
and uploading the fluid type corresponding to the fluid to be identified to a ground logging system so that the ground logging system can carry out logging operation according to the fluid type.
7. A fluid type identification device based on downhole spectral measurements, comprising:
a first acquisition module adapted to acquire fluid sample training data, wherein the fluid sample training data comprises: fluid spectral data and a labeled fluid type label corresponding to the fluid spectral data;
the model training module is suitable for performing model training on the neural network according to the fluid sample training data to obtain a fluid type identification model;
the second acquisition module is suitable for acquiring fluid spectral data of the fluid to be identified, which is measured by the downhole spectral measurement device;
and the identification module is suitable for inputting the fluid spectrum data into the fluid type identification model to carry out downhole fluid type identification so as to obtain the fluid type corresponding to the fluid to be identified.
8. A sampler, comprising: the downhole spectral measurement based fluid type identification device, the downhole spectral measurement device of claim 7.
9. A computing device, comprising: the system comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface are communicated with each other through the communication bus;
the memory is used for storing at least one executable instruction which causes the processor to execute the operation corresponding to the fluid type identification method based on the downhole spectral measurement according to any one of claims 1-6.
10. A computer storage medium having stored therein at least one executable instruction that causes a processor to perform operations corresponding to the downhole spectral measurement based fluid type identification method of any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211634147.7A CN115855872A (en) | 2022-12-19 | 2022-12-19 | Fluid type identification method and device based on downhole spectral measurement and sampling instrument |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211634147.7A CN115855872A (en) | 2022-12-19 | 2022-12-19 | Fluid type identification method and device based on downhole spectral measurement and sampling instrument |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115855872A true CN115855872A (en) | 2023-03-28 |
Family
ID=85674180
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211634147.7A Pending CN115855872A (en) | 2022-12-19 | 2022-12-19 | Fluid type identification method and device based on downhole spectral measurement and sampling instrument |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115855872A (en) |
-
2022
- 2022-12-19 CN CN202211634147.7A patent/CN115855872A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110056348B (en) | Method and system for determining formation fluid composition and properties | |
| US20200257015A1 (en) | Model based discriminant analysis | |
| EP2904207B1 (en) | Determining fluid composition downhole from optical spectra | |
| US8156800B2 (en) | Methods and apparatus to evaluate subterranean formations | |
| EP3596638A1 (en) | Collaborative sensing and prediction of source rock properties | |
| US11561215B2 (en) | Scale-coupled multiscale model simulation | |
| US9650892B2 (en) | Blended mapping for estimating fluid composition from optical spectra | |
| US20220403737A1 (en) | Determining Asphaltene Onset | |
| US20140260586A1 (en) | Method to perform rapid formation fluid analysis | |
| BR112016013211B1 (en) | METHOD TO IDENTIFY PROPERTIES OF FORMATION FLUID AND EQUIPMENT | |
| US20220404521A1 (en) | Optical sensor adaptive calibration | |
| US20220074303A1 (en) | Determining reservoir fluid properties from downhole fluid analysis data using machine learning | |
| WO2020005238A1 (en) | Methods for predicting properties of clean formation fluid using real time downhole fluid analysis of contaminated samples | |
| US11550975B2 (en) | Methods and systems for predicting interfacial tension of reservoir fluids using downhole fluid measurements | |
| WO2021108603A1 (en) | Resolution preserving methodology to generate continuous log scale reservoir permeability profile from petrographic thin section images | |
| US20230313680A1 (en) | Determination of downhole formation fluid contamination and certain component concentrations | |
| CN117825320A (en) | Non-separated ground detection method and device for drilling return liquid oil gas ratio measurement | |
| CN104880737A (en) | Multivariate Logistic method using logging information to identify type of underground fluid | |
| CN115855872A (en) | Fluid type identification method and device based on downhole spectral measurement and sampling instrument | |
| BR112019003466B1 (en) | METHOD, DEVICE AND SYSTEM FOR OPTICAL ANALYSIS USING MULTIPLE INTEGRATED COMPUTATIONAL ELEMENTS | |
| BR112019028021B1 (en) | METHOD AND SYSTEM FOR OBTAINING REAL-TIME PROPERTY MEASUREMENTS OF A FLUID | |
| WO2023196389A1 (en) | Determination of asphaltene onset condition of reservoir fluids during downhole fluid analysis | |
| WO2024102178A1 (en) | Determining ion concentration through downhole optical spectroscopy | |
| WO2024129835A1 (en) | Systems and methods for determining carbon dioxide concentrations using peak ratio-based optical spectrometric measurements | |
| WO2024043868A1 (en) | Quality assessment of downhole reservoir fluid sampling by predicted interfacial tension |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |