Shale gas development index prediction method based on engineering parameters

Abstract

The invention discloses a shale gas development index prediction method based on engineering parameters, and the method comprises the steps: obtaining historical engineering parameters corresponding to shale gas, and carrying out the preprocessing of the historical engineering parameters, and obtaining basic support data; constructing a shale gas development index prediction model based on trend prediction by adopting a deep learning model, and training the shale gas development index prediction model by adopting an improved optimization algorithm according to the basic support data to obtain a trained shale gas development index prediction model; and acquiring real-time production data corresponding to the shale gas, and predicting the shale gas yield by adopting the trained shale gas development index prediction model according to the real-time production data corresponding to the shale gas to obtain a prediction result. According to the method, the historical production data corresponding to the shale gas are collected, the shale gas development index prediction model based on trend prediction and the improved optimization algorithm are adopted to conduct deep learning on the historical data of the shale gas, and therefore laws in the historical data of the shale gas are explored, and the accuracy of shale gas yield prediction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of shale gas, and in particular to a shale gas development index prediction method based on engineering parameters. Background Art

[0002] Shale gas is an unconventional natural gas composed mainly of methane, stored in mudstone and its interlayers that are rich in organic matter. Its high quality, cleanliness and high efficiency make it a low-carbon energy source that can effectively alleviate China's current energy pressure, and it also occupies an increasingly important position in primary energy consumption.

[0003] The effective development and utilization of shale gas has important comprehensive benefits, which are manifested in the comprehensive promotion of economy, ecology and society. Due to factors such as the low recovery rate of shale gas in my country, many shale gas development projects have been on the edge of economic feasibility for a long time, and even incurred losses. However, its contribution to society and ecology cannot be ignored. Therefore, such projects need to be further comprehensively evaluated, which has important practical significance for reasonable area selection, reducing losses and reducing project risks.

[0004] In the prior art, the comprehensive benefits of shale gas are generally calculated to obtain the net profit during the entire project. When calculating the comprehensive benefits, and when the market price does not change, the total benefit of shale gas is the output multiplied by the market price, and the total benefit minus the cost is the comprehensive benefit. However, the prediction effect of shale gas production in the prior art is poor, which leads to the inability to correctly predict the total benefit of shale gas, making it impossible for managers to schedule personnel and equipment in real time according to the total benefit and profit goals, resulting in low shale gas mining efficiency and a large waste of manpower and material resources. The prior art generally directly classifies historical data to predict shale gas production, without considering the changing trend of features, resulting in inaccurate predictions, and the prior art model is poorly trained (such as training using the gradient descent method), resulting in poor final prediction results. Summary of the invention

[0005] The purpose of the present invention is to provide a shale gas development index prediction method based on engineering parameters, which solves the technical problems existing in the prior art.

[0006] The present invention is achieved through the following technical solutions:

[0007] A shale gas development index prediction method based on engineering parameters, comprising:

[0008] Obtaining historical engineering parameters corresponding to shale gas, and preprocessing the historical engineering parameters to obtain basic supporting data;

[0009] A shale gas development index prediction model based on trend prediction is constructed by using a deep learning model, and the shale gas development index prediction model is trained according to the basic support data and an improved optimization algorithm to obtain a trained shale gas development index prediction model;

[0010] The real-time production data corresponding to shale gas is obtained, and the shale gas production is predicted based on the real-time production data corresponding to shale gas and the trained shale gas development indicator prediction model to obtain the prediction result.

[0011] In a possible implementation, obtaining historical engineering parameters corresponding to shale gas and preprocessing the historical engineering parameters include:

[0012] Obtaining historical equipment characteristics, historical site characteristics, and historical production of shale gas corresponding to multiple consecutive time periods, wherein each time period corresponds to a historical equipment characteristic, a historical site characteristic, and a historical production;

[0013] Cluster the historical production corresponding to all time periods to obtain multiple cluster centers, and obtain the Euclidean distance between the historical production of each time period and each cluster center, determine the target cluster center with the shortest Euclidean distance to the historical production of the time period, and use the target cluster center to replace the historical production of the time period;

[0014] The feature matrix is ​​constructed by using the historical equipment features and historical site features corresponding to N consecutive time periods; the output data is constructed by using the target cluster center corresponding to the N+1th time end;

[0015] According to the characteristic matrix and the output data, training data is constructed, and a plurality of training data are repeatedly acquired to complete the preprocessing of the historical engineering parameters.

[0016] In a possible implementation, the historical production corresponding to all time periods is clustered to obtain multiple cluster centers, including:

[0017] Get the number of clusters as K, and randomly generate K cluster centers;

[0018] Obtain the Euclidean distance between the historical output corresponding to each time period and the K cluster centers, determine the nearest cluster center with the shortest Euclidean distance to each historical output, and classify each historical output into the corresponding nearest cluster center;

[0019] Get the mean of all historical productions corresponding to each cluster center, and determine whether there is a cluster center whose corresponding mean is different from the cluster center. If so, for the cluster center that is different from the mean, replace the cluster center with the mean and re-cluster until the means corresponding to all cluster centers are the same as the cluster center, and get K cluster centers. Otherwise, get K cluster centers.

[0020] In a possible implementation, a deep learning model is used to construct a shale gas development indicator prediction model based on trend prediction, including:

[0021] Constructing a first feature conversion module, a feature preprocessing module, a second feature conversion module, and a shale gas development index prediction module that are connected in sequence to obtain a shale gas development index prediction model;

[0022] The first feature conversion module is used to extract each column of data of the feature matrix, and sequentially send each column of extracted data to the feature preprocessing module;

[0023] The feature preprocessing module includes three parallel channels, and the three parallel channels are sequentially set to a conventional gray model, an unbiased gray model or a smoothly modified gray model, and each column of extracted data is preprocessed by the conventional gray model, the unbiased gray model or the smoothly modified gray model to obtain three preprocessed column data;

[0024] The second feature conversion module is used to perform weighted processing on the three preprocessed data to obtain final preprocessed column data, and restore all the preprocessed column data into a feature matrix;

[0025] The shale gas development index prediction module is set as a deep convolutional neural network, and the restored feature matrix is ​​used as the input of the shale gas development index prediction module to obtain a prediction result.

[0026] In a possible implementation, the shale gas development index prediction model is trained using an intelligent optimization algorithm to obtain a trained shale gas development index prediction model, including:

[0027] A1. Randomly initialize the network parameters of the shale gas development index prediction model between the upper limit and the lower limit of the network parameters to obtain a single optimal individual, and repeatedly obtain multiple optimal individuals to obtain an optimal population;

[0028] A2. Obtain the fitness value of each optimization individual, and determine the optimization individual with the largest fitness value to obtain the best individual;

[0029] A3. Update each optimal individual according to the best individual to obtain the optimal individual after initial update;

[0030] A4. Locally update the updated optimal individuals and generate new individuals to replace the M optimal individuals with the smallest fitness values ​​to obtain the optimal population after the second update;

[0031] A5. Based on the optimization population, determine whether the iteration end condition is met. If so, use the optimization individual with the largest fitness value in the optimization population as the final network parameter of the shale gas development index prediction model and end the training. Otherwise, return to step A2.

[0032] In a possible implementation, obtaining the fitness value of each optimization individual, and determining the optimization individual with the largest fitness value to obtain the best individual includes:

[0033] For each optimal individual, obtain its fitness value as follows:

[0034]

[0035] Where i = 1, 2, ..., I, I represents the total number of batch samples, j = 1, 2, ..., J, J represents the total number of outputs of the shale gas development index prediction model, and y ij represents the jth actual output of the shale gas development index prediction model after the i-th sample input, Indicates i The jth expected output of the shale gas development index prediction model after samples are input;

[0036] Determine the optimal individual with the largest fitness value and obtain the best individual.

[0037] In a possible implementation, updating each optimal individual according to the best individual to obtain the optimal individual after initial update includes:

[0038]

[0039]

[0040]

[0041]

[0042] in, represents the lth network parameter of the kth optimal individual in the optimal population during the t-1th training. Indicates updated represents the update speed corresponding to the lth network parameter in the kth optimization individual, ω1 represents the first proportional weight coefficient, ω2 represents the second proportional weight coefficient, c represents the self-learning factor, and f k represents the optimization frequency corresponding to the kth optimization individual, x best,l represents the lth network parameter in the best individual, represents the update speed corresponding to the lth network parameter in the kth optimization individual during the t-1th training, c1 represents the first speed update factor, c2 represents the second speed update factor, and p best,k,l represents the lth network parameter when the fitness of the kth optimization individual is the largest during the tth training process, ω max Represents the maximum value of the proportional weight coefficient, ω min Represents the minimum value of the proportional weight coefficient, T max represents the preset maximum number of training times, e represents a natural constant, and t represents the current number of training times.

[0043] In a possible implementation, locally updating the updated optimization individual includes:

[0044] x' best,l =x best,l +εA t

[0045] A t =αA t-1

[0046] Among them, x' best,l Represents the updated x best,l , ε represents a random number between (0,1), A t represents the average loudness of the optimal population during the t-th training, A t-1 represents the average loudness during the t-1th training; each optimal individual is assigned an initial loudness before the iteration begins; α represents a random number between (0,1).

[0047] In a possible implementation, generating new individuals to replace the M optimal individuals with the smallest fitness values ​​to obtain a secondary updated optimal population includes:

[0048] Determine the fitness value of each optimization individual in the optimization population after the initial update, and determine the M first target optimization individuals with the smallest fitness values;

[0049] For any first-target optimization individual, randomly select a second-target optimization individual from the optimization population, and perform a crossover operation on the first-target optimization individual and the second-target optimization individual to obtain two third-target optimization individuals;

[0050] Determine whether there is a third target optimization individual whose fitness value is greater than the first target optimization individual. If so, replace the first target optimization individual with the third target optimization individual with the largest fitness value to obtain a new first target optimization individual. Otherwise, perform a mutation operation.

[0051] For any first-objective optimizing individual, the parameters in the first-objective optimizing individual are mutated to obtain a new first-objective optimizing individual.

[0052] In a possible implementation manner, after each optimization individual is updated according to the best individual, the method further includes: updating the optimization frequency corresponding to the optimization individual, and the update is:

[0053]

[0054] in, represents the optimization frequency during the tth training, represents the initial optimization frequency of the kth optimization individual, and γ represents the frequency enhancement coefficient.

[0055] The present invention provides a shale gas development index prediction method based on engineering parameters. For a certain shale gas mining base, historical production data corresponding to shale gas is collected, and a shale gas development index prediction model based on trend prediction and an improved optimization algorithm are used to perform deep learning on the historical data of shale gas, thereby exploring the laws in the historical data of shale gas and improving the accuracy of shale gas production prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other relevant drawings can be obtained based on these drawings without creative work. In the drawings:

[0057] Figure 1 A flow chart of a shale gas development index prediction method based on engineering parameters provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.

[0059] Example

[0060] like Figure 1 As shown, an embodiment of the present invention provides a shale gas development index prediction method based on engineering parameters, including:

[0061] S1. Obtain historical engineering parameters corresponding to shale gas, and preprocess the historical engineering parameters to obtain basic support data.

[0062] Optionally, the historical engineering parameters corresponding to shale gas may include historical equipment characteristics and historical site characteristics. The historical equipment characteristics mainly include the type and quantity of equipment used for shale gas, and the historical site characteristics mainly include the land type, the area occupied by each type, and the basic properties of shale. The historical engineering parameters corresponding to shale gas may also include historical production. It is worth noting that in addition to the above characteristics, other characteristics may be selected to further optimize the solution. It is also possible to use only the basic properties of shale as historical engineering parameters, which can be determined by the staff according to the actual situation.

[0063] The basic properties of shale may include shale composition, mud-to-ground ratio, organic carbon content TOC, organic matter maturity Ro, porosity, net shale thickness, adsorbed gas content, and production depth. In addition to the above properties, other historical properties can also be collected, such as reservoir thickness, initial formation pressure, matrix permeability, matrix porosity, fractured volume area permeability, fractured volume area porosity, fracture permeability, fracture half-length, and fracture number.

[0064] S2. Use a deep learning model to build a shale gas development index prediction model based on trend prediction. According to the basic support data, and using an improved optimization algorithm, train the shale gas development index prediction model to obtain a trained shale gas development index prediction model.

[0065] S3. Obtain the real-time production data corresponding to shale gas, and use the trained shale gas development index prediction model to predict the shale gas production according to the real-time production data corresponding to shale gas to obtain the prediction results. Therefore, the staff can guide the gas well production allocation according to the prediction results.

[0066] In a possible implementation, obtaining historical engineering parameters corresponding to shale gas and preprocessing the historical engineering parameters include:

[0067] The historical equipment characteristics, historical site characteristics and historical production of shale gas corresponding to multiple consecutive time periods are obtained, wherein each time period corresponds to a historical equipment characteristic, a historical site characteristic and a historical production.

[0068] Optionally, during the shale gas extraction process, the geological features will not undergo sudden changes, so a short-term forecast is used to predict shale gas production, so that staff can conduct real-time scheduling and optimize the extraction plan to ensure the comprehensive benefits of shale gas.

[0069] The main purpose of the embodiment of the present invention is to predict the production of shale gas so that management personnel can predict the total benefit based on the predicted production and thus schedule personnel and equipment.

[0070] The historical equipment characteristics, historical site characteristics, and historical production corresponding to shale gas in multiple consecutive time periods may be data from a shale gas base that is currently being mined, or data from other shale gas bases.

[0071] The historical production corresponding to all time periods is clustered to obtain multiple cluster centers, and the Euclidean distance between the historical production of each time period and each cluster center is obtained. The target cluster center with the shortest Euclidean distance to the historical production of the time period is determined, and the target cluster center is used to replace the historical production of the time period.

[0072] The shale gas production and equipment usage vary between different mining bases, which will lead to different historical production. In order to reduce the amount of data, the historical production is clustered. It is worth noting that the more clusters there are, the higher the final prediction effect will be. The staff can set it according to actual needs.

[0073] The feature matrix is ​​constructed by using the historical equipment features and historical site features corresponding to N consecutive time periods. The output data is constructed by using the target cluster center corresponding to the N+1th time end.

[0074] The feature matrix can be constructed by the following method: the historical equipment features and historical site features of each time period are combined into a feature vector, and the feature vectors of multiple consecutive time periods are combined into a feature matrix. It is worth noting that in order to facilitate data processing, various types of features in this embodiment are arranged in a fixed order and are all represented by numbers.

[0075] Optionally, in order to reduce the amount of subsequent data processing, the elements in the feature matrix may be normalized to make the data easier to process and speed up the recognition process.

[0076] According to the characteristic matrix and the output data, training data is constructed, and a plurality of training data are repeatedly acquired to complete the preprocessing of the historical engineering parameters.

[0077] In a possible implementation, the historical production corresponding to all time periods is clustered to obtain multiple cluster centers, including:

[0078] The number of clusters is obtained as K, and K cluster centers are randomly generated.

[0079] Obtain the Euclidean distance between the historical output corresponding to each time period and the K cluster centers, determine the nearest cluster center with the shortest Euclidean distance to each historical output, and classify each historical output into the corresponding nearest cluster center.

[0080] Get the mean of all historical productions corresponding to each cluster center, and determine whether there is a cluster center whose corresponding mean is different from the cluster center. If so, for the cluster center that is different from the mean, replace the cluster center with the mean and re-cluster until the means corresponding to all cluster centers are the same as the cluster center, and get K cluster centers. Otherwise, get K cluster centers.

[0081] Optionally, if the historical output is only one-dimensional data, the Euclidean distance between two data is the difference between the two data.

[0082] In a possible implementation, a deep learning model is used to construct a shale gas development indicator prediction model based on trend prediction, including:

[0083] A first feature conversion module, a feature preprocessing module, a second feature conversion module and a shale gas development index prediction module are sequentially connected to obtain a shale gas development index prediction model.

[0084] The first feature conversion module is used to extract each column of data from the feature matrix, and send each column of extracted data to the feature preprocessing module in sequence.

[0085] The feature preprocessing module includes three parallel channels, and the three parallel channels are set to a conventional gray model, an unbiased gray model or a smoothly modified gray model in turn. Each column of extracted data is preprocessed by the conventional gray model, the unbiased gray model or the smoothly modified gray model to obtain three preprocessed column data.

[0086] Grey prediction refers to predicting the development and change of the characteristic value of system behavior. Therefore, the present invention adopts three parallel grey models to predict the characteristic value, so that the characteristic value in the short term can be predicted.

[0087] The second feature conversion module is used to perform weighted processing on the three preprocessed data to obtain final preprocessed column data, and restore all the preprocessed column data into a feature matrix.

[0088] By weighting the three preprocessed data, the predicted characteristic values ​​can be obtained more accurately, and the predicted shale gas production can be determined based on the predicted characteristic values.

[0089] The shale gas development index prediction module is set as a deep convolutional neural network, and the restored feature matrix is ​​used as the input of the shale gas development index prediction module to obtain a prediction result.

[0090] Optionally, the feature matrix constructed in the present invention is equivalent to a simple image, so a deep convolutional neural network can be used to identify the feature matrix. In addition to the deep convolutional neural network used in this embodiment, other classification neural networks can also be used to explore the potential connection between the feature matrix and shale gas production.

[0091] The embodiment of the present invention predicts the output, so as to obtain the total benefit of shale gas development. Although the mining module is fixed, the gas storage capacity and the usage of the equipment may change. Therefore, the gas storage capacity and the usage of the equipment in the short term can be predicted based on historical data, so as to introduce a gray model for small sample prediction, which can ensure the prediction of the required parameters without requiring too many samples. However, the gray module is easily affected by random factors, and the long-term prediction effect is not good. Therefore, this application introduces an unbiased gray module and a gray model improved by the sliding average method for prediction, which can effectively eliminate the image of random factors and solve the problem of limited prediction length. At the same time, the output can be estimated more accurately through the predicted data.

[0092] In a possible implementation, the shale gas development index prediction model is trained using an intelligent optimization algorithm to obtain a trained shale gas development index prediction model, including:

[0093] A1. Randomly initialize the network parameters of the shale gas development index prediction model between the upper limit and the lower limit of the network parameters to obtain a single optimal individual, and repeatedly obtain multiple optimal individuals to obtain an optimal population.

[0094] Optionally, in order to better find the optimal network parameters of the shale gas development index prediction model, the network parameters of the shale gas development index prediction model can be initialized using a chaotic sequence to make the optimal individuals more evenly distributed in the solution space. It is worth noting that the network parameters of the shale gas development index prediction model can include weights and biases.

[0095] A2. Obtain the fitness value of each optimizing individual, determine the optimizing individual with the largest fitness value, and obtain the best individual.

[0096] A3. Update each optimal individual according to the best individual to obtain the optimal individual after the initial update.

[0097] A4. Perform local updates on the updated optimal individuals and generate new individuals to replace the M optimal individuals with the smallest fitness values ​​to obtain the optimal population after secondary updates.

[0098] A5. Based on the optimization population, determine whether the iteration end condition is met. If so, use the optimization individual with the largest fitness value in the optimization population as the final network parameter of the shale gas development index prediction model and end the training. Otherwise, return to step A2.

[0099] In order to solve the problems of slow convergence speed, low search accuracy and weak optimization ability of the bat optimization algorithm, this embodiment improves the bat optimization algorithm. The embodiment of the present invention constructs an intelligent optimization algorithm based on the bat optimization algorithm, as follows.

[0100] In a possible implementation, obtaining the fitness value of each optimization individual, and determining the optimization individual with the largest fitness value to obtain the best individual includes:

[0101] For each optimal individual, obtain its fitness value as follows:

[0102]

[0103] Where i = 1, 2, ..., I, I represents the total number of batch samples, j = 1, 2, ..., J, J represents the total number of outputs of the shale gas development index prediction model, and y ij represents the jth actual output of the shale gas development index prediction model after the i-th sample input, Indicates i The jth expected output of the shale gas development index prediction model after samples are input.

[0104] Determine the optimal individual with the largest fitness value and obtain the best individual.

[0105] In a possible implementation, updating each optimal individual according to the best individual to obtain the optimal individual after initial update includes:

[0106]

[0107]

[0108]

[0109]

[0110] in, represents the lth network parameter of the kth optimal individual in the optimal population during the t-1th training. Indicates updated represents the update speed corresponding to the lth network parameter in the kth optimization individual, ω1 represents the first proportional weight coefficient, ω2 represents the second proportional weight coefficient, c represents the self-learning factor, and f krepresents the optimization frequency corresponding to the kth optimization individual, x best,l represents the lth network parameter in the best individual, represents the update speed corresponding to the lth network parameter in the kth optimization individual during the t-1th training, c1 represents the first speed update factor, c2 represents the second speed update factor, and p best,k,l represents the lth network parameter when the fitness of the kth optimization individual is the largest during the tth training process, ω max Represents the maximum value of the proportional weight coefficient, ω min Represents the minimum value of the proportional weight coefficient, T max represents the preset maximum number of training times, e represents a natural constant, and t represents the current number of training times.

[0111] Optionally, the self-learning factor c can be:

[0112]

[0113] Among them, c max represents the first preset update value, c min represents the second preset update value, a represents a constant between (0,1), and π represents the ratio of pi.

[0114] For the updating process of the optimizing individual, the embodiment of the present invention sets a first proportional weight coefficient and a second proportional weight coefficient. In the early stage of iteration, the second proportional weight coefficient is greater than the first proportional weight coefficient, so that the optimizing individual can quickly approach the optimal individual, which is beneficial to the global search and accelerates the convergence speed. In the later stage of iteration, the second proportional weight coefficient is less than the first proportional weight coefficient, and the position between the optimizing individual and the optimal individual is relatively close, which is beneficial to local search, balances local search and global search, and effectively avoids the algorithm premature phenomenon. By improving the update speed, the algorithm has a stronger search ability and convergence speed, thereby further improving the optimization ability of the intelligent optimization algorithm and training a more accurate shale gas development indicator prediction model.

[0115] In a possible implementation, locally updating the updated optimization individual includes:

[0116] x' best,l =x best,l +εA t

[0117] A t =αA t-1

[0118] Among them, x' best,l Represents the updated x best,l , ε represents a random number between (0,1), A trepresents the average loudness of the optimal population during the t-th training, A t-1 represents the average loudness at the t-1th training. Each optimization individual is assigned an initial loudness before the iteration begins. α represents a random number between (0,1).

[0119] When the optimization population gradually approaches the global optimal solution, the algorithm adopts a local search strategy to find the global optimal solution more accurately.

[0120] In one possible implementation, new individuals are generated to replace the M optimizing individuals with the smallest fitness values ​​to obtain a second updated optimizing population, including: determining the fitness value of each optimizing individual in the optimizing population after the initial update, and determining the M first target optimizing individuals with the smallest fitness values. For any first target optimizing individual, a second target optimizing individual is randomly selected from the optimizing population, and the first target optimizing individual is cross-operated with the second target optimizing individual to obtain two third target optimizing individuals. It is determined whether there is a third target optimizing individual whose fitness value is greater than the first target optimizing individual. If so, the first target optimizing individual is replaced with the third target optimizing individual with the largest fitness value to obtain a new first target optimizing individual, otherwise a mutation operation is performed. For any first target optimizing individual, the parameters in the first target optimizing individual are mutated to obtain a new first target optimizing individual.

[0121] During each iteration, the optimization individual with a smaller fitness can be considered as having a smaller value, so a new individual can be generated to update it to ensure the optimization ability of the entire optimization population. In order to avoid the population from becoming worse, the survival of the fittest method can be adopted when performing mutation and crossover operations, comparing the individuals before mutation with the individuals after mutation, and leaving the individuals with larger fitness values. And comparing the individuals before crossover with the individuals after mutation, leaving the individuals with larger fitness values.

[0122] Optionally, the crossover operation may be to exchange some parameters of the two individuals, thereby obtaining two new individuals. In this embodiment, the selected parameters are between 20% and 50% of the total parameters, and are selected by random selection.

[0123] Optionally, the mutation operation may be: randomly generating a decision factor between (0, 1), and determining whether the decision factor is greater than a preset decision probability; if so, performing a first mutation operation on the optimal individual; otherwise, performing a second mutation operation on the optimal individual, and the decision probability is a constant between (0, 1).

[0124] The first mutation operation can be: X * =X+w(X max -X), where X represents the individual to be mutated, X* represents the variant individual, X max represents the upper limit of the optimal individual, w represents the variation inertia weight.

[0125] The second mutation operation can be: X * =X+w(XX min ), where X min Find the lower bound of the optimal individual.

[0126] At the end of each training, the fitness values ​​of the optimization individuals before and after the crossover and mutation operations are compared, and the better individuals are selected to enter the next iteration, ensuring that the individuals retained in each iteration are elite individuals. At the same time, an adaptive decision probability is set in the mutation process to reduce the impact on the early trajectory of the algorithm optimization, increase the mutation probability in the later stage of the algorithm, and increase the probability of jumping out of the local extreme value in the later stage of optimization, so as to prevent falling into the local extreme value too early and making it difficult to search for the global optimal solution.

[0127] In a possible implementation manner, after each optimization individual is updated according to the best individual, the method further includes: updating the optimization frequency corresponding to the optimization individual, and the update is:

[0128]

[0129] in, represents the optimization frequency during the tth training, represents the initial optimization frequency of the kth optimization individual, and γ represents the frequency enhancement coefficient.

[0130] The optimization algorithm provided by the present invention can effectively avoid the problem that the prior art falls into the local optimum during the training process, and effectively improves the training speed.

[0131] The present invention provides a shale gas development index prediction method based on engineering parameters. For a certain shale gas mining base, historical production data corresponding to shale gas is collected, and a shale gas development index prediction model based on trend prediction and an improved optimization algorithm are used to perform deep learning on the historical data of shale gas, thereby exploring the laws in the historical data of shale gas and improving the accuracy of shale gas production prediction.

[0132] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A shale gas development index prediction method based on engineering parameters, characterized in that: include: Obtaining historical engineering parameters corresponding to shale gas, and preprocessing the historical engineering parameters to obtain basic supporting data; A shale gas development index prediction model based on trend prediction is constructed by using a deep learning model, and the shale gas development index prediction model is trained according to the basic support data and an improved optimization algorithm to obtain a trained shale gas development index prediction model; The real-time production data corresponding to shale gas is obtained, and the shale gas production is predicted based on the real-time production data corresponding to shale gas and the trained shale gas development indicator prediction model to obtain the prediction result.

2. The shale gas development index prediction method based on engineering parameters according to claim 1 is characterized in that: Obtaining historical engineering parameters corresponding to shale gas and preprocessing the historical engineering parameters includes: Obtaining historical equipment characteristics, historical site characteristics, and historical production of shale gas corresponding to multiple consecutive time periods, wherein each time period corresponds to a historical equipment characteristic, a historical site characteristic, and a historical production; Cluster the historical production corresponding to all time periods to obtain multiple cluster centers, and obtain the Euclidean distance between the historical production of each time period and each cluster center, determine the target cluster center with the shortest Euclidean distance to the historical production of the time period, and use the target cluster center to replace the historical production of the time period; The feature matrix is ​​constructed by using the historical equipment features and historical site features corresponding to N consecutive time periods; the output data is constructed by using the target cluster center corresponding to the N+1th time end; According to the characteristic matrix and the output data, training data is constructed, and a plurality of training data are repeatedly acquired to complete the preprocessing of the historical engineering parameters.

3. The shale gas development index prediction method based on engineering parameters according to claim 2 is characterized in that: Cluster the historical production corresponding to all time periods to obtain multiple cluster centers, including: Get the number of clusters as K, and randomly generate K cluster centers; Obtain the Euclidean distance between the historical output corresponding to each time period and the K cluster centers, determine the nearest cluster center with the shortest Euclidean distance to each historical output, and classify each historical output into the corresponding nearest cluster center; Get the mean of all historical productions corresponding to each cluster center, and determine whether there is a cluster center whose corresponding mean is different from the cluster center. If so, for the cluster center that is different from the mean, replace the cluster center with the mean and re-cluster until the means corresponding to all cluster centers are the same as the cluster center, and get K cluster centers. Otherwise, get K cluster centers.

4. The shale gas development index prediction method based on engineering parameters according to claim 2 is characterized in that: A deep learning model is used to build a shale gas development indicator prediction model based on trend prediction, including: Constructing a first feature conversion module, a feature preprocessing module, a second feature conversion module, and a shale gas development index prediction module that are connected in sequence to obtain a shale gas development index prediction model; The first feature conversion module is used to extract each column of data of the feature matrix, and sequentially send each column of extracted data to the feature preprocessing module; The feature preprocessing module includes three parallel channels, and the three parallel channels are sequentially set to a conventional gray model, an unbiased gray model or a smoothly modified gray model, and each column of extracted data is preprocessed by the conventional gray model, the unbiased gray model or the smoothly modified gray model to obtain three preprocessed column data; The second feature conversion module is used to perform weighted processing on the three preprocessed data to obtain final preprocessed column data, and restore all the preprocessed column data into a feature matrix; The shale gas development index prediction module is set as a deep convolutional neural network, and the restored feature matrix is ​​used as the input of the shale gas development index prediction module to obtain a prediction result.

5. The shale gas development index prediction method based on engineering parameters according to claim 4 is characterized in that: The shale gas development index prediction model is trained by using an intelligent optimization algorithm to obtain a trained shale gas development index prediction model, including: A1. Randomly initialize the network parameters of the shale gas development index prediction model between the upper limit and the lower limit of the network parameters to obtain a single optimal individual, and repeatedly obtain multiple optimal individuals to obtain an optimal population; A2. Obtain the fitness value of each optimization individual, and determine the optimization individual with the largest fitness value to obtain the best individual; A3. Update each optimal individual according to the best individual to obtain the optimal individual after initial update; A4. Locally update the updated optimal individuals and generate new individuals to replace the M optimal individuals with the smallest fitness values ​​to obtain the optimal population after the second update; A5. Based on the optimization population, determine whether the iteration end condition is met. If so, use the optimization individual with the largest fitness value in the optimization population as the final network parameter of the shale gas development index prediction model and end the training. Otherwise, return to step A2.

6. The shale gas development index prediction method based on engineering parameters according to claim 5 is characterized in that: Obtain the fitness value of each optimization individual, and determine the optimization individual with the largest fitness value to obtain the best individual, including: For each optimal individual, obtain its fitness value as follows: Where i = 1, 2, ..., I, I represents the total number of batch samples, j = 1, 2, ..., J, J represents the total number of outputs of the shale gas development index prediction model, and y ij represents the jth actual output of the shale gas development index prediction model after the i-th sample input, represents the jth expected output of the shale gas development index prediction model after the i-th sample input; Determine the optimal individual with the largest fitness value and obtain the best individual.

7. The shale gas development index prediction method based on engineering parameters according to claim 6, characterized in that: Each optimal individual is updated according to the best individual to obtain the optimal individual after the initial update, including: in, represents the lth network parameter of the kth optimal individual in the optimal population during the t-1th training. Indicates updated represents the update speed corresponding to the lth network parameter in the kth optimization individual, ω1 represents the first proportional weight coefficient, ω2 represents the second proportional weight coefficient, c represents the self-learning factor, and f k represents the optimization frequency corresponding to the kth optimization individual, x best,l represents the lth network parameter in the best individual, represents the update speed corresponding to the lth network parameter in the kth optimization individual during the t-1th training, c1 represents the first speed update factor, c2 represents the second speed update factor, and p best,k,l represents the lth network parameter when the fitness of the kth optimization individual is the largest during the tth training process, ω max Represents the maximum value of the proportional weight coefficient, ω min Represents the minimum value of the proportional weight coefficient, T max represents the preset maximum number of training times, e represents a natural constant, and t represents the current number of training times.

8. The shale gas development index prediction method based on engineering parameters according to claim 7 is characterized in that: Perform local updates on the updated optimal individuals, including: x' best,l =x best,l +εA t A t =αA t-1 Among them, x' best,l Represents the updated x best,l , ε represents a random number between (0,1), A t represents the average loudness of the optimal population during the t-th training, A t-1 represents the average loudness during the t-1th training; each optimal individual is assigned an initial loudness before the iteration begins; α represents a random number between (0,1).

9. The shale gas development index prediction method based on engineering parameters according to claim 8, characterized in that: Generate new individuals to replace the M optimal individuals with the smallest fitness values, and obtain the optimal population after secondary update, including: Determine the fitness value of each optimization individual in the optimization population after the initial update, and determine the M first target optimization individuals with the smallest fitness values; For any first-target optimization individual, randomly select a second-target optimization individual from the optimization population, and perform a crossover operation on the first-target optimization individual and the second-target optimization individual to obtain two third-target optimization individuals; Determine whether there is a third target optimization individual whose fitness value is greater than the first target optimization individual. If so, replace the first target optimization individual with the third target optimization individual with the largest fitness value to obtain a new first target optimization individual. Otherwise, perform a mutation operation. For any first-objective optimizing individual, the parameters in the first-objective optimizing individual are mutated to obtain a new first-objective optimizing individual.

10. The shale gas development index prediction method based on engineering parameters according to claim 7, characterized in that: After each optimization individual is updated according to the best individual, the method further includes: updating the optimization frequency corresponding to the optimization individual, and the update is: in, represents the optimization frequency during the tth training, represents the initial optimization frequency of the kth optimization individual, and γ represents the frequency enhancement coefficient.

Images