Shale gas drilling cost prediction method and system

Abstract

The invention discloses a shale gas drilling cost prediction method and system, and the method comprises the following steps: obtaining shale gas well drilling cost related data, carrying out the preprocessing of the shale gas well drilling cost related data, and obtaining a drilling cost impact factor; constructing an LSTM model, and performing time sequence analysis on the drilling cost influence factor according to the LSTM model to obtain an influence factor prediction sample; and establishing a drilling cost prediction model by using an adaptive kernel density estimation method, and inputting the influence factor prediction sample into the drilling cost prediction model to obtain drilling cost prediction data. By combining specific drilling cost data, analyzing main factors influencing the drilling cost, and performing prediction and detailed research on the drilling cost based on the LSTM model and the drilling cost prediction model, the prediction precision is improved, the cost rule in the production process is obtained, a decision basis is provided for actual production, the production efficiency is improved, and the economic benefit is increased. And the production cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of drilling exploration, and in particular to a shale gas drilling cost prediction method and system. Background Art

[0002] Shale gas refers to natural gas extracted from shale layers, which is mainly found in dark mud shale or high-carbon mud shale in an adsorbed or free state. It is an unconventional natural gas. Due to the large-scale, high-investment and high-risk of drilling production, it requires a lot of manpower, material and financial resources. At the same time, there are many unknown and uncertain factors in the drilling process, and the risk of drilling production is also very high. The drilling operation will increase the drilling investment due to the influence of factors such as the complex underground conditions. At present, the investment risk of shale gas development in my country is relatively high. Cost is a key factor restricting the development of the shale gas industry. The cost-effectiveness estimation of a single shale gas well is the basis for the economic evaluation of shale gas block development, and it is also a necessary reference for shale gas investment decisions. It is of great significance to reduce the risk of shale gas development. Therefore, controlling drilling costs is crucial to reducing drilling project investment and improving economic benefits. Summary of the invention

[0003] The technical problem to be solved by the present invention is that the drilling process is complicated, resulting in high drilling costs, and it is impossible to evaluate the drilling costs. The purpose is to provide a shale gas drilling cost prediction method and system. By analyzing data characteristics, fitting a learning rate model, and combining industry information to solve research results, the shale gas well cost analysis research is preliminarily completed, the drilling costs are refined, and the cost laws in the production process are obtained, which provides a decision-making basis for actual production, improves production efficiency, and reduces production costs.

[0004] The present invention is achieved through the following technical solutions:

[0005] A first aspect of the present invention provides a shale gas drilling cost prediction method, comprising the following steps:

[0006] S1. Acquire data related to shale gas well drilling costs, pre-process the data related to shale gas well drilling costs, and obtain drilling cost influencing factors;

[0007] S2. Build an LSTM model, perform time series analysis on the factors affecting drilling costs based on the LSTM model, and obtain prediction samples of the factors affecting drilling costs;

[0008] S3. Use the adaptive kernel density estimation method to establish a drilling cost prediction model, input the influencing factor prediction samples into the drilling cost prediction model, and obtain the drilling cost prediction data.

[0009] The present invention analyzes the main factors affecting the drilling cost by combining specific drilling cost data, constructs an LSTM model to perform time series analysis on the factors affecting the drilling cost, and uses an adaptive kernel density estimation method to establish a drilling cost prediction model to predict the drilling cost, thereby improving the prediction accuracy. According to the historical drilling cost data, the future drilling cost data is obtained to conduct a detailed study on the drilling cost, and the cost law in the production process is obtained, which provides a decision-making basis for actual production, improves production efficiency, and reduces production costs.

[0010] Further, preprocessing the shale gas well drilling cost related data includes: determining the weight of each factor in the shale gas well related data according to the probability distribution, combining them to obtain a first factor group set, wherein the first factor group set includes a combination of multiple groups of factors that affect the drilling cost;

[0011] Checking the causal relationship between each factor group in the first factor group set and the drilling cost, and saving the factor group having a causal relationship with the drilling cost into the second factor group set;

[0012] For each factor group in the second factor group set, the drilling cost corresponding to each factor group in the first future time period is predicted according to the characteristic value of each factor in the factor group, and the drilling cost influencing factor is determined.

[0013] Furthermore, the time series analysis of the factors affecting drilling costs according to the LSTM model specifically includes:

[0014] Normalize the factors affecting drilling costs;

[0015] The time series data is passed through the convolution layer, and the result is the extracted features;

[0016] Put each dimension of features obtained into the neurons of LSTM to obtain the LSTM hidden layer output vector;

[0017] Use the attention mechanism to automatically focus on the LSTM hidden layer output vector, and automatically give larger weights to parameters that are significantly related to the target value;

[0018] Expand the feature vector processed by the attention mechanism into a one-dimensional vector to meet the requirements of the fully connected layer input;

[0019] The one-dimensional vector obtained by the expansion operation is input into the fully connected layer, and the fully connected layer outputs the impact factor prediction sample.

[0020] Furthermore, the S3 specifically includes:

[0021] The historical drilling cost is obtained as the dependent variable, and the influencing factors of the historical drilling cost are obtained as the independent variables to construct a vector;

[0022] Determine training samples and test samples based on the vector;

[0023] The drilling cost prediction model is established by using the adaptive kernel density estimation method, and the drilling cost prediction model is trained;

[0024] Performing model verification on the trained drilling cost prediction model;

[0025] The verified drilling cost prediction model is used in combination with the influencing factor sequence to predict the drilling cost.

[0026] Further, determining the training sample set according to the vector includes:

[0027] According to the historical data of drilling costs, the input variables and output variables are determined, and the training samples and test samples are constructed.

[0028] The cumulative number of wells, drilling cycle, and industrial producer purchase price index are selected as input variables and drilling cost as output variable.

[0029] Furthermore, the method of establishing a drilling cost prediction model by using an adaptive kernel density estimation method and training the drilling cost prediction model includes:

[0030] Normalization is used to normalize the sample data of training samples and test samples, eliminating the influence of different dimensions and units of various factors, and normalizing the input and output data of the samples so that the normalized values ​​are between [-1,1];

[0031] The kernel function is determined, and the drilling cost prediction model is established using the adaptive kernel density estimation method.

[0032] Furthermore, the determining of the kernel function specifically includes: determining an insensitive loss function coefficient, a penalty parameter and a Gaussian radial basis kernel function parameter, and the determining step includes:

[0033] Determine the range of the three parameters;

[0034] Determine the step size for each parameter change;

[0035] Form all combinations of three parameters;

[0036] Traverse all combinations and select the optimal parameters.

[0037] Furthermore, the training of the drilling cost prediction model specifically includes:

[0038] Set the initial values ​​of the network weights and thresholds, calculate the actual output of the network and the state of the hidden units, and calculate the network output error until the output error is less than the threshold or reaches the specified number of iterations. End the training, otherwise perform error back propagation;

[0039] Calculate the training error, modify the weights and thresholds according to the training error, and repeat the above steps until the training error is less than the set value.

[0040] A second aspect of the present invention provides a shale gas drilling cost prediction system, comprising an interface layer, a business layer and a data layer;

[0041] The data layer is used to obtain data related to shale gas well drilling costs, and the business layer is used to pre-process the data related to shale gas well drilling costs to obtain drilling cost influencing factors;

[0042] The business layer is used to construct an LSTM model, and perform time series analysis on the factors affecting drilling costs according to the LSTM model to obtain prediction samples of the factors affecting drilling costs;

[0043] The business layer is used to establish a drilling cost prediction model using an adaptive kernel density estimation method, and the interface layer is used to input influencing factor prediction samples into the drilling cost prediction model to obtain drilling cost prediction data.

[0044] Furthermore, the interface layer includes a model control unit and a data maintenance unit;

[0045] The business layer includes a data processing unit, a model building unit and a parameter optimization unit;

[0046] The data layer includes drilling cost related data units and model data units;

[0047] The drilling cost related data unit is used to obtain shale gas well drilling cost related data and input the data into the business layer, and the data processing unit is used to pre-process the shale gas well drilling cost related data to obtain drilling cost influencing factors;

[0048] The model building unit is used to build an LSTM model, and perform time series analysis on the factors affecting drilling costs according to the LSTM model to obtain prediction samples of the factors affecting drilling costs;

[0049] The model building unit is also used to establish a drilling cost prediction model using an adaptive kernel density estimation method. The model control unit and the data maintenance unit are used to input the influencing factor prediction samples into the drilling cost prediction model. The data processing unit is used to perform data processing on the influencing factor prediction samples input into the drilling cost prediction model to obtain drilling cost prediction data.

[0050] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0051] 1. Combined with the specific drilling cost data, the main factors affecting the drilling cost are analyzed, and the LSTM model is constructed to conduct time series analysis on the factors affecting the drilling cost. The drilling cost prediction model is established using the adaptive kernel density estimation method to predict the drilling cost, which improves the prediction accuracy. Based on the historical drilling cost data, the future drilling cost data is obtained to conduct a detailed study on the drilling cost, and the cost law in the production process is obtained, which provides a decision-making basis for actual production, improves production efficiency, and reduces production costs.

[0052] 2. Use the adaptive kernel density estimation method to establish a drilling cost prediction model, introduce the kernel function, and establish a support vector machine model to predict drilling costs by analyzing the factors affecting drilling costs. Introducing more drilling cost variables in the probability model can improve the accuracy of the prediction results. On the other hand, after certain specific events have occurred, conditional probabilities can be obtained to better quantify predictability. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] 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:

[0054] Figure 1 is a flow chart of a method in an embodiment of the present invention;

[0055] Figure 2 is the total cost and average cost of the shale gas well in the embodiment of the present invention;

[0056] Figure 3 is the cumulative average cost and drilling day cost of the shale gas well in the embodiment of the present invention;

[0057] Figure 4 is the cumulative average cost of shale gas wells and the drilling cost factor in the embodiment of the present invention;

[0058] Figure 5 This is the learning curve of drilling cost factors from 2015 to 2017 in the embodiment of the present invention. DETAILED DESCRIPTION

[0059] 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.

[0060] Example 1

[0061] like Figure 1 As shown, this embodiment provides a shale gas drilling cost prediction method, comprising the following steps:

[0062] S1. Acquire data related to shale gas well drilling costs, pre-process the data related to shale gas well drilling costs, and obtain drilling cost influencing factors;

[0063] S2. Build an LSTM model, perform time series analysis on the factors affecting drilling costs based on the LSTM model, and obtain prediction samples of the factors affecting drilling costs;

[0064] S3. Use the adaptive kernel density estimation method to establish a drilling cost prediction model, input the influencing factor prediction samples into the drilling cost prediction model, and obtain the drilling cost prediction data.

[0065] By combining specific drilling cost data, the main factors affecting drilling costs are analyzed, and an LSTM model is built to perform time series analysis on the factors affecting drilling costs. The drilling cost prediction model is established using the adaptive kernel density estimation method to predict drilling costs, thereby improving the prediction accuracy. Based on the historical drilling cost data, the future drilling cost data is obtained to conduct a detailed study on the drilling cost, and the cost law in the production process is obtained, which provides a decision-making basis for actual production, improves production efficiency, and reduces production costs.

[0066] In some possible embodiments, preprocessing the shale gas well drilling cost related data includes: determining the weight of each factor in the shale gas well related data according to the probability distribution, combining them to obtain a first factor group set, the first factor group set including a combination of multiple groups of factors that affect the drilling cost;

[0067] Test the causal relationship between each factor group in the first factor group set and the drilling cost, and save the factor group that has a causal relationship with the drilling cost into the second factor group set;

[0068] For each factor group in the second factor group set, the corresponding drilling cost of each factor group in the first future time period is predicted according to the characteristic value of each factor in the factor group, and the drilling cost influencing factor is determined.

[0069] In some possible embodiments, the shale gas well drilling cost-related data specifically include: cumulative number of wells, cumulative total cost, average cost, drilling cycle, industrial producer purchase price index (previous year = 100), industrial producer purchase price index (1990 = 100), and the industrial producer purchase price index is used to eliminate price impacts.

[0070] It also includes the annual settlement table of shale gas wells from 2015 to 2020: 228 data related to 228 shale gas wells from 2015 to 2020, including 63 indicators, provided by the Natural Gas Research Institute;

[0071] Industrial Producer Purchase Price Index: Industrial Producer Purchase Price Index (previous year=100) and Industrial Producer Purchase Price Index (1990=100) are downloaded from the National Bureau of Statistics website.

[0072] The factors that affect drilling costs include macroeconomic factors, technical factors, policy factors and term structure factors, including: cumulative number of wells, cumulative total cost, average cost, drilling cycle, industrial producer purchase price index (previous year = 100), industrial producer purchase price index (1990 = 100), and the industrial producer purchase price index is used to eliminate the impact of prices. The impact of cumulative total cost includes: financial market data, bank credit data, financial terminal data, government website information and news media information. As shown in Table 1:

[0073] Table 1 Annual data of shale gas wells

[0074]

[0075] Among them, the visualization shows the number of wells calculated by year, the total cost (billion) and the average cost of shale gas wells, such as Figure 2 As shown in Figure 2, the average annual cost of shale gas wells has not decreased year by year.

[0076] Based on the basic idea of ​​the learning curve method (with the increase of production and the increase of operator proficiency, the product cost will decrease, and the cost reduction is gradually accumulated rather than annual average. Observation of data found that the drilling cycle varies greatly, and the daily average cost is introduced.

[0077] Based on the accumulation idea and unit cost, the cumulative number of wells, cumulative total cost, cumulative average cost, drilling cycle, and drilling day cost are calculated, as shown in Table 2:

[0078] Table 2 Annual data of shale gas wells

[0079]

[0080] Among them, the cumulative average cost and drilling day cost are visualized, such as Figure 3 As shown in the data, the daily cost of shale gas well drilling has been in line with the downward trend year by year, but there was a turning point in 2018.

[0081] The introduction of the industrial producer purchase price index specifically includes: after verification, it was found that there was a financial crisis in 2018, and it was necessary to introduce a price factor to try to eliminate the price impact. Shale gas wells belong to the industrial field, so the "industrial producer purchase price index (1990 = 100)" was introduced, and the price index used the annual average price index, the drilling cost factor (drilling day cost / price index * 100), and the calculation results are shown in Table 3:

[0082] Table 3 Cumulative average cost of shale gas wells and drilling cost factors

[0083]

[0084] Visualize the cumulative average cost and drilling cost factors, such as Figure 4 as shown.

[0085] like Figure 5 As shown in the figure, the drilling cost factor decreased year by year from 2015 to 2017, which is consistent with the learning curve. The fitting results are as follows: Y = a*exp(-x / t), a = 4.927, t = 2.946. The model accuracy is very high, and it can be concluded that:

[0086] The drilling cost factor decreased year by year from 2015 to 2017, and it was more in line with the learning rate curve;

[0087] The 2018 financial crisis caused a shock to the cost reduction trend;

[0088] 2019-2020 saw a gradual decline in the new balance;

[0089] The annual data volume is too small and the data set needs to be expanded for verification.

[0090] First, build the LSTM model. The construction steps include:

[0091] 1. Recording of past information: The current input and the output of the previous moment determine how much past information the current memory unit needs to store: f t =σ(W f ·[h t-1 ,x t ]+b f )The activation function is often a sigmoid function, which maps to a number between 0 and 1, where 1 means that all past information is recorded and 0 means that all past information is discarded.

[0092] 2. Recording of current information: The current input and the output of the previous moment determine what new information is stored in the state unit. This includes two parts: how much information the current memory unit needs to store, and how much information the current information represents. t =σ(W i ·[ht-1 ,x t ]+b i ), The activation function is the sigmoid function, which maps to a number between 0 and 1, where 1 means that all current information is recorded and 0 means that all current information is discarded. It is the representation of the current information and is a candidate for the current state, obtained by mapping with the activation function tanh.

[0093] 3. Current state: Get the current state from current information and past information In this step, the sum of two weights is used to determine how much previous information and current information to retain. If it is larger, the subsequent model will use more information from more distant times. If it is larger, the subsequent model will use more information from more recent times, thereby avoiding the forgetting of information from more distant time points in the ordinary RNN network structure.

[0094] 4. Output result: Based on the output at the previous moment, the input at the current moment and the current state, the current output is obtained: o t =σ(W o ·[h t-1 ,x t ]+b o )h t =o t *tanh(C t )

[0095] The time series analysis of the factors affecting drilling costs is carried out based on the LSTM model, including:

[0096] The factors affecting drilling costs are normalized; normalization is helpful to avoid the magnitude differences between factors and eliminate the impact of factors due to different dimensions, and to normalize the input and output data of the samples.

[0097] The time series data is passed through the convolution layer, and the result is the extracted features;

[0098] Put each dimension of features obtained into the neurons of LSTM to obtain the LSTM hidden layer output vector;

[0099] Use the attention mechanism to automatically focus on the LSTM hidden layer output vector, and automatically give larger weights to parameters that are significantly related to the target value;

[0100] Expand the feature vector processed by the attention mechanism into a one-dimensional vector to meet the requirements of the fully connected layer input;

[0101] The one-dimensional vector obtained by the expansion operation is input into the fully connected layer, and the fully connected layer outputs the impact factor prediction sample.

[0102] Among them, in the forward calculation process of LSTM, the operation of each memory cell is controlled by the input gate, forget gate, and output gate. Each gate operates on the current input data respectively, and decides whether to pass the operation result downward according to the judgment logic of the activation function. In the current state, the memory cell receives the input vector xt and the hidden state ht-1 of the cell at the previous moment. The input gate it multiplies the newly generated information and superimposes it with the cell state ft·Ct-1 processed by the forget gate ft to generate the cell state Ct. After the output gate operation and control, the retained information ot of the current cell state and the output ht of the hidden state are generated. The reverse error propagation process of LSTM is similar to BPNN. By calculating the error of each cell, it is back-propagated in two directions of time and network level, the weight gradient is updated, the connection weight is optimized, and the network accuracy is improved.

[0103] The drilling cost prediction model is established by using the adaptive kernel density estimation method. The influencing factor sequence is input into the drilling cost prediction model to obtain the drilling cost prediction data, including:

[0104] The historical drilling cost is obtained as the dependent variable, and the influencing factors of the historical drilling cost are obtained as the independent variables to construct a vector;

[0105] Determine the training sample and the test sample according to the vector. Determine the training sample set according to the vector, including: determine the input variable and the output variable according to the historical drilling cost data, construct the training sample and the test sample, wherein,

[0106] The cumulative number of wells, drilling cycle, and industrial producer purchase price index are selected as input variables and drilling cost as output variable.

[0107] The drilling cost prediction model is established by using the adaptive kernel density estimation method, and the drilling cost prediction model is trained;

[0108] Set the initial values ​​of the network weights and thresholds, calculate the actual output of the network and the state of the hidden units, and calculate the network output error until the output error is less than the threshold or reaches the specified number of iterations. End the training, otherwise perform error back propagation;

[0109] Calculate the training error, correct the weights and thresholds according to the training error, and repeat the above steps.

[0110] Conduct model validation on the trained drilling cost prediction model;

[0111] The verified drilling cost prediction model is used in combination with the influencing factor sequence to predict the drilling cost.

[0112] The drilling cost prediction model is established by using the adaptive kernel density estimation method, and the drilling cost prediction model is trained, including: using normalization processing to normalize the sample data of training samples and test samples, eliminating the influence of different dimensions and units of various factors, and normalizing the input and output data of the samples so that the normalized values ​​are between [-1,1].

[0113] Determine the kernel function, use the adaptive kernel density estimation method to establish the drilling cost prediction model, determine the insensitive loss function coefficient, penalty parameter and Gaussian radial basis kernel function parameter, and the determination steps include: determining the range of the three parameters; determining the step size of each parameter change; forming all combinations of the three parameters; traversing all combinations to select the optimal parameters.

[0114] The steps of establishing a drilling cost prediction model using the adaptive kernel density estimation method include:

[0115]

[0116] Among them, {x (j ):j=1,2,3,...,M} represents the M sample data for constructing the density function, n represents the dimension of the variable space, w represents the number of windows, k(x) represents the kernel probability density function, specifically, Where S is the sample set {x (j)}, In the formula Describe the mean of M sample data.

[0117] Fixed the wide kernel density function of the adaptive kernel density estimation and modified the window width parameter to wλ j , which is in the form of:

[0118]

[0119] Among them, λ j is the local bandwidth factor, w is the window width parameter

[0120]

[0121] Where f(x) represents the joint probability density of sample x, 0≤α<1 is the sensitivity factor, usually α=0.5, M d is the number of different samples in M ​​sample data (M d ≤M).

[0122] Example 2

[0123] This embodiment provides a shale gas drilling cost prediction system, including an interface layer, a business layer and a data layer;

[0124] The interface layer includes,model control unit and data maintenance unit;

[0125] The business layer includes data processing unit, model building unit and parameter optimization unit;

[0126] The data layer includes drilling cost related data units and model data units;

[0127] The drilling cost related data unit is used to obtain the shale gas well drilling cost related data and input the data into the business layer. The data processing unit is used to pre-process the shale gas well drilling cost related data and obtain the drilling cost influencing factors.

[0128] The model building unit is used to build an LSTM model, and perform time series analysis on the factors affecting drilling costs according to the LSTM model to obtain prediction samples of the factors affecting drilling costs;

[0129] The model building unit is also used to establish a drilling cost prediction model using an adaptive kernel density estimation method. The model control unit and the data maintenance unit are used to input the influencing factor prediction samples into the drilling cost prediction model. The data processing unit is used to perform data processing on the influencing factor prediction samples input into the drilling cost prediction model to obtain drilling cost prediction data.

[0130] The data layer is used to add data, modify data, delete data, and query data;

[0131] Data addition enables users to add sample data. Users can manually enter the basic information of drilling cost data, or import the data into the system by saving it in .

[0132] Data modification enables users to modify sample data. Sample data does not need to be modified, but sometimes the data reported by the drilling team may have individual errors, requiring the administrator to modify the data.

[0133] Data query, to realize data query, you can find the data that meets the conditions according to the block, well type, technology, well number, etc. Users can choose to output it to the table.

[0134] Data statistics, to realize various statistical analyses of drilling cost data, such as total cost in a certain period of time, number of wells with average cost, and total number of wells.

[0135] The business layer is used to create new models, modify models, delete models, automatically train models, and predict drilling costs.

[0136] Create a new model and establish a new prediction model. After the new block is added, you need to establish the model and set the model. The required sample data range, sample change number and other basic information are required to facilitate users to use the model to predict the data.

[0137] Modify the model to enable users to modify the model file, such as modifying the change range of sample data, the number of changes in sample data, and other information.

[0138] Model viewing allows users to view the model and understand model information.

[0139] Automatic model training enables training of new models, finding the optimal parameter combination, establishing a support vector machine regression model and automatically saving the optimal parameters and training results. At the same time, it can automatically modify the model according to the data change conditions set by the user to automatically find the optimal parameters.

[0140] Drilling cost prediction enables users to predict drilling costs. Users manually or in batches input drilling cost influencing factors, and the system can automatically output the predicted value. If there is actual cost, the prediction results can be statistically analyzed.

[0141] User authority management enables different users to control the use of the system. Administrators can modify drilling cost data and model data, while other users can only use the system for cost forecasting.

[0142] 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 drilling cost prediction method, characterized in that: The following steps are involved: S1. Acquire data related to shale gas well drilling costs, pre-process the data related to shale gas well drilling costs, and obtain drilling cost influencing factors; S2. Build an LSTM model, perform time series analysis on the factors affecting drilling costs based on the LSTM model, and obtain prediction samples of the factors affecting drilling costs; S3. Use the adaptive kernel density estimation method to establish a drilling cost prediction model, input the influencing factor prediction samples into the drilling cost prediction model, and obtain the drilling cost prediction data.

2. The shale gas drilling cost prediction method according to claim 1, characterized in that: Preprocessing the shale gas well drilling cost related data includes: determining the weight of each factor in the shale gas well related data according to probability distribution, combining them to obtain a first factor group set, wherein the first factor group set includes a combination of multiple groups of factors that affect the drilling cost; Checking the causal relationship between each factor group in the first factor group set and the drilling cost, and saving the factor group having a causal relationship with the drilling cost into the second factor group set; For each factor group in the second factor group set, the drilling cost corresponding to each factor group in the first future time period is predicted according to the characteristic value of each factor in the factor group, and the drilling cost influencing factor is determined.

3. The shale gas drilling cost prediction method according to claim 1, characterized in that: The time series analysis of the factors affecting drilling costs according to the LSTM model specifically includes: Normalize the factors affecting drilling costs; The time series data is passed through the convolution layer, and the result is the extracted features; Put each dimension of features obtained into the neurons of LSTM to obtain the LSTM hidden layer output vector; Use the attention mechanism to automatically focus on the LSTM hidden layer output vector, and automatically give larger weights to parameters that are significantly related to the target value; Expand the feature vector processed by the attention mechanism into a one-dimensional vector to meet the requirements of the fully connected layer input; The one-dimensional vector obtained by the expansion operation is input into the fully connected layer, and the fully connected layer outputs the impact factor prediction sample.

4. The shale gas drilling cost prediction method according to claim 1, characterized in that: The S3 specifically includes: The historical drilling cost is obtained as the dependent variable, and the influencing factors of the historical drilling cost are obtained as the independent variables to construct a vector; Determine training samples and test samples based on the vector; The drilling cost prediction model is established by using the adaptive kernel density estimation method, and the drilling cost prediction model is trained; Performing model verification on the trained drilling cost prediction model; The verified drilling cost prediction model is used in combination with the influencing factor sequence to predict the drilling cost.

5. The shale gas drilling cost prediction method according to claim 4, characterized in that: Determining the training sample set according to the vector includes: According to the historical data of drilling costs, the input variables and output variables are determined, and the training samples and test samples are constructed. The cumulative number of wells, drilling cycle, and industrial producer purchase price index are selected as input variables and drilling cost as output variable.

6. The shale gas drilling cost prediction method according to claim 4, characterized in that: The use of the adaptive kernel density estimation method to establish a drilling cost prediction model specifically includes: Normalization is used to normalize the sample data of training samples and test samples, eliminating the influence of different dimensions and units of various factors, and normalizing the input and output data of the samples so that the normalized values ​​are between [-1,1]; The kernel function is determined, and the drilling cost prediction model is established using the adaptive kernel density estimation method.

7. The shale gas drilling cost prediction method according to claim 6, characterized in that: Determining the kernel function specifically includes: determining an insensitive loss function coefficient, a penalty parameter, and a Gaussian radial basis kernel function parameter, and the determining step includes: Determine the range of the three parameters; Determine the step size for each parameter change; Form all combinations of three parameters; Traverse all combinations and select the optimal parameters.

8. The shale gas drilling cost prediction method according to claim 4, characterized in that: The training of the drilling cost prediction model specifically includes: Set the initial values ​​of the network weights and thresholds, calculate the actual output of the network and the state of the hidden units, and calculate the network output error until the output error is less than the threshold or reaches the specified number of iterations. End the training, otherwise perform error back propagation; Calculate the training error, modify the weights and thresholds according to the training error, and repeat the above steps until the training error is less than the set value.

9. A shale gas drilling cost prediction system, characterized in that: Includes interface layer, business layer and data layer; The data layer is used to obtain data related to shale gas well drilling costs, and the business layer is used to pre-process the data related to shale gas well drilling costs to obtain drilling cost influencing factors; The business layer is used to construct an LSTM model, and perform time series analysis on the factors affecting drilling costs according to the LSTM model to obtain prediction samples of the factors affecting drilling costs; The business layer is used to establish a drilling cost prediction model using an adaptive kernel density estimation method, and the interface layer is used to input influencing factor prediction samples into the drilling cost prediction model to obtain drilling cost prediction data.

10. The shale gas drilling cost prediction system according to claim 9, characterized in that: The interface layer includes a model control unit and a data maintenance unit; The business layer includes a data processing unit, a model building unit and a parameter optimization unit; The data layer includes drilling cost related data units and model data units; The drilling cost related data unit is used to obtain shale gas well drilling cost related data and input the data into the business layer, and the data processing unit is used to pre-process the shale gas well drilling cost related data to obtain drilling cost influencing factors; The model building unit is used to build an LSTM model, and perform time series analysis on the factors affecting drilling costs according to the LSTM model to obtain prediction samples of the factors affecting drilling costs; The model building unit is also used to establish a drilling cost prediction model using an adaptive kernel density estimation method. The model control unit and the data maintenance unit are used to input the influencing factor prediction samples into the drilling cost prediction model. The data processing unit is used to perform data processing on the influencing factor prediction samples input into the drilling cost prediction model to obtain drilling cost prediction data.

Images