An intelligent simulation method for synchronous expansion of multi-cluster fractures in large-segment horizontal wells

Abstract

The present invention discloses an intelligent simulation method for synchronous expansion of multi-cluster hydraulic fracturing cracks in a large section of a horizontal well, and relates to the technical field of numerical simulation of oil reservoirs. The method comprises the following steps: firstly, a prediction model of the rupture pressure is established, and the rupture pressure of the target well is predicted, thereby obtaining the total number of clusters of fractures initiated during the fracturing process; a simulation model for synchronous expansion of multi-cluster hydraulic fracturing cracks is established, and the model is discretized to obtain a discretized model; based on an adaptive elite genetic algorithm, the flow distribution results of the position of the crack mouth of each cluster under different time conditions are first obtained, and then the three dimensions of each cluster of cracks are predicted. The method of the present invention can grasp the expansion of multi-cluster hydraulic fracturing cracks in real time, provide guidance for the optimization of the hydraulic fracturing process and the adjustment of the construction plan, and effectively improve the recovery rate of horizontal wells.

Description

Technical Field

[0001] The invention relates to the technical field of numerical simulation of oil reservoirs, and in particular to an intelligent simulation method for synchronous expansion of large-segment multi-cluster fracturing cracks in horizontal wells. Background Art

[0002] Staged fracturing of horizontal wells is an important technical means to achieve the efficient development of unconventional oil and gas. With the advancement of technology, large-scale multi-cluster hydraulic fracturing with more than 10 clusters in a single stage can be achieved, which greatly improves the construction efficiency. However, the development case of tight conglomerate reservoirs in the Mahu area shows that the production effect after large-scale multi-cluster fracturing of horizontal wells is not always ideal. The fundamental reason is the low cluster utilization rate and poor fracture effectiveness. Therefore, evaluating and mastering the expansion of multi-cluster fractures is the key to analyzing and solving these problems. Microseismic monitoring technology can directly obtain fracture information of horizontal wells, but this technology can only be implemented after fracturing, and it is impossible to evaluate the expansion of hydraulic fractures during the fracturing process, and it will also increase costs. Carrying out a simulation study on the synchronous expansion of fractures in large-scale multi-cluster fracturing of horizontal wells can grasp the expansion of multi-cluster fracturing fractures in real time, provide effective guidance for fracturing design optimization and construction parameter adjustment, and thus improve the cluster utilization rate and fracture effectiveness of horizontal wells.

[0003] Existing research methods include two categories: hydraulic fracturing physical simulation experiments and numerical simulations. Among them, physical simulation experiments rely on complex equipment and have strict requirements on the size of experimental samples. Rock samples are usually prepared by artificial synthesis, and there is a big difference between the experimental environment and the real complex formation environment. The experimental results need to be verified on site. Numerical simulations are mostly based on finite element (FEM), discrete element (DEM) and other methods, which is a more economical and effective means. However, most of the existing studies propose optimization directions for segmented clustering and perforation processes based on assumed working conditions. Not only is there a lack of verification with actual working conditions on site, but the calculation process is also complicated, especially when it comes to large-scale multi-cluster fracture simulations, and the solution efficiency needs to be improved urgently. Summary of the invention

[0004] To solve at least one of the above problems, the present invention proposes an intelligent simulation method for synchronous expansion of large-section multi-cluster hydraulic fracturing cracks in horizontal wells, which can quickly simulate the expansion of multi-cluster hydraulic fracturing cracks based on actual working condition data.

[0005] The technical solution of the present invention is: an intelligent simulation method for synchronous expansion of multi-cluster hydraulic fracturing cracks in a large section of a horizontal well, comprising the following steps:

[0006] S1. Obtain the well logging data of the developed horizontal wells in the target block and the fracture pressure of the segmented and clustered fracturing, and match the well logging data with the fracture pressure;

[0007] S2. Using logging data as input and fracture pressure as output, a fracture pressure prediction model is established based on gated recurrent units and attention mechanism, and the logging data and fracture pressure prediction model of the developed horizontal wells in the current block are used for training;

[0008] S3. Based on the continuity equation, pressure drop equation, crack width equation, and crack height equation, combined with the ground stress data of the target block, a multi-cluster hydraulic fracture synchronous expansion simulation model is established, and discretized to obtain a discretized model. The coupling conditions of the model are established according to the volume balance and pressure balance. The discretized model is as follows:

[0009]

[0010] The coupling conditions are:

[0011] Volume Balance Equation , the pressure balance equation ; Wherein, subscripts j and k are the crack cluster numbers, and j, k = 1, 2, …, m; P j is the flow pressure in fracture cluster j; t and l are the time step and length step respectively; l j is the length of crack cluster j; t is time, T is the total construction time; h j is the crack height of crack cluster j; L f,j is the total length of the jth cluster of cracks; Q represents the construction displacement; represents the bottom hole pressure calculated from the opening of the j-th cluster of fractures to the reference point O; Indicates that the fracturing fluid flows from the reference point O to the k Pipe friction of cluster fractures; P b,j represents the bottom hole pressure at the jth cluster perforation position; P p,j represents the hole friction resistance at the jth cluster perforation position; f 1= , h j ( l j , t )yes t Time next j Cluster cracks l j The seam height at C L is the comprehensive filtration coefficient; t 0( l j ) is the time when the fracturing fluid reaches the crack lj The time required for , w j ( l j , t )yes t Time next j Cluster cracks l j The width of the seam at f 2= ; yes t Time next j Cluster cracks l j The 2n'+1 power of the slit width at n’ is the fracturing fluid flow index; f 3= ; E is the elastic modulus; v is Poisson's ratio; s h ( l j ) is the j Cluster crack extension l j Minimum principal stress at the length; P j ( l , t )yes t Time next j Cluster cracks l j The pressure of

[0012] S4, obtaining the logging parameters and bottom hole pressure during the fracturing construction of the target well, predicting the fracture pressure of the target well based on the trained prediction model, and obtaining the total number of fracture initiation clusters during the fracturing process of the target well in combination with the pressure balance equation;

[0013] S5. Based on the adaptive elite genetic algorithm, a fitness function is established with the minimum error of the bottom hole pressure from the opening position of each cluster of fractures in the target well to the reference point O. The volume balance equation is used as a constraint and the flow distribution of the opening position of each cluster of fractures is used as a goal. The flow distribution results of the opening position of each cluster of fractures under different time conditions are obtained by solving the problem.

[0014] S6. Based on the adaptive elite genetic algorithm, the fitness function is established with the absolute minimum error of the flow pressure at the fracture mouth of each cluster of fractures and the bottom hole pressure at the perforation position of each cluster. The discretized model is used to solve the flow pressure, flow rate, fracture width and fracture height from the end of each cluster of fractures to the fracture mouth. The total length of each cluster of fractures is taken as the target, and the length of each cluster of fractures under different time conditions is solved, and then the three-dimensional size of each cluster of fractures under different time conditions is obtained.

[0015] Beneficial effects: The method of the present invention can grasp the expansion of multiple clusters of fracturing cracks in real time, provide guidance for fracturing process optimization and construction plan adjustment, and effectively improve the recovery rate of horizontal wells. DETAILED DESCRIPTION

[0016] The specific implementation modes of the present invention will be described clearly and completely below with reference to examples. Obviously, the examples described are only some embodiments of the present invention, rather than all embodiments.

[0017] An intelligent simulation method for synchronous expansion of multi-cluster hydraulic fracturing cracks in a horizontal well comprises the following steps:

[0018] S1. Obtain the well logging data of the developed horizontal wells in the target block and the fracture pressure of the segmented and clustered fracturing, and match the well logging data with the fracture pressure;

[0019] Specifically, in this step, the selected logging data are depth, density, resistivity, natural gamma, natural potential, acoustic time difference, and compensated neutron.

[0020] Meanwhile, in this embodiment, the method for obtaining the bursting pressure includes the following steps:

[0021] Obtain segmented cluster data and fracturing construction data for developing horizontal wells in target blocks, and calculate the bottom hole pressure of each cluster perforation position in each well section; the segmented cluster data include the number of sections, section length, number of clusters, cluster spacing, and cluster depth; the fracturing construction data include wellhead casing pressure, displacement, and sand ratio;

[0022] Based on the fracturing operation data, the first peak on the bottom hole pressure curve when the displacement is in an increasing or constant state is recorded. This peak corresponds to the fracture pressure of the cluster with the lowest minimum principal stress.

[0023] The method for obtaining the bottom hole pressure is as follows: Under the condition that the total number of fracturing clusters is n and the total fracturing time is T, for t (t=1,2,…, T ) Clusters under the conditions of time i ( i =1,2,…, n ), the construction pressure can be obtained through fracturing construction data P 0( t ) ; Calculate the liquid column pressure, pipe friction and hole friction: P l ( t )=10 -6 × r ( t ) gH , , , whereP l(t) is the pressure of the liquid column at time t; p(t) is the density of the fracturing fluid in the wellbore at time t; g is the acceleration due to gravity; H is the height of the liquid column; P f,i ( t )yes t At this moment, the fracturing fluid flows to the i Pipe string friction at the cluster perforation location; f is the friction coefficient; u ( t )yes t The flow rate of fracturing fluid in the wellbore at the moment; L is the distance that the fracturing fluid flows from the wellhead to the perforation location of the cluster; D is the pipe diameter; P p,i ( t )yes t At this moment, the fracturing fluid flows to the i The hole friction of cluster holes, MPa; Q i ( t )yes t Time next i Flow rate of cluster holes, m 3 / min; N is the number of perforations; d is the aperture, m; C is the perforation flow coefficient; at the same time, when no cracks occur in the i-th cluster, its perforation flow is 0 and the perforation friction is 0.

[0024] Calculate the bottom hole pressure based on the above results: P b,i ( t )= P 0( t )+ P l ( t )- P f,i ( t )- P p,i ( t ). P b,i ( t )yes t Time next i Bottom hole pressure at cluster perforation location.

[0025] After calculating the bottom hole pressure, we make a curve of the bottom hole pressure in chronological order and judge the fracture pressure based on the curve.

[0026] When the fracture pressure of the cluster is obtained, the logging data of the cluster needs to be matched with the fracture pressure to obtain a complete data matrix.

[0027] S2. Taking logging data as input and fracture pressure as output, a fracture pressure prediction model is established based on gated recurrent units and attention mechanism. The logging data and fracture pressure prediction model of the developed horizontal wells in the current block are used for training. The trained model can be used to predict the fracture pressure of the remaining wells in the same block.

[0028] Based on the prior art, those skilled in the art can perform operations according to the description of this step, which will not be described in detail here.

[0029] S3. Based on the continuity equation, pressure drop equation, fracture width equation and fracture height equation, a simulation model for synchronous expansion of multiple clusters of hydraulic fractures is established, and discretized to obtain a discretized model. The coupling conditions of the model are established according to volume balance and pressure balance. The discretized model is as follows:

[0030]

[0031] The coupling conditions are:

[0032] Volume Balance Equation , the pressure balance equation ; Wherein, subscripts j and k are the crack cluster numbers, and j, k = 1, 2, …, m; P j is the flow pressure in fracture cluster j; t and l are the time step and length step respectively; l j is the length of crack cluster j; t is time, T is the total construction time; h j is the crack height of crack cluster j; L f,j is the total length of the jth cluster of cracks; Q represents the construction displacement; represents the bottom hole pressure calculated from the opening of the j-th cluster of fractures to the reference point O; Indicates that the fracturing fluid flows from the reference point O to the k Pipe friction of cluster fractures; P b,j represents the bottom hole pressure at the jth cluster perforation position; P p,j represents the hole friction resistance of the j-th cluster perforation position; f 1= , h j ( l j , t )yes t Time nextj Cluster cracks l j The seam height at C L is the comprehensive filtration coefficient; t 0( l j ) is the time when the fracturing fluid reaches the crack l j The time required for , w j ( l j , t )yes t Time next j Cluster cracks l j The width of the seam at f 2= ; yes t Time next j Cluster cracks l j The 2n'+1 power of the slit width at n’ is the fracturing fluid flow index; f 3= ; E is the elastic modulus; v is Poisson's ratio; s h ( l j ) is the j Cluster crack extension l j Minimum principal stress at the length; P j ( l , t )yes t Time next j Cluster cracks l j The pressure of

[0033] The derivation process of the above model is as follows:

[0034] During the fracturing operation, m clusters of fractures are formed in the target well section. For any cluster of fractures j, j =1,2,…, m , the continuity equation can be obtained from the mass conservation law: , where q j ( l , t )yes t Time next j Cluster cracksl j The flow rate at h j ( l j , t )yes t Time next j Cluster cracks l j The seam height at C L is the comprehensive filtration coefficient; t It is the time of fracturing; t 0( l j ) is the time when the fracturing fluid reaches the crack l j The time required for A j ( l j , t )yes t Time next j Cluster cracks l j The cross-sectional area at . , w j ( l j , t )yes t Time next j Cluster cracks l j The width of the seam.

[0035] The pressure drop equation for fluid flow inside the fracture is: , where P j ( l , t )yes t Time next j Cluster cracks l j Pressure at, MPa; n’ is the fracturing fluid flow index.

[0036] The crack width equation is obtained according to the England and Green formula, and the crack width at any position is obtained by combining the ground stress of the target block: , where v is Poisson's ratio; E is the elastic modulus; s h ( l j ) is the j Cluster crack extension lj Minimum principal stress at the length.

[0037] The crack height equation is obtained based on the linear elastic fracture mechanics theory, and the crack height at any position of each cluster of cracks is calculated in combination with the ground stress data of the target block: , where

[0038] Then, the crack width equation is substituted into the continuity equation, pressure drop equation and crack height equation, and the initial conditions and boundary conditions are given to obtain the model:

[0039]

[0040] In the formula, f 1 is the equation on the right side of the continuity equation; f 2 is the equation on the right side of the pressure drop equation; f 3 is the crack height equation, the right side of the crack length l j Find the formula for partial derivatives; L f,j It is j Total length of cluster cracks; s h ( L f,j ) is the j Minimum principal stress at the end of a cluster crack.

[0041] Then, according to the time step △t and space step △ l The above model is discretized, and a discretized model is obtained by combining the volume balance equation and the pressure balance equation.

[0042] S4, obtaining the logging parameters and bottom hole pressure during the fracturing construction of the target well, predicting the fracture pressure of the target well based on the trained prediction model, and obtaining the total number of fracture initiation clusters during the fracturing process of the target well in combination with the pressure balance equation; this step includes the following sub-steps:

[0043] S41. predicting the fracture pressure of each cluster using the trained prediction model according to the logging parameters of each cluster of the target well at time t;

[0044] S42. According to the fracturing construction data of the target well at time t, the bottom hole pressure of each cluster perforation position is obtained. When the bottom hole pressure of a cluster is not less than the predicted value of the fracture pressure of the cluster, the cluster is fractured;

[0045] S43. recalculating the bottom hole pressure of each cluster that has not been fractured according to the pressure balance equation, and comparing it with the predicted fracture pressure value to determine whether fracture has occurred in each cluster;

[0046] S44. Repeat S43 until no new clusters start to crack;

[0047] S45. Repeat S41 to S44 until the entire fracturing time T is completed.

[0048] S5. Based on the adaptive elite genetic algorithm, a fitness function is established with the minimum error of the bottom hole pressure from the opening position of each cluster of fractures in the target well to the reference point O, and the flow distribution of the opening position of each cluster of fractures is taken as the target to solve the flow distribution results of the opening position of each cluster of fractures under different time conditions; this step includes the following sub-steps:

[0049] S51. Based on the adaptive elite genetic algorithm, the volume balance equation is used as a constraint, and the initial population of flow distribution at the location of each cluster of crack openings is randomly generated;

[0050] S52. Establish a fitness function based on the minimum error between the bottom hole pressure of each cluster fracture opening and the reference point O: , where f Q is the first fitness value, indicating the bottom hole pressure error from the fracture position to the reference point O; m t is the total number of cracking clusters; x and y represent m t The xth and yth clusters in the cluster.

[0051] S53. Calculate the first fitness value of the initial population, and according to the elite retention strategy, save the individual with the smallest fitness value in the current population as the elite individual and do not perform inheritance;

[0052] S54. Take the non-elite individuals in the initial population, perform crossover and mutation operations on them to obtain a new population;

[0053] S55. Repeat S53 and S54 until the evolutionary generation is reached, and the final result is the flow distribution result of the crack opening position of each cluster at time t;

[0054] S56. Repeat S51-S55 to obtain the flow distribution results of the crack opening positions of each cluster under different time conditions.

[0055] In this step, the "flow distribution" refers to the individual flow of the fracturing fluid distributed to each cluster of fractures. When the fracturing fluid enters the fracture, the specific amount of flow distributed to each fracture is different, which is related to factors such as the length and width of the fracture, and can reflect the condition of the fracture to a certain extent. Among them, the obtained flow distribution result can not only serve as a necessary condition in S6, but also can be used for auxiliary verification by volume balance: the sum of the flow in each cluster of fractures should be equal to the flow distribution result.

[0056] S6. Based on the adaptive elite genetic algorithm, the fitness function is established with the absolute minimum error of the flow pressure at the opening of each cluster of fractures and the bottom hole pressure at the perforation position of each cluster. The discretized model is used to solve the flow pressure, flow rate, fracture width and fracture height from the end of each cluster of fractures to the opening of the fracture. The total length of each cluster of fractures is taken as the target to solve the length of each cluster of fractures under different time conditions, and then the three-dimensional size of each cluster of fractures under different time conditions is obtained. This step includes the following sub-steps:

[0057] S61. Based on the adaptive elite genetic algorithm, the initial population of the total length of each cluster of cracks is randomly generated;

[0058] S62. According to the total length of each cluster of cracks, l is the step length, and using the discretized model, starting from the end of each cluster of fractures, the flow pressure, flow rate, fracture width and fracture height are calculated step by step until the fracture opening;

[0059] S63. Establish a fitness function by minimizing the absolute difference between the flow pressure at the opening of each cluster of fractures and the bottom hole pressure at the perforation position: , where f f is the second fitness value, which represents the absolute difference between the flow pressure at the fracture position and the bottom hole pressure at the perforation position; m t is the total number of crack initiation clusters;

[0060] S64. Calculate the second fitness value of the initial population, and according to the elite retention strategy, save the individual with the smallest fitness value in the current population as the elite individual and do not perform inheritance;

[0061] S65. Take the non-elite individuals in the initial population, perform crossover and mutation operations on them to obtain a new population;

[0062] S66. Repeat S64 and S65 until the evolutionary generation is reached, and the final result is the total length of each cluster of cracks at time t;

[0063] S67. Repeat S61 to S66 to obtain the total length of each cluster of cracks under different time conditions.

[0064] After obtaining the total length of each cluster of cracks under different time conditions, combined with the crack width and crack height under different time conditions calculated in the S62 process, the method of the embodiment of the present invention can be used to grasp the expansion of multiple clusters of fracturing cracks in real time, provide guidance for fracturing process optimization and construction plan adjustment, and thus effectively improve the recovery rate of horizontal wells.

[0065] The above description is only a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any technician familiar with the profession can make some changes or modifications to equivalent embodiments of equivalent changes using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of the technical solution of the present invention.

Claims

1. An intelligent simulation method for synchronous expansion of multi-cluster hydraulic fracturing cracks in a horizontal well, characterized in that: The following steps are involved: S1. Obtain the well logging data of the developed horizontal wells in the target block and the fracture pressure of the segmented and clustered fracturing, and match the well logging data with the fracture pressure; S2. Using logging data as input and fracture pressure as output, a fracture pressure prediction model is established based on gated recurrent units and attention mechanism, and the logging data and fracture pressure prediction model of the developed horizontal wells in the current block are used for training; S3. Based on the continuity equation, pressure drop equation, crack width equation, and crack height equation, combined with the ground stress data of the target block, a multi-cluster hydraulic fracture synchronous expansion simulation model is established, and discretized to obtain a discretized model. The coupling conditions of the model are established according to the volume balance and pressure balance. The discretized model is as follows: ； The coupling conditions are: Volume Balance Equation , the pressure balance equation ; Wherein, subscripts j and k are the crack cluster numbers, and j, k = 1, 2, …, m; P j is the flow pressure in fracture cluster j; f 1 is the equation on the right side of the continuity equation; f 2 is the equation on the right side of the pressure drop equation; f 3 is the crack height equation, the right side of the crack length l j Find the partial derivative formula; t and l are the time step and length step respectively; l j is the length of crack cluster j; t is time, T is the total construction time; h j is the crack height of crack cluster j; L f,j is the total length of the jth cluster of cracks; Q represents the construction displacement; represents the bottom hole pressure calculated from the opening of the j-th cluster of fractures to the reference point O; Indicates that the fracturing fluid flows from the reference point O to the k Pipe friction of cluster fractures; P b,j represents the bottom hole pressure at the jth cluster perforation position; P p,j represents the hole friction resistance at the jth cluster perforation position; f 1= , h j ( l j , t )yes t Time next j Cluster cracks l j The seam height at C L is the comprehensive filtration coefficient; τ 0( l j ) is the time when the fracturing fluid reaches the crack l j The time required for , w j ( l j , t )yes t Time next j Cluster cracks l j The width of the seam at A j ( l j , t )yes t Time next j Cluster cracks l j The cross-sectional area at f 2= ; yes t Time next j Cluster cracks l j The 2n'+1 power of the slit width at n’ is the fracturing fluid flow index; f 3= ; E is the elastic modulus; v is Poisson's ratio; σ h ( l j ) is the j Cluster crack extension l j Minimum principal stress at the length; P j ( l , t )yes t Time next j Cluster cracks l j The pressure of S4, obtaining the logging parameters and bottom hole pressure during the fracturing construction of the target well, predicting the fracture pressure of the target well based on the trained prediction model, and obtaining the total number of fracture initiation clusters during the fracturing process of the target well in combination with the pressure balance equation; S5. Based on the adaptive elite genetic algorithm, a fitness function is established with the minimum error of the bottom hole pressure from the opening position of each cluster of fractures in the target well to the reference point O. The volume balance equation is used as a constraint and the flow distribution of the opening position of each cluster of fractures is used as a goal. The flow distribution results of the opening position of each cluster of fractures under different time conditions are obtained by solving the problem. S6. Based on the adaptive elite genetic algorithm, the fitness function is established with the absolute minimum error of the flow pressure at the fracture mouth of each cluster of fractures and the bottom hole pressure at the perforation position of each cluster. The discretized model is used to solve the flow pressure, flow rate, fracture width and fracture height from the end of each cluster of fractures to the fracture mouth. The total length of each cluster of fractures is taken as the target, and the length of each cluster of fractures under different time conditions is solved, and then the three-dimensional size of each cluster of fractures under different time conditions is obtained.

2. The method according to claim 1, characterized in that In S1, the acquisition of the fracturing pressure includes the following sub-steps: obtaining the segmented and clustered data and fracturing operation data of the developed horizontal wells in the target block, and calculating the bottom hole pressure of each cluster perforation position in each well section; Based on the fracturing operation data, the first peak on the bottom hole pressure curve when the displacement is in an increasing or constant state is recorded. This peak corresponds to the fracture pressure of the cluster with the lowest minimum principal stress.

3. The method according to claim 2, characterized in that The segmentation and clustering data include the number of segments, segment length, number of clusters, cluster spacing, and cluster depth; the fracturing construction data include wellhead casing pressure, displacement, and sand ratio.

4. The method according to claim 1, characterized in that: In S1 and S2, the logging data include depth, density, resistivity, natural gamma, natural potential, acoustic time difference, and compensated neutron.

5. The method according to claim 1, characterized in that S4 includes the following sub-steps: S41. predicting the fracture pressure of each cluster using the trained prediction model according to the logging parameters of each cluster of the target well at time t; S42. According to the fracturing construction data of the target well at time t, the bottom hole pressure of each cluster perforation position is obtained. When the bottom hole pressure of a cluster is not less than the predicted value of the fracture pressure of the cluster, the cluster is fractured; S43. recalculating the bottom hole pressure of each cluster that has not been fractured according to the pressure balance equation, and comparing it with the predicted fracture pressure value to determine whether fracture has occurred in each cluster; S44. Repeat S43 until no new clusters start to crack; S45. Repeat S41 to S44 until fracturing is completed.

6. The method according to claim 1, characterized in that S5 includes the following sub-steps: S51. Based on the adaptive elite genetic algorithm, the volume balance equation is used as a constraint, and the initial population of flow distribution at the location of each cluster of crack openings is randomly generated; S52. Establish a fitness function based on the minimum error between the bottom hole pressure of each cluster fracture opening and the reference point O: , where f Q is the first fitness value, indicating the bottom hole pressure error from the fracture position to the reference point O; m t is the total number of cracking clusters; x and y represent m t The xth cluster and the yth cluster in the cluster; S53. Calculate the first fitness value of the initial population, and according to the elite retention strategy, save the individual with the smallest fitness value in the current population as the elite individual and do not perform inheritance; S54. Take the non-elite individuals in the initial population, perform crossover and mutation operations on them to obtain a new population; S55. Repeat S53 and S54 until the evolutionary generation is reached, and the final result is the flow distribution result of the crack opening position of each cluster at time t; S56. Repeat S51-S55 to obtain the flow distribution results of the crack opening positions of each cluster under different time conditions.

7. The method according to claim 1, characterized in that S6 includes the following sub-steps: S61. Based on the adaptive elite genetic algorithm, the initial population of the total length of each cluster of cracks is randomly generated; S62. According to the total length of each cluster of cracks, l is the step length, and using the discretized model, starting from the end of each cluster of fractures, the flow pressure, flow rate, fracture width and fracture height are calculated step by step until the fracture opening; S63. Establish a fitness function by minimizing the absolute difference between the flow pressure at the opening of each cluster of fractures and the bottom hole pressure at the perforation position: , where f f is the second fitness value, which represents the absolute difference between the flow pressure at the fracture position and the bottom hole pressure at the perforation position; m t is the total number of crack initiation clusters; S64. Calculate the second fitness value of the initial population, and according to the elite retention strategy, save the individual with the smallest fitness value in the current population as the elite individual and do not perform inheritance; S65. Take the non-elite individuals in the initial population, perform crossover and mutation operations on them to obtain a new population; S66. Repeat S64 and S65 until the evolutionary generation is reached, and the final result is the total length of each cluster of cracks at time t; S67. Repeat S61 to S66 to obtain the total length of each cluster of cracks under different time conditions, and combine the crack width and crack height to obtain the three-dimensional size of each cluster of cracks under different time conditions.

Images