Multi-target mining optimization and intelligent control system for coal-bed gas well

Abstract

The invention relates to the field of mining engineering, in particular to a coal-bed gas well multi-target mining optimization and intelligent control system which comprises a channeling layer pollution monitoring and tracing module, a self-adaptive profile control and displacement control module, a pulverized coal blockage prevention and control module, a coal seam stress self-adaptive compensation module and a management database. The system establishes a pollution diffusion path prediction model through acoustic signals, searches for an optimal pollution diffusion path based on a permeability gradient field, constructs a risk coordinate system to divide a four-quadrant strategy to implement dynamic regulation and control according to the diffusion area growth rate and fluid channeling velocity of pollution, and monitors the pulverized coal concentration and particle size of the output liquid to achieve the optimal control of the pollution diffusion path. According to the method, a critical flow velocity model is constructed, a safe extraction rate threshold value of pulverized coal is determined, and gas injection parameters are adjusted according to fracture extension characteristics by monitoring coal seam strain, so that the accuracy and flexibility of pollution prevention and control are improved, the capability of coping with complex and changeable pollution conditions is enhanced, and efficient exploitation of resources is realized.

Description

Technical Field

[0001] The present invention relates to the field of mining engineering, and more specifically, it is an optimization and intelligent control system for multi-objective exploitation of coalbed methane wells. Background Art

[0002] Coalbed methane, as an efficient and clean unconventional natural gas resource, is playing an increasingly important role in the global energy structure. With the gradual scarcity of traditional fossil energy and the rapid growth of the demand for clean energy, the development and utilization of coalbed methane have received extensive attention.

[0003] China is rich in coalbed methane resources. The geological resources of coalbed methane with a burial depth shallower than 2,000 meters are approximately 36.81 trillion cubic meters, possessing huge development potential. However, the exploitation of coalbed methane faces many challenges. How to achieve efficient, safe, and environmentally friendly multi-objective exploitation has become a key issue that needs to be solved urgently.

[0004] However, the existing technologies still have deficiencies. For example, the Chinese patent with the application number 202410669101.1 discloses a remote control system and method for coalbed methane well exploitation. This solution monitors the environmental parameters of the target coalbed methane well in real time during the exploitation process of the target coalbed methane well, determines in real time whether there is a need for regulation of the pumping equipment in the target coalbed methane well, and performs corresponding regulation and treatment on the pumping equipment in the target coalbed methane well according to the analysis. By using the environmental parameters of the target coalbed methane well, it determines whether the regulation and treatment of the pumping equipment in the target coalbed methane well are effective, realizes the precise and efficient regulation of the pumping equipment in the coalbed methane well, maintains the stability and safety of the coalbed methane well exploitation, and thus optimizes the management of the coalbed methane well.

[0005] This solution has the following deficiencies: This solution mainly focuses on the wellbore environmental parameters and geological environmental parameters, without considering pollution crossflow. Pollution crossflow may change the geological environment and wellbore environment around the coalbed methane well, and it may not be able to detect pollution crossflow in time and make corresponding regulation decisions.

[0006] For example, the Chinese patent with the application number 202310450311.7 discloses a goaf coalbed methane extraction well and extraction method based on the transformation of abandoned coalbed methane wells. This solution surveys the damaged parts of the vertical well section structure of the abandoned well and the coalbed methane content in the goaf, sets a packer near the damaged part of the vertical well section, repairs the intact vertical well section, drills a horizontal well in the relatively stable rock layer near the fracture zone above the goaf and close to the caving zone, and the horizontal section is completed with a full-hole casing or screen pipe. For casing completion, directional perforation is required. Finally, a negative pressure extraction device is installed at the wellhead of the vertical well, and the gas in the goaf is transported to the surface pipeline through the horizontal well and vertical well by means of negative pressure to complete the extraction, making full use of the cracks generated by the pressure relief of the overlying rock layer in the goaf for extraction and reducing the leakage of goaf gas to the ground.

[0007] The following are the deficiencies of this solution: The gas drainage method reformed by this solution only drains the goaf gas through negative pressure drainage. In the regulation of mining parameters, it only involves the setting of the horizontal well section and the determination of the horizontal section length according to different positions, lacking dynamic and precise regulation of key parameters such as drainage speed and pressure. Summary of the Invention

[0008] To overcome the deficiencies in the background technology, the embodiments of the present invention provide an optimization and intelligent control system for multi-target mining of coalbed methane wells, which can effectively solve the problems involved in the above background technology.

[0009] The object of the present invention can be achieved through the following technical solutions: The present invention provides an optimization and intelligent control system for multi-target mining of coalbed methane wells, including: a cross-layer pollution monitoring and tracing module, which is used to detect acoustic signals in real time, establish a pollution diffusion path prediction model through acoustic signal analysis, and search for the optimal pollution diffusion path based on the permeability gradient field.

[0010] An adaptive displacement control module, which is used to construct a risk coordinate system to divide the four-quadrant strategy according to the growth rate of the pollution diffusion area and the crossflow velocity, and implement dynamic regulation.

[0011] A pulverized coal blockage prevention and control module, which is used to construct a critical flow velocity model by monitoring the pulverized coal concentration and particle size in the produced fluid, and determine the safe drainage rate threshold of pulverized coal.

[0012] A coal seam stress adaptive compensation module, which is used to monitor the coal seam strain and adjust the gas injection parameters according to the fracture propagation characteristics.

[0013] A management database, which is used to store historical crossflow event records and control strategy parameters.

[0014] Compared with the prior art, the present invention has the following beneficial effects: First, the present invention establishes a pollution diffusion path prediction model through acoustic signal analysis, and searches for the optimal pollution diffusion path based on the permeability gradient field, fully considering the influence of coal seam geological characteristics on pollution diffusion, and greatly improving the prediction accuracy.

[0015] Second, the present invention constructs a risk coordinate system to divide the four-quadrant strategy according to the growth rate of the pollution diffusion area and the crossflow velocity, and implements dynamic regulation for different quadrant situations, and can adopt the most suitable coping strategy according to the specific characteristics of the pollution.

[0016] Third, the present invention constructs a critical flow velocity model by monitoring the pulverized coal concentration and particle size in the produced fluid, and determines the safe drainage rate threshold of pulverized coal. A stable drainage rate can make the entire drainage system operate more smoothly, and at the same time can effectively avoid the deposition of pulverized coal in equipment such as the wellbore.

[0017] IV. The present invention monitors the coal seam strain and adjusts the gas injection parameters according to the fracture propagation characteristics. Reasonably adjusting the gas injection parameters can optimize the distribution and flow of gas in the coal seam and improve the recovery rate of coalbed methane. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0019] Figure 1 It is a connection diagram of the system modules of the present invention.

[0020] Figure 2 It is Figure 1 a flowchart of the cross-layer pollution monitoring and tracing module in

[0021] Figure 3 It is Figure 2 a flowchart of step S3 in

[0022] Figure 4 It is Figure 2 a flowchart of step S4 in DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0024] Please refer to Figure 1 as shown, a multi-objective exploitation optimization and intelligent control system for coalbed methane wells includes a cross-layer pollution monitoring and tracing module, an adaptive displacement control module, a coal powder plugging prevention and control module, a coal seam stress adaptive compensation module, and a management database.

[0025] The management database is connected to the cross-layer pollution monitoring and tracing module, the adaptive displacement control module, the coal powder plugging prevention and control module, and the coal seam stress adaptive compensation module, and the cross-layer pollution monitoring and tracing module is connected to the adaptive displacement control module.

[0026] The cross-layer pollution monitoring and tracing module is used to detect acoustic signals in real time, establish a pollution diffusion path prediction model through acoustic signal analysis, and search for the optimal pollution diffusion path based on the permeability gradient field.

[0027] Please refer to Figure 2As shown, the cross-layer pollution monitoring and tracing module includes the steps of: S1. Arranging a distributed acoustic sensing array along the wellbore and key positions between layers, and continuously collecting the original acoustic signals of the interlayer fluid at a preset sampling frequency; continuously collecting the original acoustic signals at a preset sampling frequency realizes real-time dynamic monitoring of the state of the interlayer fluid and can timely capture information on fluid changes.

[0028] S2. Preprocessing the original acoustic signals, performing Fourier transform on the preprocessed acoustic signals, converting the acoustic signals from the time domain to the frequency domain, analyzing the frequency spectrum of the acoustic signals in the frequency domain, searching for relevant characteristic signals of interlayer fluid crossflow, classifying and storing the characteristic signals according to the acquisition time and acquisition position, and establishing a characteristic signal database; by studying the characteristic signals at different times and positions, the laws and development trends of interlayer fluid crossflow can be deeply understood, providing a scientific basis for further prevention and control measures.

[0029] It should be noted that the fast Fourier transform is adopted, and the number of points of the power of 2 is selected as the number of points for the Fourier transform. The frequency resolution is calculated by dividing the sampling rate by the number of transform points, and then the Fourier transform is performed on the original acoustic signals to convert them from the time domain to the frequency domain.

[0030] It should be noted that the specific analysis method for searching for relevant characteristic signals of interlayer fluid crossflow is: analyzing the spectral peak value. The interlayer fluid crossflow causes obvious spectral peak values at specific frequencies. Due to the change in the fluid flow state and the pressure fluctuation during the crossflow process, the natural frequency vibration is induced.

[0031] Exemplarily, through the analysis of a large amount of experimental data or actual monitoring data, when crossflow occurs, one or more significant peaks appear in the frequency band of 100 - 200 Hz. The frequency positions and amplitude sizes of these peaks are used as part of the characteristic signals for identifying interlayer fluid crossflow.

[0032] Including analyzing the change in the spectral bandwidth. Under normal circumstances, the spectral bandwidth of the acoustic signal is relatively stable. When interlayer fluid crossflow occurs, the spectral bandwidth changes due to the irregular flow and interaction of the fluid. For example, the increase in high-frequency components causes the spectral bandwidth to broaden, or the suppression of some frequency components causes the bandwidth to narrow. By calculating the spectral bandwidth and comparing it with the bandwidth in the normal state, the change amount of the bandwidth is used as the characteristic signal for identifying interlayer fluid crossflow.

[0033] Analyzing the harmonic structure. The interlayer fluid crossflow destroys the periodicity and symmetry of the original fluid system, thereby causing changes in the harmonic structure in the frequency spectrum. Extra harmonic components appear in the original regular harmonic sequence, or the amplitude ratio of some harmonics changes. By analyzing the characteristics such as the distribution, amplitude ratio, and whether there are abnormal harmonics of the harmonics, it is used to identify the interlayer fluid crossflow signal.

[0034] S3. Collect geological data of the coal seam, including porosity and permeability, calculate the permeability gradient field of the coal seam using numerical simulation methods, and construct a pollution diffusion path prediction model based on the calculated permeability gradient field using algorithms such as the shortest path algorithm; constructing a pollution diffusion path prediction model based on the permeability gradient field using algorithms such as the shortest path algorithm can predict the diffusion path of pollution in the coal seam according to the geological characteristics of the coal seam and the law of fluid flow.

[0035] S4. According to the pollution diffusion path prediction model, determine the starting position of the pollution, use the corresponding node as the starting point, and search for the optimal path from the starting point to the target point to form the optimal pollution diffusion path. Searching for the optimal path from the starting point to the target point to form the optimal pollution diffusion path can more accurately predict the diffusion direction and range of the pollution, help to reasonably arrange monitoring points and treatment measures, improve the treatment efficiency, and reduce the impact of pollution on the environment and resources.

[0036] Please refer to Figure 3 As shown, the specific content of step S3 is as follows: S31. Determine the calculation area of the numerical model according to the actual range of the coal seam, divide this calculation area into several grid cells, and at the same time define the boundary conditions of the coal seam. The boundary conditions include flow boundary conditions, pressure boundary conditions, and closed boundary conditions; defining the boundary conditions of the coal seam provides accurate constraint conditions for the numerical model, enabling the model to more realistically simulate the flow state of fluids in the coal seam under different boundary conditions, and helping to obtain more accurate calculation results.

[0037] It should be noted that the specific analysis method for dividing this calculation area into several grid cells is as follows: Divide the three-dimensional space area where the coal seam is located in three directions. Divide it into grid cells in the direction, with the distance between adjacent grid nodes being . Divide it into grid cells in the direction, with the grid spacing being . Divide it into grid cells in the direction, with the grid spacing being . In this way, a three-dimensional grid system composed of several small grid cells is formed. Each grid node has a corresponding coordinate. At the same time, divide the entire time process into time steps, and the length of each time step is .

[0038] It should be noted that based on the actual distribution range of the coal seam, the boundary of the calculation area is determined by appropriately expanding a certain distance outward around it. The expansion distance can be estimated based on the average thickness of the coal seam or the scale of the main geological structure. For example, it is expanded by 1-5 times the average thickness of the coal seam.

[0039] It should be noted that the flow boundary conditions include inflow boundaries and outflow boundaries. If external fluid flows into the coal seam, it is necessary to set the inflow flow boundary conditions. The flow rate can be estimated by measuring the head difference, permeability and water-passing cross-sectional area of ​​the aquifer using Darcy's law.

[0040] For example, it is known that the hydraulic head difference between the adjacent aquifer and the coal seam is The permeability of the fault is The cross-sectional area of ​​water flow is , then the inflow ,in is the flow path length. In the numerical model, the calculated flow value is assigned to the grid cell of the inflow boundary.

[0041] When the fluid in the coal seam flows out through a specific channel, the outflow flow boundary condition is set. For the drainage well, the outflow flow is determined according to the drainage capacity and operation time of the drainage equipment.

[0042] The pressure boundary conditions include a constant head boundary and a pressure gradient boundary. If the coal seam boundary is connected to a water body with a stable head, the boundary can be set as a constant head boundary condition, and the head value at the boundary can be determined based on the actual measured water level of the water body.

[0043] If the pressure change at the coal seam boundary follows a certain gradient law, such as in areas where the coal seam burial depth varies greatly, the pressure at the boundary changes linearly with depth. The pressure gradient is determined based on the relationship between pressure and depth, and the pressure gradient boundary condition is set in the numerical model.

[0044] S32. Collect geological data of coal seams, including porosity and permeability, use the geological data as input parameters, assign corresponding values ​​to corresponding grid cells in the numerical model, and construct a mathematical model of fluid flow in coal seams through simultaneous equations; integrate geological information into the numerical model so that the model can take into account the impact of the physical properties of the coal seam on the fluid flow, and more accurately describe the flow patterns of the fluid in the coal seam.

[0045] It should be noted that the specific construction of the mathematical model of fluid flow in the coal seam is as follows: according to the physical process of fluid flow in the coal seam, the corresponding control equations are selected, the selected control equations are combined to form a group of equations, and the group of equations is numerically solved using a solver.

[0046] S33. The finite difference method is used to discretize the mathematical model, thereby obtaining an algebraic equation system. By solving this algebraic equation system, the pressure values at each grid node are obtained, the flow velocity vectors of each grid cell are calculated, and further, based on the relationship between the flow velocity vector and the permeability, the permeability gradient field of the coal seam is calculated; it can quantitatively describe the distributions of pressure, flow velocity, and permeability gradient in the coal seam, providing key physical quantity information for analyzing the flow characteristics of fluids in the coal seam and predicting pollution diffusion.

[0047] It should be noted that the specific analysis method for the permeability gradient field of the coal seam is as follows: for the gradients of pressure in all directions, finite difference discretization is performed, and combined with Darcy's law, an algebraic equation system regarding the pressures of each grid node is obtained. The obtained algebraic equation system is solved by the direct method to obtain the pressure values of each grid node at each time step. According to the deformation formula of Darcy's law, when the flow velocity vector and the pressure gradient are known, the permeability of each grid cell is calculated, and the central difference approximation is used to calculate the gradients of the permeability in all directions, thereby obtaining the permeability gradients of the coal seam in all directions, and further obtaining the permeability gradient field of the entire coal seam.

[0048] S34. The coal seam calculation area is constructed as a graph based on grid cells. Each node in the graph corresponds to a grid cell, and adjacent nodes are connected by edges to represent the connectivity relationship between nodes. Based on the permeability gradient field of the coal seam, the reciprocal of the permeability gradient amplitude is used as the basic weight to assign values to each edge in the graph; the graph structure can intuitively represent the spatial relationship and connectivity between different grid cells in the coal seam, providing a basic framework for subsequent applications of graph theory algorithms.

[0049] S35. In the constructed graph structure, the source node of the pollution source is determined and used as the starting point for calculating the pollution diffusion path. The shortest path algorithm is used to obtain the shortest paths from the starting node to other nodes in the graph structure, and all the paths calculated by the shortest path algorithm are integrated. Based on these paths, a pollution diffusion path prediction model is constructed; determining the source node of the pollution source in the constructed graph structure and using it as the starting point for calculating the pollution diffusion path clarifies the source location of the pollution, provides a clear starting point for subsequent path calculations, makes the prediction results more targeted and practically significant, and can accurately simulate the diffusion process of pollution starting from the source.

[0050] Please refer to Figure 4As shown in the figure, the specific content of step S4 is as follows: S41. Obtain the actual pollution-related information, where the pollution-related information includes the time and location of pollution occurrence and the source of pollutants, so as to accurately locate the starting position of pollution in the pollution diffusion path prediction model, and determine the node corresponding to this starting position in the model as the starting point for subsequent search operations; this avoids path deviation caused by uncertain starting positions, improves the accuracy and reliability of prediction, and provides precise starting conditions for effectively preventing and controlling pollution diffusion.

[0051] S42. Define the target area and target position according to actual needs, and determine the target points corresponding to this target area and target position in the pollution diffusion path prediction model; this helps to specifically analyze the diffusion of pollution in the coal seam to a specific area or position, meeting different actual application requirements.

[0052] S43. Use the determined starting point and target point as input parameters, and perform a search in the pollution diffusion path prediction model using a path search algorithm. Calculate different paths from the starting point to the target point based on the weights of the edges in the model, and calculate the total weights of each path. Screen out the path with the minimum total weight and determine it as the optimal pollution diffusion path of pollution in the coal seam; this can comprehensively consider physical characteristics such as the permeability gradient of the coal seam (reflected by the weights of the edges), comprehensively analyze possible pollution diffusion paths, and find the most likely diffusion path.

[0053] It should be noted that in a specific embodiment, there is a large coal mine area where a pollution diffusion path prediction model based on grid cells has been constructed. The coal mine area is 500 meters long and 300 meters wide, and is divided into 50×30 grid cells, with each grid cell being 10 meters×10 meters in size. The edge weights in the model have been assigned according to the reciprocal of the amplitude of the coal seam permeability gradient.

[0054] At 10 am one day, the monitoring system detected a pollution incident at the position with an abscissa of 200 meters and an ordinate of 150 meters in the coal mine area. After investigation, the pollutants originated from the illegal discharge of a coal processing workshop near this position.

[0055] According to the grid division, the position with an abscissa of 200 meters and an ordinate of 150 meters corresponds to the grid cell in the 20th column (200÷10) and the 15th row (150÷10). In the pollution diffusion path prediction model, the node corresponding to this grid cell is the pollution starting node.

[0056] Since there is an important water source protection area downstream of the coal mine area, it is necessary to determine whether the pollution will spread to this protection area and the most likely spread path. The water source protection area is located in the lower right corner of the coal mine area, with a range of 400 - 500 meters in the abscissa and 200 - 300 meters in the ordinate. In the model, the water source protection area corresponds to multiple grid cells. To simplify the analysis, we select the grid cell node corresponding to the upper left corner (abscissa 400 meters, ordinate 200 meters) of the protection area as the target point, and this node is located in the 40th column and the 20th row.

[0057] Use Dijkstra's algorithm to conduct path search. Take the determined starting point (the node corresponding to the 20th column and the 15th row) and the target point (the node corresponding to the 40th column and the 20th row) as the input parameters of Dijkstra's algorithm. The algorithm starts to run. Starting from the starting point, according to the weights of the edges in the model, gradually explore the paths to other nodes. During the exploration process, the algorithm records the current shortest path from the starting point to each node and its total weight.

[0058] When the algorithm finishes running, find all possible paths from the starting point to the target point and their corresponding total weights. After comparison, it is found that the total weight of one path is the smallest. This path successively passes through nodes such as the 22nd column and the 16th row, the 25th column and the 18th row, the 30th column and the 19th row, etc., and finally reaches the target point. This path with the smallest total weight is the optimal pollution spread path of the pollution in the coal seam.

[0059] The adaptive displacement control module is used to construct a risk coordinate system to divide the four - quadrant strategy for dynamic regulation according to the growth rate of the pollution spread area and the cross - flow velocity.

[0060] The specific analysis method of the said adaptive displacement control module is as follows: A1. According to the optimal pollution spread path and spread velocity of the pollution in the coal seam, gradually update the states of each node, record the set of nodes that have been polluted at each time step, which constitutes the pollution spread boundary at the current time, calculate the growth rate of the pollution spread area, and at the same time extract the cross - flow velocity according to the change of the signal frequency in the frequency spectrum of the acoustic wave signal; it can grasp the dynamic spread situation of the pollution in the coal seam in real time. Calculating the growth rate of the pollution spread area can quantify the speed change of the pollution spread, and extracting the cross - flow velocity according to the acoustic wave signal frequency spectrum supplements the dynamic information of the pollution propagation from another angle, providing accurate data support for comprehensively understanding the pollution spread process.

[0061] It should be noted that the specific analysis method for calculating the growth rate of the pollution diffusion area is as follows: Set the time step and initialize the time. Only mark the starting node as polluted. Within each time step, update the pollution status of each node according to the optimal pollution diffusion path and the diffusion speed. If the adjacent node of a certain node has been polluted and meets the diffusion conditions, then mark this node as polluted. After each time step, record the set of polluted nodes. For the nodes in the set, if there are unpolluted adjacent nodes, then this node is a boundary node, and these boundary nodes constitute the pollution diffusion boundary at the current time. Define the pollution area as the area covered by the polluted nodes. Within each time step, calculate the change in the pollution area, and through the formula Calculate the growth rate of the pollution diffusion area , where is the pollution area of the previous time step.

[0062] A2. Taking the crossflow velocity as the horizontal axis and the growth rate of the pollution diffusion area as the vertical axis, construct a two-dimensional risk coordinate system, preset the velocity threshold and the area growth rate threshold, and divide the four-quadrant control strategy; through this coordinate system, the risk state of the pollution can be clearly judged, which is convenient for taking corresponding control measures and improves the scientificity and accuracy of risk assessment.

[0063] A3. Respectively set the judgment criteria for pollutant accumulation and diffusion trends. Determine the pollution channels by real-time collecting pollution data, compare the optimal pollution diffusion path of the pollution in the coal seam with the position of the pollution channels. If there is an overlapping part between the two and at the same time meet the judgment conditions of quadrant 1 or 2, then determine that this channel is in a high-risk diffusion path; determine the pollution channels by real-time collecting data and compare them with the optimal pollution diffusion path, which can accurately identify the pollution diffusion paths in high risk.

[0064] A4. When the pollution data of the pollution channel is greater than the set danger threshold, trigger the reverse water injection operation, continuously inject clear water into the pollution channel through a high-pressure pump until the pollution concentration drops to the set proportion of the danger threshold; injecting clear water through a high-pressure pump can effectively dilute the pollution concentration until it drops to a safe level, thereby avoiding the further spread of pollution, protecting the coal seam and the surrounding environment from the harm of pollution, and realizing the effective control and treatment of pollution.

[0065] The specific content of the quadrant control strategy in step A2 is as follows: When the crossflow velocity is greater than the preset velocity threshold and the growth rate of the pollution diffusion area is greater than the area growth rate threshold, it is demarcated into quadrant 1, and an emergency pulse is triggered. According to the default adjustment amplitude of the emergency pulse mode, the pressure wave frequency of the pulse gas injection device is increased and the intensity is increased to block the main crossflow channel; this helps to generate a stronger impact force in a short time, effectively block the main crossflow channel, prevent the pollution from rapidly spreading to a larger area, and minimize the harm caused by the pollution to the greatest extent.

[0066] It should be noted that by analyzing the historical crossflow event data, the pressure wave frequency increase range and the pressure wave intensity increase range are respectively obtained when the median suppression time is the shortest, and the median value of the range is taken as the default adjustment amplitude of the emergency pulse mode. At the same time, the system is allowed to dynamically fine-tune through reinforcement learning.

[0067] Specifically refer to Table 1

[0068]

[0069] In the effective working conditions, for group 8, when the frequency is increased by 30% and the intensity is increased by 20%, the blocking time is the shortest, which is 8 minutes, and the safety margin is 0.3 MPa, meeting the equipment safety requirements. Compared with other effective working conditions, it is determined as the default adjustment amplitude.

[0070] When the crossflow velocity is less than or equal to the preset velocity threshold and the growth rate of the pollution diffusion area is greater than the area growth rate threshold, it is demarcated into quadrant 2, and the osmotic gradient displacement is started. The geological data of the coal seam is read, the permeability data is extracted from the geological data of the coal seam, and the high-permeability areas are screened out from each region according to the permeability data, and the viscoelastic displacement agent is injected in the opposite direction of the permeability gradient. The injection volume is positively correlated with the amplitude of the permeability gradient; injecting more displacement agent in the area with a larger permeability gradient can more effectively adjust the permeability structure of the coal seam, achieve precise control of pollution diffusion, and at the same time avoid the cost increase caused by excessive use of the displacement agent and the unnecessary impact on the coal seam.

[0071] It should be noted that the calculation formula for the injection volume is: injection volume = basic injection volume + amplitude of permeability gradient × proportionality coefficient.

[0072] When the crossflow velocity is greater than the preset velocity threshold and the growth rate of the pollution diffusion area is less than or equal to the area growth rate threshold, it is demarcated into quadrant 3, and local pulse oscillation is executed. The pressure wave frequency is increased, and the pressure intensity fluctuates periodically with a set amplitude based on the current value to destroy the attachment of pollutants at the crossflow inlet; this can effectively destroy the attachment of pollutants at the crossflow inlet, prevent pollutants from further entering the crossflow channel, and thus control the diffusion path and range of pollution.

[0073] It should be noted that the lower limit of the increased range of the pressure wave frequency can be set as the current frequency plus (the current crossflow velocity - the preset velocity threshold) × the frequency adjustment coefficient, and the upper limit is the lower limit plus a fixed frequency band increment.

[0074] When the crossflow velocity is less than or equal to the preset velocity threshold and the growth rate of the pollution diffusion area is less than or equal to the area growth rate threshold, it is delimited to quadrant 4 and enters the adaptive learning mode; the system can utilize this relatively stable period to accumulate data and experience. By continuously monitoring and analyzing the pollution diffusion situation, it learns the pollution characteristics and diffusion laws under different conditions, optimizes the treatment strategy for more complex pollution situations that may occur in the future, and improves the intelligence and adaptability of the system.

[0075] The specific operation content of the adaptive learning mode is as follows: Read the historical crossflow event records from the management database, and then form a training data set. Use the velocity threshold and the area growth rate threshold in the four-quadrant control strategy as decision variables, and take the shortest crossflow blocking time and the minimum dosage of the displacement agent as the optimization objectives. Iteratively train with the help of a learning algorithm to generate a threshold adjustment strategy and an optimized rule for the gas injection parameters; by continuously adjusting the threshold, the four-quadrant control strategy can more accurately adapt to different coal seam pollution diffusion situations.

[0076] Collect the crossflow velocity, the growth rate of the diffusion area, and the pollution concentration data after displacement under the current working conditions in real time, and calculate the crossflow control efficiency per unit time. If the crossflow control efficiency is lower than the preset efficiency threshold in multiple consecutive preset monitoring periods, then optimize the threshold parameters and the gas injection parameters of the four-quadrant control strategy; Collecting data in real time can timely reflect the actual situation of the current coal seam pollution diffusion, accurately grasp the dynamic changes of the pollution, and by calculating the crossflow control efficiency, the effectiveness of the current control strategy can be quantitatively evaluated, providing a basis for subsequent optimization.

[0077] Generate an optimized threshold adjustment plan and a fine-tuning strategy for the gas injection parameters according to the reinforcement learning algorithm, and perform synchronous adjustment based on this; The reinforcement learning algorithm can automatically adjust the threshold and the gas injection parameters according to the real-time collected data and the evaluation results, enabling the control strategy to adapt to different coal seam working conditions and pollution changes, improving the pollution control effect, and better coping with complex and changeable actual situations.

[0078] The specific analysis method for the judgment criteria of pollutant accumulation and diffusion trends is as follows: Calculate the concentration change amount between two adjacent monitoring time points. If this change amount exceeds the pre-set concentration change threshold and the current concentration is higher than the reference concentration of the background environment, it is determined that there is a pollutant accumulation trend. By paying attention to the concentration change amount, it is possible to timely detect whether there is an accumulation phenomenon of pollutants in a local area. If the change amount exceeds the threshold and the current concentration is higher than the background reference concentration, it can early warn that pollutants may form an accumulation, providing a basis for taking preventive measures in advance to prevent the accumulation from causing more serious pollution, and helping to effectively intervene at the initial stage of pollution and avoid the accumulation of pollutants to a dangerous level.

[0079] Use the pollution diffusion path prediction model to determine the diffusion direction of pollution in the coal seam. In this direction, by continuously monitoring the positions where the pollution front reaches these points at different time points, calculate the pollution diffusion speed. When the calculated diffusion speed exceeds the set diffusion speed threshold, it is determined that there is an obvious diffusion trend. By using the model to determine the diffusion direction and calculate the speed, it is possible to dynamically master the propagation situation of pollution. When the diffusion speed exceeds the set threshold, it is determined that there is an obvious diffusion trend, which is conducive to arranging prevention and control measures in advance, timely adjusting prevention and control resources according to the direction and speed of pollution diffusion, carrying out targeted pollution control, improving the accuracy and effectiveness of pollution prevention and control, and reducing the pollution influence range and harm degree.

[0080] It should be noted that in a specific embodiment, there is a coal seam area with a length of 1000 meters and a width of 800 meters, which is divided into 100×80 grid cells with a side length of 10 meters, and a corresponding pollution diffusion path prediction model is constructed. The reference concentration of the background environment has been determined to be 5 mg / m³, the pre-set concentration change threshold is 3 mg / m³, and the set diffusion speed threshold is 0.5 m / hour.

[0081] At two adjacent monitoring time points, 9 am and 10 am, the pollutant concentration of grid cell (30, 40) is monitored. At 9 am, the concentration monitoring value of this grid cell is 8 mg / m³, and at 10 am, the concentration monitoring value becomes 12 mg / m³.

[0082] Calculate the concentration change amount: 12 mg / m³ - 8 mg / m³ = 4 mg / m³. Since 4 mg / m³ > 3 mg / m³ (concentration change threshold) and the current concentration 12 mg / m³ > 5 mg / m³ (reference concentration of the background environment), it is determined that there is a pollutant accumulation trend at grid cell (30, 40).

[0083] Through the constructed pollution diffusion path prediction model, it is determined that at the current moment, the diffusion direction of pollution in the coal seam at grid cell (30, 40) is roughly southeast.

[0084] In the southeast direction, two points A and B that are 10 meters apart are selected (assuming A is closer to the pollution source). At 10:30 am, it is monitored that the pollution front reaches point A, and at 11:30 am, it is monitored that the pollution front reaches point B.

[0085] Calculate the pollution diffusion speed: The distance between the two points is 10 meters, and the time interval is 1 hour. So the diffusion speed = 10 meters ÷ 1 hour = 10 meters / hour. Since 10 meters / hour > 0.5 meters / hour (the set diffusion speed threshold), it is determined that there is an obvious diffusion trend in this diffusion direction.

[0086] The coal powder blockage prevention and control module is used to construct a critical flow velocity model and determine the safe extraction rate threshold of coal powder by monitoring the coal powder concentration and particle size in the produced fluid.

[0087] The specific operation content of the coal powder blockage prevention and control module is as follows: By emitting a laser beam through the produced fluid, the concentration and particle size distribution information of coal powder in the produced fluid are obtained in real time using the laser scattering principle; accurately grasping the concentration and particle size distribution of coal powder helps to analyze the changes in the coal seam and the impact on the extraction process, so as to adjust the extraction strategy in a timely manner and ensure the smooth progress of the extraction work.

[0088] Based on the preset wellbore parameters and coal seam characteristic parameters, a critical flow velocity calculation model is constructed. By simulating the movement state of coal powder in the produced fluid at different extraction rates, the safe extraction rate threshold to avoid coal powder deposition is determined; it can effectively prevent coal powder from depositing in the wellbore, reduce problems such as pipeline blockage and equipment wear caused by coal powder deposition, reduce maintenance costs, and improve the stability and reliability of the extraction system.

[0089] It should be noted that during the simulation process, the wellbore and coal seam areas are divided into a large number of grid cells, the fluid and coal powder particles in each grid cell are calculated, and by continuously changing the extraction rate, the deposition situation of coal powder particles is observed. The minimum extraction rate corresponding to when the coal powder no longer deposits is the safe extraction rate threshold.

[0090] According to the safe extraction rate threshold, adjust the working parameters of the high-frequency pulsation generator. By controlling the frequency and intensity of the pressure fluctuation, the extraction rate is maintained within the safe threshold range; ensure the stability of the extraction process and avoid a series of problems caused by too high or too low extraction rate.

[0091] It should be noted that the extraction rate is monitored in real time and compared with the safe threshold. If there is a deviation between the two, the working parameters of the high-frequency pulsation generator are adjusted according to the magnitude and direction of the deviation.

[0092] Exemplarily, when the extraction rate is lower than 10% of the safety threshold, increase the frequency of pressure fluctuation by 20% and the intensity by 15%. When the extraction rate is higher than 15% of the safety threshold, reduce the frequency of pressure fluctuation by 30% and the intensity by 25%.

[0093] The coal seam stress adaptive compensation module is used to monitor the coal seam strain and adjust the gas injection parameters according to the fracture propagation characteristics.

[0094] The specific operation content of the coal seam stress adaptive compensation module is as follows: Drill holes in the coal seam of the coal mining face, place a fiber Bragg grating micro-strain monitoring network composed of fiber Bragg grating sensors in the holes, detect the micro-strain of the surrounding rock mass when fractures occur in the coal seam through the fiber Bragg grating sensors, and collect the reflected light signals at a set frequency; It can comprehensively understand the strain situation of the coal seam in the coal mining face, avoid the limitations of single-point measurement, and more accurately reflect the overall state of the coal seam.

[0095] Calculate the strain value based on the change in the reflected light wavelength and the relevant parameters of the fiber Bragg grating sensor. If the strain change in a certain area exceeds the pre-set threshold, mark this area as a suspicious fracture area, and determine the fracture propagation direction and speed by comparing the data collected at different times; Determining the fracture propagation direction and speed by comparing the data collected at different times helps to understand the development trend of coal seam fractures and provides a basis for formulating reasonable coal mining plans and safety measures.

[0096] It should be noted that the strain value is calculated based on the change in the reflected light wavelength and the relevant parameters of the fiber Bragg grating sensor. Its basic principle is based on the strain-wavelength sensitive characteristic of the fiber Bragg grating. When the rock mass undergoes strain, the grating pitch of the fiber Bragg grating will change, resulting in a drift in the wavelength of its reflected light. Through the formula Calculate the strain value , where is the wavelength change amount, is the central wavelength, is the effective elasto-optic coefficient.

[0097] Obtain the stress data and pore pressure data in real time, subtract the pore pressure from the total stress to get the effective stress, and compare the calculated effective stress with the preset range, and then adjust the gas injection pressure; It can optimize the gas injection process, improve the gas injection effect, and avoid problems such as coal seam damage or gas leakage caused by too high or too low gas injection pressure.

[0098] It should be noted that if the effective stress is lower than the lower limit of the preset range, it indicates that the coal seam may be in an under-compacted state, and the gas injection pressure needs to be appropriately increased to enhance the stability and gas permeability of the coal seam. If the effective stress is higher than the upper limit of the preset range, it may cause coal seam fracture or other safety problems, and the gas injection pressure should be reduced. When adjusting the gas injection pressure, the PID control algorithm is adopted, and the gas injection pressure is automatically adjusted according to the deviation between the effective stress and the preset range and the rate of change of the deviation, so that the effective stress is always maintained within the preset range.

[0099] The management database is used to store historical cross-flow event records and control strategy parameters.

[0100] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention, and still be covered by the protection scope of the present invention.

Claims

1. A multi-objective mining optimization and intelligent control system for coalbed methane wells, characterized in that: The system specifically includes the following modules: The layer pollution monitoring and tracing module is used to detect acoustic signals in real time, establish a pollution diffusion path prediction model through acoustic signal analysis, and search for the optimal pollution diffusion path based on the permeability gradient field; Adaptive control module, which is used to construct a risk coordinate system to divide the four-quadrant strategy and implement dynamic control according to the growth rate of pollution diffusion area and crossflow speed; The pulverized coal blockage prevention and control module is used to build a critical flow rate model by monitoring the pulverized coal concentration and particle size of the output liquid, and determine the safe extraction rate threshold of pulverized coal; Coal seam stress adaptive compensation module, used to monitor coal seam strain and adjust gas injection parameters according to fracture expansion characteristics; Management database, used to store historical crossflow event records and control strategy parameters.

2. The multi-objective mining optimization and intelligent control system for coalbed methane wells according to claim 1 is characterized in that: The layer contamination monitoring and tracing module includes the following steps: S1. Distributed acoustic wave sensor arrays are arranged along the wellbore and at key locations between layers to continuously collect the original acoustic signals of the interlayer fluid at a preset sampling frequency; S2. Performing Fourier transform on the original acoustic signal, converting the acoustic signal from the time domain to the frequency domain, analyzing the spectrum of the acoustic signal in the frequency domain, finding the relevant characteristic signals of the interlayer fluid crossflow, classifying and storing the characteristic signals according to the acquisition time and acquisition position, and establishing a characteristic signal database; S3. Collect geological data of coal seams, including porosity and permeability, calculate the permeability gradient field of coal seams using numerical simulation methods, and construct a pollution diffusion path prediction model based on the calculated permeability gradient field using the shortest path algorithm; S4. According to the pollution diffusion path prediction model, determine the starting position of the pollution, take its corresponding node as the starting point, search for the optimal path from the starting point to the target point, and construct the optimal pollution diffusion path.

3. The multi-objective mining optimization and intelligent control system for coalbed methane wells according to claim 2 is characterized in that: The specific content of step S3 is: S31. Determine the calculation area of ​​the numerical model according to the actual range of the coal seam, and divide the calculation area into a number of grid cells, and define the boundary conditions of the coal seam, wherein the boundary conditions include flow boundary conditions, pressure boundary conditions and closed boundary conditions; S32. Collect geological data of the coal seam, the geological data including porosity and permeability, use the geological data as input parameters, assign values ​​to corresponding grid cells in the numerical model, and construct a mathematical model of fluid flow in the coal seam through simultaneous equations; S33. The mathematical model is discretized by using a finite difference method to obtain a set of algebraic equations, and the pressure value on each grid node is obtained by solving the set of algebraic equations, and the velocity vector of each grid unit is calculated, and the permeability gradient field of the coal seam is further calculated based on the relationship between the velocity vector and the permeability; S34. The coal seam calculation area is constructed into a graph based on the grid units, wherein each node in the graph corresponds to a grid unit, and adjacent nodes are connected by edges, so as to characterize the connectivity relationship between the nodes. Based on the permeability gradient field of the coal seam, the inverse of the permeability gradient amplitude is used as a basic weight to assign a value to each edge in the graph; S35. Determine the starting node of the pollution source in the constructed graph structure, use it as the starting point for calculating the pollution diffusion path, use the shortest path algorithm to obtain the shortest path from the starting node to other nodes in the graph structure, integrate all paths calculated by the shortest path algorithm, and construct a pollution diffusion path prediction model based on the path.

4. The multi-objective mining optimization and intelligent control system for coalbed methane wells according to claim 2 is characterized in that: The specific content of step S4 is: S41. Acquire actual pollution-related information, including the time and location of pollution and the source of pollutants, so as to accurately locate the starting position of pollution in the pollution diffusion path prediction model, and determine the node corresponding to the starting position in the model as the starting point of subsequent search operations; S42. Clarify the target area and target location according to actual needs, and determine the target point corresponding to the target area and target location in the pollution diffusion path prediction model; S43. Take the determined starting point and target point as input parameters, use the path search algorithm to search in the pollution diffusion path prediction model, calculate the different paths from the starting point to the target point based on the weights of the edges in the model, and calculate the total weight of each path, screen out the path with the smallest total weight, and determine it as the optimal pollution diffusion path for pollution in the coal seam.

5. The multi-objective mining optimization and intelligent control system for coalbed methane wells according to claim 1 is characterized in that: The specific analysis method of the adaptive drive control module is as follows: A1. According to the optimal pollution diffusion path and diffusion speed of pollution in the coal seam, the state of each node is gradually updated, the set of polluted nodes at each time step is recorded, the pollution diffusion boundary at the current time is formed, the growth rate of the pollution diffusion area is calculated, and the crossflow velocity is extracted according to the signal frequency change in the spectrum of the acoustic wave signal; A2. With the crossflow velocity as the horizontal direction and the pollution diffusion area growth rate as the vertical direction, a two-dimensional risk coordinate system is constructed, the velocity threshold and area growth rate threshold are preset, and the four-quadrant control strategy is divided; A3. Set the judgment criteria for pollutant accumulation and diffusion trends respectively, determine the pollution channel by collecting pollution data in real time, compare the optimal pollution diffusion path of the pollution in the coal seam with the position of the pollution channel, and if there is an overlap between the two and the judgment conditions of quadrant 1 or 2 are met at the same time, then the channel is judged to be in a high-risk diffusion path; A4. When the pollution data of the pollution channel is greater than the set danger threshold, the reverse water injection operation is triggered, and clean water is continuously injected into the pollution channel through a high-pressure pump until the pollution concentration drops to a set proportion of the danger threshold.

6. A multi-objective mining optimization and intelligent control system for coalbed methane wells according to claim 5, characterized in that: The specific content of the four-quadrant control strategy in step A2 is: When the crossflow velocity is greater than the preset velocity threshold, and the pollution diffusion area growth rate is greater than the area growth rate threshold, it is demarcated to quadrant 1, and an emergency pulse is triggered. The pressure wave frequency and intensity of the pulse gas injection device are increased according to the default adjustment range of the emergency pulse mode to block the main crossflow channel; When the crossflow velocity is less than or equal to the preset velocity threshold, and the pollution diffusion area growth rate is greater than the area growth rate threshold, it is demarcated to quadrant 2, and the permeability gradient control and displacement is started. The geological data of the coal seam is read, and the permeability data is extracted from the geological data of the coal seam. According to the permeability data, the high permeability area is screened out from each area, and the viscoelastic control and displacement agent is injected in the reverse direction of the permeability gradient, wherein the injection amount is positively correlated with the permeability gradient amplitude; When the crossflow velocity is greater than the preset velocity threshold, and the pollution diffusion area growth rate is less than or equal to the area growth rate threshold, it is demarcated to quadrant 3, and local pulse oscillation is performed to increase the pressure wave frequency, and the pressure intensity is periodically fluctuated with a set amplitude based on the current value to destroy the attachment of pollutants at the crossflow entrance; When the crossflow velocity is less than or equal to the preset velocity threshold, and the pollution diffusion area growth rate is less than or equal to the area growth rate threshold, it is demarcated to quadrant 4 and enters the adaptive learning mode.

7. The multi-objective mining optimization and intelligent control system for coalbed methane wells according to claim 5 is characterized in that: The specific operation content of the adaptive learning mode is: The historical crossflow event records are read from the management database to form a training data set. The velocity threshold and area growth rate threshold in the four-quadrant control strategy are used as decision variables, and the shortest crossflow blocking time and the minimum amount of displacement agent are used as optimization goals. Iterative training is performed with the help of a learning algorithm to generate a threshold adjustment strategy and gas injection parameter optimization rules. Collect the crossflow velocity, diffusion area growth rate and pollution concentration data after adjustment and flooding under the current working conditions in real time, calculate the crossflow control efficiency per unit time, and if the crossflow control efficiency is lower than the preset efficiency threshold in multiple consecutive preset monitoring cycles, optimize the threshold parameters and gas injection parameters of the four-quadrant control strategy; The optimized threshold adjustment scheme and gas injection parameter fine-tuning strategy are generated according to the reinforcement learning algorithm, and synchronous adjustments are performed based on them.

8. The multi-objective mining optimization and intelligent control system for coalbed methane wells according to claim 5 is characterized in that: The specific analysis method for judging the pollutant accumulation and diffusion trend is as follows: Calculate the concentration change between two adjacent monitoring time points. If the change exceeds the preset concentration change threshold and the current concentration is higher than the baseline concentration of the background environment, it is determined that there is a trend of pollutant accumulation. The pollution diffusion path prediction model is used to determine the diffusion direction of pollution in the coal seam. In this direction, the pollution diffusion speed is calculated by continuously monitoring the positions of the pollution front reaching these points at different time points. When the calculated diffusion speed exceeds the set diffusion speed threshold, it is determined that there is an obvious diffusion trend.

9. The multi-objective mining optimization and intelligent control system for coalbed methane wells according to claim 1 is characterized in that: The specific operation contents of the coal powder blockage prevention and control module are as follows: By emitting a laser beam through the produced fluid, the concentration and particle size distribution information of coal powder in the produced fluid is obtained in real time using the principle of laser scattering; A critical velocity calculation model is constructed based on preset wellbore parameters and coal seam characteristic parameters. By simulating the movement of coal powder in the produced fluid at different extraction rates, the safe extraction rate threshold to avoid coal powder deposition is determined. According to the safe extraction rate threshold, the working parameters of the high-frequency pulsation generator are adjusted, and the extraction rate is maintained within the safe threshold range by controlling the frequency and intensity of pressure fluctuations.

10. The multi-objective mining optimization and intelligent control system for coalbed methane wells according to claim 1, characterized in that: The specific operation contents of the coal seam stress adaptive compensation module are as follows: Drilling operations are performed on the coal seam of the coal mining face, and a fiber Bragg grating micro-strain monitoring network composed of fiber Bragg grating sensors is placed in the borehole. The fiber Bragg grating sensors are used to detect the micro-strain of the surrounding rock mass when cracks occur in the coal seam, and the reflected light signals are collected at a set frequency; The strain value is calculated based on the change in the wavelength of the reflected light and the relevant parameters of the fiber grating sensor. If the strain change in a certain area exceeds the preset threshold, the area is marked as a suspected crack area, and the crack expansion direction and speed are determined by comparing the data collected at different times; The stress data and pore pressure data are acquired in real time. The effective stress is obtained by subtracting the total stress from the pore pressure. The calculated effective stress is compared with the preset range to adjust the gas injection pressure.

Images