Intelligent construction sequencing optimization method and system for underground cavern

Abstract

The invention provides an underground cavern intelligent construction sequencing optimization method and system, and the method comprises the steps: firstly integrating geological parameters, construction machinery parameters and historical construction cases to construct a construction method database, and then determining a layering height parameter set through a dynamic matching algorithm according to the spatial topological relation between geomechanical indexes and a three-dimensional geological model in the database, according to construction machinery and layering height parameters, a transverse partition and axial partition parameter set is generated by using an octree space subdivision algorithm, then composite coding is performed based on multiple attributes of the three-dimensional geologic model and the parameters, and a unique construction block coding set is generated; and finally, performing construction sequence optimization on the code set by using a multi-objective optimization algorithm to generate a Pareto optimal solution set containing construction period, cost and stability optimization paths, and dynamically adjusting construction sequence parameters in the solution set according to surrounding rock deformation data to realize intelligent construction sequencing optimization of the underground cavern.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular, to an intelligent construction sequence optimization method and system for underground chambers. Background Art

[0002] Large-span underground chambers (such as underground powerhouses of hydropower stations) face challenges such as complex surrounding rock stability and multi-process coordination during construction due to their large structural spans and high stratification heights. Traditional excavation sequence designs mostly adopt the layered and zoned method, dividing through multi-level planes and multi-processes in elevation to reduce the disturbance of the surrounding rock and control deformation. However, the spatial topological relationships of such projects are complex, and small deviations in construction parameters (advance, stratification height, support timing) may trigger chain reactions, resulting in construction delays or safety risks. Therefore, how to scientifically plan the construction sequence and dynamically optimize resource allocation has become the core problem in the efficient and safe construction of large-span underground chambers.

[0003] Current construction sequence decisions mainly rely on engineering experience and static planning models, which have significant defects: First, manual sequence division is limited by experts' subjective experience and it is difficult to quantitatively evaluate the game relationship among multiple objectives (construction period, cost, safety); Second, static models cannot respond in real time to sudden changes in geological conditions (such as fault exposure, rock mass fissure expansion) or construction progress deviations (such as mechanical failures, resource shortages), resulting in the disconnection between the sequence division plan and the actual site; Third, traditional optimization algorithms (such as linear programming) have low computational efficiency when facing high-dimensional and non-linear constrained sequence division problems and are difficult to meet the dynamic adjustment requirements of the project. Summary of the Invention

[0004] In view of the problems mentioned above, in combination with the first aspect of the present invention, embodiments of the present invention provide an intelligent construction sequence optimization method for underground chambers, and the method includes: Integrate the geological parameter set, construction machinery parameter set, and historical construction case set to construct a construction method database; Based on the geomechanical indexes in the construction method database and the spatial topological relationship of the three-dimensional geological model, determine the set of stratification height parameters through a dynamic matching algorithm; According to the construction machinery parameter set and the set of stratification height parameters, use the octree spatial subdivision algorithm to generate a set of transverse partition parameters and a set of axial block parameters; Based on the spatial coordinate attributes, surrounding rock grade attributes, the set of transverse partition parameters, and the set of axial block parameters of the three-dimensional geological model, perform composite coding processing to generate a unique set of construction block codes including hierarchical identifiers, axial identifiers, and block identifiers; The construction sequence of the unique construction block coding set is optimized by a multi-objective optimization algorithm to generate a Pareto optimal solution set including a construction period optimization path, a cost optimization path, and a stability optimization path, and the construction sequence parameters in the Pareto optimal solution set are dynamically adjusted according to the surrounding rock deformation data monitored in real time.

[0005] In another aspect, an embodiment of the present invention further provides an intelligent construction sequence optimization system for underground caverns, including a processor and a machine-readable storage medium. The machine-readable storage medium is connected to the processor. The machine-readable storage medium is used to store programs, instructions, or codes, and the processor is used to execute the programs, instructions, or codes in the machine-readable storage medium to implement the above method.

[0006] Based on the above aspects, the embodiments of the present application construct a construction method database by integrating geological parameters, construction machinery parameters, and historical construction cases. On this basis, a dynamic matching algorithm based on the spatial topological relationship between geomechanical indexes and three-dimensional geological models determines a set of layer height parameters, which not only considers the complexity of geological conditions but also realizes the adaptive matching of the layering strategy and geological structure, significantly improving the rationality and scientificity of layered construction. Further, an octree spatial subdivision algorithm is used to generate a set of horizontal partition and axial block parameters in combination with construction machinery parameters and layer height parameters. The efficient spatial partitioning ability of the octree spatial subdivision algorithm makes the division of the construction area more refined and in line with actual construction requirements. Through composite coding processing, the spatial coordinates of the three-dimensional geological model, the surrounding rock grade, and the horizontal partition and axial block parameters are uniformly coded to generate a unique construction block coding set, simplifying the expression and management of construction information. Importantly, a multi-objective optimization algorithm is used to optimize the construction sequence of the construction block coding set to generate a Pareto optimal solution set including construction period, cost, and stability optimizations, effectively balancing the contradiction between the construction period and cost while ensuring construction safety and stability, and realizing the maximization of construction benefits. In addition, by monitoring the surrounding rock deformation data in real time and dynamically adjusting the construction sequence parameters, the changes in geological conditions can be responded to in real time, ensuring the continuous optimization and dynamic adjustment of the construction process, thus significantly improving the intelligent level and overall efficiency of underground cavern construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 is a schematic execution flowchart of the intelligent construction sequence optimization method for underground caverns provided by an embodiment of the present invention.

[0008] Figure 2 is a schematic diagram of exemplary hardware and software components of the intelligent construction sequence optimization system for underground caverns provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0009] The present invention will be specifically described below in conjunction with the accompanying drawings of the specification. Figure 1 It is a schematic flow chart of an intelligent construction sequence optimization method for underground chambers provided by an embodiment of the present invention. The intelligent construction sequence optimization method for underground chambers will be introduced in detail below.

[0010] Step S110: Integrate the geological parameter set, construction machinery parameter set, and historical construction case set to construct a construction method database.

[0011] In this embodiment, step S110 may include: Step S111: Obtain the rock mass integrity coefficient and in-situ stress distribution data in the geological parameter set, and establish a first mapping relationship set between the rock mass grade and the bench height. Among them, the first mapping relationship set is obtained by non-linearly fitting the bench height records in the historical construction case set through a machine learning algorithm.

[0012] Specifically, for the rock mass integrity coefficient, the acoustic wave testing technology can be used to select multiple test points at different positions in the chamber, and the acoustic wave propagation velocity is measured multiple times at each test point. For example, 5 test points can be selected at the entrance of the chamber, and the acoustic wave propagation velocity is measured 10 times at each test point, and then the rock mass integrity coefficient is calculated according to the standard acoustic wave velocity of the rock. For example, after a large number of tests and data processing, it is obtained that the rock mass integrity coefficient of the chamber fluctuates between 0.2 and 0.9. For the in-situ stress distribution data, it can be obtained by long-term monitoring by arranging stress sensors at different depths and positions in the chamber. At key positions such as the top, side wall, and bottom of the chamber, 3-5 stress sensors can be arranged respectively, and continuous monitoring for 3 months is carried out to obtain the data of the magnitude and direction of the in-situ stress changing with time and space.

[0013] According to the obtained rock mass integrity coefficient and in-situ stress distribution data, the rock mass grade of different regions of the chamber can be determined in combination with the geological classification standards of related technologies. For example, when the rock mass integrity coefficient is relatively high and the in-situ stress is relatively small, the rock mass grade may be class I or class II; on the contrary, when the rock mass integrity coefficient is relatively low and the in-situ stress is relatively large, the rock mass grade may be class IV or class V.

[0014] Furthermore, based on the above-determined rock mass grade, a historical construction case set of several (more than 20) similar underground chamber projects can be collected, which details the bench height corresponding to different rock mass grades, and then the first mapping relationship set is established by using the support vector regression (SVR) algorithm.

[0015] Preprocessing is carried out first. For example, the surrounding rock grades can be quantitatively encoded. Class I surrounding rock is encoded as 1, Class II surrounding rock is encoded as 2, and so on. In addition, the data of the layer height can adopt the minimum-maximum normalization method. First, find the minimum and maximum values in the layer height data. Suppose the minimum value is 5 meters and the maximum value is 20 meters. For any layer height data x, the normalized value x_norm = (x - 5) / (20 - 5).

[0016] Next, the processed data can be divided into a training set and a test set according to the ratio of 80% and 20%. When training the SVR model, the radial basis kernel function (RBF) can be selected. The penalty factor C is determined to be 10 through multiple experiments and cross-validation, and the kernel coefficient γ is determined to be 0.1. During the training process, the SVR algorithm continuously adjusts the position and shape of the hyperplane to make the interval between the training data points and the hyperplane the largest, while minimizing the prediction error. After training, the test set is used to evaluate the model performance. When calculating the mean square error, the difference between the true layer height value and the predicted layer height value of each test sample can be calculated first, the differences are squared and summed, and then divided by the number of test samples. Suppose there are 100 samples in the test set, and the calculated mean square error is 0.02 and the coefficient of determination R² is 0.9, indicating that the SVR model has a good fitting effect. Finally, according to the trained SVR model, when the surrounding rock grade is Class III, assuming the predicted layer height is 9.5 meters, a first mapping relationship set between the surrounding rock grade and the layer height is established.

[0017] Step S112: Analyze the operation radius parameter of the three-boom rock drill jumbo in the construction machinery parameter set, and establish a second mapping relationship set between the mechanical operation range and the lateral partition width, where the second mapping relationship set includes the maximum lateral partition width constraint condition and the minimum safe operation spacing constraint condition.

[0018] In this embodiment, the construction machinery parameter set details various parameters of the three-boom rock drill jumbo and the drilling equipment. For example, by referring to the equipment manual and actual tests, it is determined that the maximum operation radius of the three-boom rock drill jumbo is 6.0 meters.

[0019] To establish the second mapping relationship set, consider the maximum lateral partition width constraint condition and the minimum safe operation spacing constraint condition. The maximum lateral partition width constraint condition is determined based on the operation radius of the three-boom rock drill jumbo, and the maximum lateral partition width ≤ 2×6.0 meters = 12.0 meters. The minimum safe operation spacing constraint condition is set to 1.0 meter to ensure the safety of construction personnel and equipment.

[0020] Assume that the horizontal partition is divided into three areas: left, middle, and right. To meet the maximum horizontal partition width constraint condition and the minimum safe operation spacing constraint condition, the following planning can be carried out in this embodiment. If the width of the middle area is set to 6 meters, then the maximum sum of the widths of the left and right areas is 12 - 6 - 2×1 = 4 meters (subtracting two safety spacings). The width of the left area can be set to 2 meters, and the width of the right area can be set to 2 meters. Such a partition plan not only ensures the normal operation of the three-boom rock drill jumbo but also meets the safety requirements.

[0021] Step S113: Extract the fault fracture zone treatment cases and support time sequence cases from the historical construction case set, and construct a dynamic rule base through an incremental learning algorithm. Among them, the dynamic rule base includes the layered height adjustment rule and the block size reduction rule under geological mutation conditions.

[0022] For example, from multiple collected historical construction case sets, extract the fault fracture zone treatment cases and support time sequence cases, which cover the treatment situations of fault fracture zones with different scales and complexities and the corresponding support time sequence arrangements.

[0023] Then, use the online gradient descent method in the incremental learning algorithm to construct a dynamic rule base. For example, first perform feature extraction and quantization coding on the case data of the extracted fault fracture zone treatment cases and support time sequence cases. For the fault fracture zone, quantify its width, length, rock strength, etc. For example, the width of the fault fracture zone is measured in meters, and the rock strength is determined through a rock compressive strength test. For the support time sequence cases, convert information such as the support time node and support method into numerical coding.

[0024] Then, use the online gradient descent method to train the features of the fault fracture zone treatment cases and support time sequence cases. During the training process, continuously adjust the parameters of the model so that the model can accurately predict the layered height adjustment rule and the block size reduction rule under different geological conditions. When new construction data, such as the fault information revealed by on-site drilling, is added, merge the new data with the existing data and recalculate the parameters of the model.

[0025] For example, in a certain historical case, when the width of the fault fracture zone is 3 meters and the rock strength is 5 MPa, measures such as reducing the block size to 70% of the original and increasing the temporary support are taken. Incorporate the above case data into the training. When the new on-site data shows that the width of the fault fracture zone is 3.5 meters and the rock strength is 4.5 MPa, the model generates a new block size reduction rule according to the updated parameters, which may reduce the block size to 65% of the original and increase the corresponding temporary support strength. Through continuous incremental learning, the dynamic rule base can adapt to different geological mutation situations and provide accurate rule guidance for construction.

[0026] Step S114: Perform relational data fusion on the first mapping relationship set, the second mapping relationship set, and the dynamic rule base to generate a construction method database that includes a surrounding rock grade field, a mechanical parameter field, and a construction constraint field.

[0027] In this embodiment, a relational database management system MySQL can be used to implement data fusion. First, database design is carried out, and table structures such as a surrounding rock grade table, a mechanical parameter table, and a construction constraint table are designed.

[0028] Among them, the surrounding rock grade table records the code, description, and corresponding layer height of the surrounding rock grade. For example, the code for Class I surrounding rock is 1, the description is "intact and hard rock", and the corresponding layer height is determined to be 15 meters according to the first mapping relationship set. The mechanical parameter table records information such as the operation radius of the three-boom rock drill jumbo and the parameters of the drilling equipment. The construction constraint table records the maximum lateral partition width constraint condition, the minimum safe operation spacing constraint condition, etc.

[0029] Then, the data in the first mapping relationship set, the second mapping relationship set, and the dynamic rule base are respectively imported into the corresponding tables. During the import process, the data is cleaned and preprocessed to remove duplicate data and error data. For example, it is found that the layer height data of a certain surrounding rock grade record is abnormally large. After verification, it is an input error and is removed from the data.

[0030] By establishing the association relationships between the tables, the data in different tables is associated. For example, an association is established between the surrounding rock grade table and the construction constraint table, and the corresponding construction constraint conditions can be queried through the surrounding rock grade. To improve the query efficiency and performance of the database, the database is optimized. Indexes are created, such as creating an index on the surrounding rock grade code field of the surrounding rock grade table to speed up the query of relevant information according to the surrounding rock grade. The query statements are optimized to avoid complex nested queries and reduce the calculation amount and response time of the database.

[0031] Step S115: Update the surrounding rock grade field in the construction method database through the real-time collected on-site geological exploration data, and trigger the layer height adjustment rule in the dynamic rule base to recalculate the crown arch layer height parameter, the rock anchor beam layer height parameter, and the traffic tunnel associated layer height parameter in the layer height parameter set.

[0032] In this embodiment, an unmanned aerial vehicle is used for low-altitude flight to obtain more accurate terrain and geological data. At the same time, multiple geological sensors are arranged in the cave to collect on-site geological exploration data in real time.

[0033] When new geological exploration data is collected, the geological exploration data is analyzed and processed in detail. For example, rock samples can be analyzed, combined with sensor data, to determine whether the surrounding rock grade has changed. For example, in a certain area of the cavern, the original surrounding rock grade was Class III. The new exploration data shows that the rock fractures in this area have increased and the rock strength has decreased. After comprehensive evaluation, the surrounding rock grade in this area is updated to Class IV.

[0034] Then, trigger the layer height adjustment rule in the dynamic rule library and recalculate each parameter in the set of layer height parameters.

[0035] For the top arch layer height parameter, according to the updated surrounding rock grade of Class IV, query the corresponding predicted layer height value of 8.0 meters from the first mapping relationship set. At the same time, consider the constraint condition of the bolt installation height. The bolt installation height is 3.0 meters. To ensure that the bolts can effectively fix the surrounding rock, the top arch layer height should be no less than 2 times the bolt installation height, that is, 6.0 meters. Therefore, the top arch layer height parameter is finally determined to be 8.0 meters.

[0036] For the rock anchor beam layer height parameter, first measure the vertical distance between the upper inflection point coordinates of the rock anchor beam and the bottom of the top arch layer. Use a high-precision measuring instrument, such as a total station, to take multiple measurements and calculate the average value. Suppose the measured vertical distance is 10.0 meters, and the preset threshold in the construction method database is 8.0 meters. Since the vertical distance exceeds the preset threshold, trigger the layer division decision rule to generate the rock anchor beam layer height parameter with a double-layer division mark. Divide the rock anchor beam layer into two layers, considering the operation radius of the construction machinery and the blasting vibration safety distance constraint. The operation radius of the construction machinery is 6.0 meters, and the blasting vibration safety distance is determined to be 2.0 meters according to relevant standards and calculations. After detailed calculation and analysis, the height of the first layer is set to 4.0 meters, and the height of the second layer is set to 6.0 meters.

[0037] For the traffic tunnel associated layer height parameter, it is determined by combining the traffic tunnel elevation parameter and the installation bay position parameter in the construction method database through spatial topology analysis. The elevation of the traffic tunnel is 50.0 meters, and the position of the installation bay is 10.0 meters away from the lower inflection point of the rock anchor beam. To ensure that the blasting vibration will not affect the traffic tunnel and the installation bay, calculate according to the blasting vibration propagation law and safety standards. After calculation, the lower layer height parameter of the rock anchor beam is determined to be 5.0 meters.

[0038] Step S120: Based on the geomechanical indexes in the construction method database and the spatial topology relationship of the three-dimensional geological model, determine the set of layer height parameters through a dynamic matching algorithm.

[0039] In this embodiment, step S120 may include: Step S121: Perform spatial coordinate analysis on the three-dimensional geological model to identify the upper inflection point coordinates and the lower inflection point coordinates of the rock anchor beam. Among them, the upper inflection point coordinates are located in the plane rectangular coordinate system through the surrounding rock integrity coefficient and the maximum span comparison algorithm in the construction method database.

[0040] For example, the SolidWorks software can be used to generate a three-dimensional geological model, which presents the spatial structure and geological characteristics of the underground cavern. Thus, spatial coordinate analysis can be performed on the three-dimensional geological model to establish a plane rectangular coordinate system. Take the central axis of the part above the rock anchor beam as the y-axis, the intersection of the y-axis and the bottom of the contour as the coordinate origin, and the horizontal direction as the x-axis.

[0041] To identify the upper inflection point coordinates of the rock anchor beam, the surrounding rock integrity coefficient and the maximum span comparison algorithm in the construction method database can be used. First, accurately measure the maximum span of the part above the rock anchor beam in the three-dimensional geological model. Through the measurement tool in the three-dimensional geological model, multiple measurements are taken and averaged to obtain a maximum span of 20.0 meters. Then, according to the data in the construction method database, analyze the span range corresponding to different surrounding rock integrity coefficients, and determine the surrounding rock integrity coefficient threshold as 0.6.

[0042] In the plane rectangular coordinate system, traverse all key points of the part above the rock anchor beam. For each key point, obtain its surrounding rock integrity coefficient and span information from the construction method database. Screen out the key points with a surrounding rock integrity coefficient greater than or equal to 0.6 and a span close to 20.0 meters. Suppose 5 key points are screened out, and record the absolute values of their abscissas respectively. Compare the absolute values of the abscissas of these 5 key points to find the maximum value. Suppose the absolute value of the abscissa of key point A is the largest, which is 8.0 meters, and its ordinate is 50.0 meters, then determine the coordinates of the upper inflection point as (8.0, 50.0).

[0043] For the determination of the lower inflection point coordinates of the rock anchor beam, since the slope between the lower inflection point and the upper inflection point is inconsistent, calculate the slope between each key point and the upper inflection point. From the key points in the rock anchor beam area, select a key point in turn to connect with the upper inflection point, and calculate the ratio of the difference in ordinates to the difference in abscissas between the two points, which is the slope. Suppose 20 key points are selected for calculation. When calculating to key point B, it is found that the slope between it and the upper inflection point has a significant change compared with the slopes calculated before. After further verification and analysis, determine the coordinates of key point B as (6.0, 45.0), which is the lower inflection point coordinates.

[0044] Step S122: Calculate the initial height parameter of the crown layer according to the surrounding rock grade field and geomechanical index in the construction method database. Among them, the initial height parameter of the crown layer is obtained through a dynamic calculation formula combined with the bolt installation height constraint condition.

[0045] For example, according to the surrounding rock grade field in the construction method database, it is determined that the surrounding rock grade of the current area is Class IV. At the same time, the corresponding geomechanical indexes of this surrounding rock grade are obtained. For example, the compressive strength of the rock is 15 MPa and the elastic modulus is 5 GPa.

[0046] The initial height parameter of the crown arch layer is determined by combining the dynamic calculation formula with the bolt installation height constraint condition. The specific steps are as follows: First, obtain the tensile strength parameter of the bolt material and the cohesion parameter of the surrounding rock grade field in the construction method database. For example, assume that the tensile strength parameter of the bolt material is 300 MPa and the cohesion parameter of Class IV surrounding rock is 1.5 MPa.

[0047] Then, generate the bolt spacing constraint range according to the ratio relationship between the tensile strength parameter and the cohesion parameter. For example, calculate the ratio of the tensile strength parameter to the cohesion parameter as 300÷1.5 = 200. According to past engineering experience and relevant research, the bolt spacing constraint range is generally related to this ratio. Assume that through a large amount of data statistics and analysis, the relationship between the bolt spacing d and this ratio k is d = 0.01k ± 0.2. Substitute k = 200 into it, and the bolt spacing constraint range is obtained as 1.8 - 2.2 meters.

[0048] Then, combine the curvature radius parameter of the crown arch area in the 3D geological model to calculate the coupling coefficient of the bolt installation height and the crown arch layer thickness. For example, measure the curvature radius of the crown arch area through the 3D geological model as 10 meters. According to relevant theories and empirical formulas, the coupling coefficient λ is related to the curvature radius R and the bolt spacing d. Assume the empirical formula is λ = 0.5 + 0.05R / d. Substitute R = 10 meters and d take the middle value 2.0 meters of the bolt spacing constraint range into it, and the coupling coefficient λ = 0.5 + 0.05×10÷2 = 0.75 is obtained.

[0049] Next, based on the coupling coefficient and the bolt spacing constraint range, dynamically match the optimal solution of the support density in the historical construction case set. For example, in the historical construction case set, find the cases similar to the current coupling coefficient and bolt spacing constraint range. Assume that 3 cases are found and record their support densities respectively. The support density of Case 1 is 3 bolts per square meter, the support density of Case 2 is 3.5 bolts per square meter, and the support density of Case 3 is 4 bolts per square meter. Through comprehensive analysis and evaluation, select the support density of Case 2 with the best support effect and relatively low cost as the optimal solution, that is, 3.5 bolts per square meter.

[0050] Finally, perform an intersection operation on the matching results and the blasting vibration safety distance constraint in the construction method database to generate the feasible region interval of the initial height parameter of the crown arch layer. The blasting vibration safety distance constraint in the construction method database is that the thickness of the crown arch layer is not less than 2.0 meters. According to the optimal solution of the support density and relevant mechanical calculations, the minimum value of the crown arch layer thickness that meets the support requirements is 2.5 meters. Take the larger value of these two values, and determine that the feasible region interval of the initial height parameter of the crown arch layer is not less than 2.5 meters. Considering the convenience of construction and other factors, the initial height parameter of the crown arch layer is finally determined to be 3.0 meters.

[0051] Step S130: According to the operation radius parameter in the construction machinery parameter set and the layered height parameter set, use the octree spatial subdivision algorithm to generate the transverse partition parameter set and the axial block parameter set.

[0052] In this embodiment, step S130 may include: Step S131: Based on the crown arch layer height parameter and the rock anchor beam layer height parameter in the layered height parameter set, determine the horizontal projection contour data of each layer of the three-dimensional geological model.

[0053] For example, it has been determined that the crown arch layer height parameter is 3.0 meters, and the rock anchor beam layer height parameter is divided into two layers. The first layer is 4.0 meters, and the second layer is 6.0 meters. Using the three-dimensional geological model, through the projection function of the software, project each layer vertically onto the horizontal plane to obtain the horizontal projection contour of each layer.

[0054] For the crown arch layer, in the three-dimensional geological model, project vertically downward from the top of the crown arch layer. The software tool in the three-dimensional geological model will automatically identify the boundary of the crown arch layer and project it onto the horizontal plane to form a closed contour. Through the measurement tool, obtain the boundary coordinate points of this contour. For example, the boundary coordinate points of the horizontal projection contour of the crown arch layer are (0, 0), (10, 0), (10, 8), (0, 8), etc. Connect the above coordinate points to determine the horizontal projection contour of the crown arch layer.

[0055] For the first layer of the rock anchor beam layer, perform the same vertical projection operation. Since the structure of the rock anchor beam layer is relatively complex, it may be necessary to locally magnify and finely process the three-dimensional geological model to accurately identify its boundary. For example, the boundary coordinate points of the horizontal projection contour of the first layer of the rock anchor beam layer are (2, 2), (8, 2), (8, 6), (2, 6), etc. For the second layer of the rock anchor beam layer, perform projection and coordinate acquisition in the same way to obtain the boundary coordinate points of its horizontal projection contour.

[0056] Step S132: According to the trolley width parameter and the operation radius parameter in the construction machinery parameter set, calculate the maximum allowable transverse partition width corresponding to each layer.

[0057] In this embodiment, the width parameter and the operation radius parameter of the jumbo are clearly recorded in the construction machinery parameter set. Assume that the width of the jumbo is 3.0 meters and the operation radius is 6.0 meters. To ensure that the jumbo can operate smoothly within the lateral partition, the calculation of the maximum allowable lateral partition width needs to fully consider the operation range of the jumbo. Generally speaking, the maximum allowable lateral partition width should be less than or equal to twice the operation radius of the jumbo, and at the same time, the width of the jumbo itself and the necessary operation space should be considered.

[0058] During specific calculation, twice the operation radius of the jumbo can be used as a preliminary reference value first, that is, 2 × 6.0 meters = 12.0 meters. However, in order to reserve enough operation space for the jumbo and avoid the jumbo colliding with the partition boundary during operation, this value needs to be adjusted. Considering that the width of the jumbo is 3.0 meters and at least 0.5 meters of operation space is required on each side, a total of 1.0 meter of additional space is needed on both sides. Therefore, the maximum allowable lateral partition width is 12.0 meters - 1.0 meter = 11.0 meters.

[0059] For each layer, due to the possible differences in the geological conditions and construction requirements of different layers, fine-tuning also needs to be carried out according to the actual situation. For example, due to the special structure of the crown arch layer, the lateral partition width may need to be appropriately reduced to ensure construction safety. After comprehensive evaluation, it is determined that the maximum allowable lateral partition width of the crown arch layer is 10.0 meters; the geological conditions of the first and second layers of the rock anchor beam layer are relatively stable, and the maximum allowable lateral partition width can be taken as 11.0 meters.

[0060] Step S133: Perform spatial grid division on the horizontal projection contour data through the octree spatial subdivision algorithm to generate the initial lateral partition width parameter and the initial axial block length parameter, where the initial lateral partition width parameter is less than or equal to the maximum allowable lateral partition width.

[0061] Step S1331: Extract the set of irregular boundary vertex coordinates in the horizontal projection contour data.

[0062] In this embodiment, the horizontal projection contour data of each layer has been obtained. Next, the measurement and analysis functions of the 3D geological model can be used to extract the irregular boundary vertex coordinates of the horizontal projection contour. Taking the crown arch layer as an example, through the coordinate measurement tool in the 3D geological model, each vertex on the boundary of the horizontal projection contour of the crown arch layer is marked. For example, after measurement, the set of irregular boundary vertex coordinates of the horizontal projection contour of the crown arch layer is obtained, such as (0, 0), (2, 1), (5, 2), (8, 1), (10, 0), etc. For the first and second layers of the rock anchor beam layer, the vertex coordinate extraction is also carried out in the same way, and the respective sets of irregular boundary vertex coordinates are obtained.

[0063] Step S1332: Set the recursive depth threshold of the octree subdivision algorithm according to the spatial distribution density of the vertex coordinate set.

[0064] In this embodiment, first, the extracted vertex coordinate set is analyzed to count the distribution quantity of vertices in different regions. For example, in the horizontal projection contour of the crown arch layer, it is found that the vertices are more densely distributed in the area near the center of the cavern, while the vertices in the edge area are relatively sparse. By calculating the ratio of the number of vertices in different regions to the area of the region, the spatial distribution density of the vertex coordinate set is obtained.

[0065] Then, the recursive depth threshold of the octree subdivision algorithm is set according to the spatial distribution density. If the vertex distribution density is large, it indicates that the geological structure and shape of this region are relatively complex and more refined division is required, so the recursive depth threshold can be set higher; on the contrary, if the vertex distribution density is small, the recursive depth threshold can be set lower. After comprehensive evaluation, for the crown arch layer, since the vertex distribution density in its central region is large, the recursive depth threshold of the octree subdivision algorithm is set to 5; for the first and second layers of the rock anchor beam layer, according to their vertex distribution density conditions, the recursive depth thresholds are set to 4 and 3 respectively.

[0066] Step S1333: During each recursive subdivision process, detect the mutation mark of the surrounding rock grade field contained in the grid cell.

[0067] In this embodiment, during the recursive process of the octree spatial subdivision algorithm, each grid cell is carefully inspected. Specifically, the surrounding rock grade field information corresponding to each grid cell can be obtained from the 3D geological model and the construction method database. After the initial division of the large grid cells, the recursive subdivision begins. For example, for a large grid cell in the crown arch layer, it contains multiple small regions. By querying the construction method database, it is found that the surrounding rock grade in some regions within this large grid cell is class III, while in another part it is class IV, and this change in the surrounding rock grade is the mutation mark.

[0068] Such detection is repeated every time a newly generated grid cell is recursively subdivided. If a mutation mark of the surrounding rock grade field is detected in a certain grid cell, it indicates that there are obvious differences in the geological conditions of this region and special treatment is required.

[0069] Step S1334: When a mutation mark is detected, terminate the recursive subdivision of the current grid cell and lock the grid boundary.

[0070] In this embodiment, once a mutation mark of the surrounding rock grade field is detected in a certain grid cell, the further recursive dissection of this grid cell is immediately stopped. For example, in a grid cell of the top arch layer, it is detected that the surrounding rock changes from class III to class IV, and at this time, the subdivision operation of this grid cell is stopped. Then, using the boundary locking function in the model, the boundary of this grid cell is fixed to ensure that subsequent construction plans can fully consider such changes in geological conditions. This can avoid over-subdivision in areas with large differences in geological conditions and ensure the safety and rationality of construction.

[0071] Step S1335: Continuously perform spatial dissection on the grid cells that have not triggered mutation marks until the recursive depth threshold is reached.

[0072] In this embodiment, for those grid cells in which no mutation marks of the surrounding rock grade field are found during the detection process, the recursive dissection operation continues. Taking a grid cell in the first layer of the rock anchor beam layer as an example, the surrounding rock grade within this cell is all class III, and no mutation marks are detected. According to the rules of the octree dissection algorithm, this grid cell is further subdivided into smaller sub-grid cells.

[0073] After each subdivision, it is necessary to check whether the previously set recursive depth threshold has been reached. If it has not been reached, continue with the next round of subdivision operations. For example, the recursive depth threshold for the first layer of the rock anchor beam layer is 4. After 3 recursive subdivisions, the threshold has not been reached, so the 4th subdivision continues. Stop dissecting this grid cell until the recursive depth threshold is reached.

[0074] Step S1336: Compare the width parameters of the grid cells in the final dissection result with the maximum allowable lateral partition width level by level, and generate the initial lateral partition width parameters after removing the over-limit cells.

[0075] In this embodiment, after the octree spatial dissection is completed, a series of grid cells are obtained. Measure and record the width parameters of each grid cell. Compare the above grid cell width parameters with the maximum allowable lateral partition width of each layer calculated previously one by one.

[0076] Taking the top arch layer as an example, the maximum allowable lateral partition width is 10.0 meters. For the grid cells obtained by dissecting the top arch layer, check their widths one by one. If the width of a certain grid cell exceeds 10.0 meters, remove this cell from the result. After comparison and screening, the remaining grid cell width parameters form the initial lateral partition width parameters of the top arch layer. The same method is applied to the first and second layers of the rock anchor beam layer to obtain their initial lateral partition width parameters respectively. For example, after screening, the initial lateral partition width parameters obtained for the top arch layer are 3.0 meters, 4.0 meters, 3.0 meters, etc.

[0077] Step S1337: Extract the reference direction vector of the chamber axis in the three-dimensional geological model according to the axial extension length of the horizontal projection contour data.

[0078] In this embodiment, in the three-dimensional geological model, measure the axial extension length of the horizontal projection contour data. Taking the horizontal projection contour of the entire chamber as an example, use the measurement tool in the three-dimensional geological model to measure from one end of the chamber to the other end, and the obtained axial extension length is 50.0 meters.

[0079] To determine the reference direction vector of the chamber axis, mark the starting point and the ending point of the chamber axis in the three-dimensional geological model. By calculating the coordinate difference between the starting point and the ending point, a vector is obtained. For example, if the starting point coordinates are (0, 0, 0) and the ending point coordinates are (0, 0, 50), then the reference direction vector is (0, 0, 1), indicating that the chamber axis extends along the positive z-axis direction.

[0080] Step S1338: Determine the minimum cutting step length of the axial block based on the telescopic stroke parameter of the robotic arm in the construction machinery parameter set.

[0081] The telescopic stroke parameter of the robotic arm is recorded in the construction machinery parameter set. Assume that the maximum telescopic stroke of the robotic arm is 5.0 meters. To ensure that the robotic arm can fully play its role in the axial block operation and considering the construction accuracy and efficiency, determine the minimum cutting step length of the axial block.

[0082] Generally speaking, the minimum cutting step length should be less than the maximum telescopic stroke of the robotic arm. After comprehensive consideration, the minimum cutting step length of the axial block is set to 40% of the maximum telescopic stroke of the robotic arm, that is, 5.0 meters × 40% = 2.0 meters. Such a setting can ensure that the robotic arm has enough movement space in each cutting operation and avoid low construction efficiency caused by too small cutting step length.

[0083] Step S1339: During the octree space subdivision process, axially extend and cut the horizontal projection contour along the reference direction vector to generate an initial mesh unit containing an axial cutting plane.

[0084] Based on the octree space subdivision, axially extend and cut the horizontal projection contour along the previously determined reference direction vector (0, 0, 1) of the chamber axis. Taking the horizontal projection contour of the top arch layer as an example, start from the starting position of the top arch layer and cut according to the minimum cutting step length of 2.0 meters for the axial block.

[0085] Each cut generates an initial mesh element that includes an axial cutting plane. For example, at the first cutting position in the top arch layer, the horizontal projection profile is extended 2.0 meters in the positive z-axis direction to form a mesh element with a certain height. The bottom of this mesh element is the horizontal projection profile of the top arch layer, the top is the new profile after axial extension, and the sides are the axial cutting planes. In this way, cuts are made sequentially until the axial end point of the cavern is reached, generating a series of initial mesh elements that include axial cutting planes.

[0086] Step S13310: According to the extension length parameter of the initial mesh element in the reference direction, extract the candidate values of the axial block length that meet the minimum cutting step constraint.

[0087] Measure and record the extension length of the generated initial mesh element in the reference direction (z-axis direction). For example, among a series of initial mesh elements generated in the top arch layer, the measured extension lengths in the z-axis direction are 1.5 meters, 2.0 meters, 2.5 meters, 3.0 meters, etc.

[0088] According to the minimum cutting step constraint (2.0 meters) of the axial block, screen out the candidate values of the axial block length that meet the conditions from the above extension length parameters. In the above example, the candidate values of the axial block length that meet the conditions are 2.0 meters, 2.5 meters, 3.0 meters, etc.

[0089] Step S13311: Combine the axial block records of historical cases in the construction method database, and screen out the axial block length parameter that matches the current surrounding rock grade field as the initial axial block length parameter.

[0090] In the construction method database, search for the axial block records of historical cases. The above axial block records contain the axial block length information under different surrounding rock grades. Taking the surrounding rock grade of the current top arch layer as Class III as an example, screen out the historical cases with a surrounding rock grade of Class III in the database.

[0091] Then, extract the axial block length parameter from the above historical cases and compare it with the previously obtained candidate values of the axial block length. Assuming that the axial block length parameters of the historical cases with a surrounding rock grade of Class III are mainly concentrated between 2.0 meters and 2.5 meters, combining with the candidate values, screen out 2.0 meters and 2.5 meters as the axial block length parameters that match the current surrounding rock grade field. After comprehensive evaluation and analysis, determine the initial axial block length parameter as 2.0 meters. For the first and second layers of the rock anchor beam layer, also use the same method to determine the initial axial block length parameters according to their respective surrounding rock grades and candidate values, combined with the historical case records.

[0092] Step S134: Dynamically reduce the initial lateral partition width parameter according to the maximum span standard corresponding to the surrounding rock grade field in the construction method database to generate a lateral partition width parameter that meets the surrounding rock stability constraint.

[0093] The construction method database records the maximum span standards corresponding to different surrounding rock grades. For the crown arch layer, its surrounding rock grade is class III, and the corresponding maximum span standard is 8.0 meters. Compare the previously generated initial lateral partition width parameter with this maximum span standard.

[0094] Suppose the initial lateral partition width parameters of the crown arch layer are 3.0 meters, 4.0 meters, 3.0 meters, etc. Although the above parameters do not exceed the maximum allowable lateral partition width itself, in order to ensure the stability of the surrounding rock, it is necessary to further check whether they meet the maximum span standard. Since all the initial lateral partition width parameters are less than 8.0 meters, there is no need to reduce them for the time being.

[0095] However, for the second layer of the rock anchor beam layer, its surrounding rock grade is class IV, and the corresponding maximum span standard is 6.0 meters. The initial lateral partition width parameters of this layer are 4.0 meters, 5.0 meters, 3.0 meters, etc. Among them, although 5.0 meters is less than the maximum allowable lateral partition width, it exceeds the maximum span standard of class IV surrounding rock. Therefore, dynamically reduce this parameter to less than 6.0 meters, such as adjusting it to 5.5 meters. After such dynamic reduction operations, a lateral partition width parameter that meets the surrounding rock stability constraint is obtained.

[0096] Step S135: Iteratively optimize the initial axial block length parameter based on the daily maximum advance parameter and the mechanical utilization rate parameter in the construction method database to generate an axial block length parameter that meets the mechanical operation cycle constraint.

[0097] In this embodiment, the construction method database records the daily maximum advance parameter and the mechanical utilization rate parameter. Suppose the daily maximum advance parameter is 2.5 meters and the mechanical utilization rate parameter is 80%. Take the initial axial block length parameter of 2.0 meters of the crown arch layer as an example for iterative optimization.

[0098] First, calculate the actually achievable daily advance according to the mechanical utilization rate parameter. The actual daily advance = daily maximum advance parameter × mechanical utilization rate parameter, that is, 2.5 meters × 80% = 2.0 meters. Since the initial axial block length parameter of 2.0 meters is exactly equal to the actually achievable daily advance, there is no need to adjust it for the time being.

[0099] For the first layer of the rock-anchored beam layer, the initial axial block length parameter is 2.2 meters. Since 2.2 meters is greater than the actual achievable daily advance of 2.0 meters, it needs to be reduced. Through iterative adjustment, the axial block length parameter is adjusted to 2.0 meters to meet the mechanical operation cycle constraint. After such iterative optimization operations, the axial block length parameter that meets the mechanical operation cycle constraint is obtained.

[0100] Step S136: When there is a fault fracture zone in the construction method database, synchronously and proportionally reduce the transverse partition width parameter and the axial block length parameter to generate the adjusted transverse partition width parameter and axial block length parameter in the fault zone.

[0101] In this embodiment, upon inspection in the construction method database, it is found that there is a fault fracture zone in a certain area of the cavern. After investigation and analysis, the scope and geological characteristics of the fault fracture zone are determined. To ensure the construction safety in the fault fracture zone area, it is necessary to synchronously and proportionally reduce the transverse partition width parameter and the axial block length parameter in this area.

[0102] Suppose the original transverse partition width parameter in the area where the fault fracture zone is located is 4.0 meters, and the axial block length parameter is 2.0 meters. According to the severity of the fault fracture zone and relevant safety standards, the proportional reduction coefficient is determined to be 0.6.

[0103] Reduce the transverse partition width parameter. The reduced transverse partition width parameter = 4.0 meters × 0.6 = 2.4 meters. Reduce the axial block length parameter. The reduced axial block length parameter = 2.0 meters × 0.6 = 1.2 meters. In this way, the adjusted transverse partition width parameter and axial block length parameter in the fault zone are generated.

[0104] Step S137: Combine the transverse partition width parameter with the adjusted transverse partition width parameter in the fault zone into a transverse partition parameter set, combine the axial block length parameter with the adjusted axial block length parameter in the fault zone into an axial block parameter set, and insert the support marking parameter into the combined parameter set.

[0105] In this embodiment, the transverse partition width parameter in the normal area and the adjusted transverse partition width parameter in the fault zone can be integrated. For example, the transverse partition width parameters in the normal area are 3.0 meters, 4.0 meters, etc., and the adjusted transverse partition width parameter in the fault zone is 2.4 meters. Combine them into a transverse partition parameter set, namely 3.0 meters, 4.0 meters, 2.4 meters, etc.

[0106] Similarly, the axial block length parameters of the normal area and the adjusted axial block length parameters of the fault area are combined into an axial block parameter set. The axial block length parameter of the normal area is 2.0 meters, and the adjusted axial block length parameter of the fault area is 1.2 meters. The combined axial block parameter set is 2.0 meters, 1.2 meters, etc.

[0107] To better guide the construction, support marking parameters are inserted into the combined parameter set. For the parameters of the fault area, special support markings such as "reinforced support" are inserted; for the parameters of the normal area, ordinary support markings such as "conventional support" are inserted. In this way, the lateral zoning parameter set and the axial block parameter set contain complete construction information, providing a detailed basis for subsequent construction planning.

[0108] Step S140: Based on the spatial coordinate attributes, surrounding rock grade attributes, lateral zoning parameter set, and axial block parameter set of the 3D geological model, perform composite coding processing to generate a unique construction block coding set containing hierarchical identifiers, axial identifiers, and block identifiers.

[0109] In this embodiment, in the local underground chamber project, generating a unique construction block coding set helps to accurately manage and identify each construction block, improving the organization efficiency and coordination of the construction.

[0110] Step S141: Analyze the spatial coordinate attribute data of each layer in the 3D geological model, and combine the surrounding rock category parameters in the surrounding rock grade attributes to generate a hierarchical number containing a Roman numeral sequence and a special support marking, where the special support marking is dynamically generated according to the stability threshold in the surrounding rock grade attributes.

[0111] In this embodiment, in the 3D geological model, a coordinate measurement tool is used to obtain the spatial coordinate attribute data of each layer. For example, the spatial coordinate range of the top arch layer is 0 - 3.0 meters in the z-axis direction and has specific boundary coordinates in the x-y plane. At the same time, the surrounding rock category parameters in the surrounding rock grade attributes of each layer are obtained from the construction method database. The surrounding rock grade of the top arch layer is class III.

[0112] According to the order of the layers, each layer is numbered using a Roman numeral sequence. The top arch layer is the first layer, the first layer of the rock anchor beam layer is the second layer, the second layer of the rock anchor beam layer is the third layer, etc.

[0113] For special support markings, they are dynamically generated based on the stability thresholds in the surrounding rock grade attributes. In the construction method database, the stability thresholds for Class III surrounding rock are set as certain deformation amounts and stress values. Through the geomechanical analysis and simulation of the crown arch layer, it is found that the stress values in some areas of this layer are close to the stability thresholds. Therefore, a special support marking "SP" is added to the layer numbering of the crown arch layer, and the generated layer number is "Ⅰ-SP". For the first and second layers of the rock anchor beam layer, in the same way, according to their surrounding rock grades and stability conditions, the layer numbers "Ⅱ" and "Ⅲ-SP" are generated respectively.

[0114] Step S142: Based on the left area width parameter, middle area width parameter, and right area width parameter in the lateral partition parameter set, combined with the azimuth coordinate data in the spatial coordinate attributes of the 3D geological model, generate an azimuth feature code, which includes the spatial topological relationship of the left area, middle area, and right area and the mechanical movement path association parameters.

[0115] In this embodiment, during the construction of the underground cavern in this project, a lateral partition parameter set has been obtained, which includes the left area width parameter, middle area width parameter, and right area width parameter. Assume that the left area width parameter is 3 meters, the middle area width parameter is 4 meters, and the right area width parameter is 3 meters. At the same time, the 3D geological model provides detailed azimuth coordinate data, and the above data clarify the specific position and direction of the cavern in space.

[0116] First, determine the reference direction of the cavern according to the azimuth coordinate data. For example, take the axis direction of the cavern as the longitudinal direction, and the direction perpendicular to the axis as the lateral direction. In the lateral direction, divide according to the width parameters of the left area, middle area, and right area. Obtain the boundary coordinate points of each area from the 3D geological model to accurately define the scope of each area.

[0117] To generate the azimuth feature code, this embodiment needs to analyze the spatial topological relationship of the left area, middle area, and right area. The left area is located on the left side of the cavern, the middle area is in the middle, and the right area is on the right side. They are adjacent and juxtaposed. At the same time, considering the movement path of the construction machinery between different areas, this embodiment needs to determine the shortest path and possible movement directions for the machinery to move from one area to another.

[0118] For the azimuth feature code, this embodiment adopts a set coding rule. Use the letter "L" to represent the left area, "M" to represent the middle area, and "R" to represent the right area. To reflect the mechanical movement path association parameters, this embodiment adds numbers to the code to represent the order and direction of movement. For example, if the machinery moves from the left area to the middle area and then from the middle area to the right area, the azimuth feature code can be represented as "L-1M-2R", where "1" and "2" represent the order of movement.

[0119] During actual encoding generation, this embodiment will be adjusted according to specific construction plans and mechanical layout situations. For example, if the machinery needs to perform round-trip operations between the left area and the middle area multiple times in the construction plan, the encoding may become "L-1M-2L-3M-4R". Through such an encoding method, the azimuth feature encoding can accurately reflect the spatial topological relationships of the left area, the middle area, and the right area, as well as the mechanical movement path correlation parameters, providing clear guidance for construction organization and management.

[0120] Step S143: According to the daily footage block length parameter and the volume constraint parameter in the axial block parameter set, combined with the axial spatial coordinate attributes of the three-dimensional geological model, generate an axial sequence encoding including an excavation sequence priority mark and a resource allocation correlation mark.

[0121] In this embodiment, the axial block parameter set includes a daily footage block length parameter and a volume constraint parameter. Suppose the daily footage block length parameter is 2 meters, and the volume constraint parameter stipulates that the volume of each block cannot exceed 100 cubic meters. At the same time, the three-dimensional geological model provides axial spatial coordinate attributes, clarifying the position and scope of the cavern in the axial direction.

[0122] First, according to the daily footage block length parameter, the cavern is divided into blocks in the axial direction. Starting from the starting end of the cavern, a block is divided every 2 meters. During the division process, the volume constraint parameter needs to be considered. For each block, its volume is calculated according to its spatial scope in the three-dimensional geological model. If the volume of a certain block exceeds 100 cubic meters, the block needs to be further subdivided to meet the volume constraint conditions.

[0123] To determine the excavation sequence priority mark, this embodiment will consider multiple factors comprehensively. For example, blocks with better geological conditions and closer to the cave entrance can be excavated first. In this project, through the analysis of the geological data in the three-dimensional geological model, it is found that the surrounding rock grades of the first few blocks close to the cave entrance are higher and the stability is better. Therefore, the excavation sequence priority mark of the above-mentioned blocks is marked as "1". For blocks with complex geological conditions and fault fracture zones, their excavation sequence priority marks will be relatively lower, such as marked as "3".

[0124] The resource allocation correlation mark is related to the construction resources required for each block. Different blocks will have different requirements for construction resources (such as machinery and equipment, manpower, materials, etc.) due to factors such as geological conditions and volume sizes. For blocks that require large machinery and equipment for excavation, their resource allocation correlation mark may be "heavy equipment"; while for some small blocks, only small machinery and equipment and a small amount of manpower are required to complete the excavation, and their resource allocation correlation mark may be "light equipment".

[0125] When generating the axial sequence code, in this embodiment, the excavation sequence priority mark and the resource allocation association mark are combined with the number of the divided block. For example, the excavation sequence priority mark of the first divided block is "1", and the resource allocation association mark is "light equipment", and its axial sequence code can be expressed as "1-L1", where "1" represents the excavation sequence priority, and "L1" represents the first divided block with the resource allocation association mark of light equipment. In this way, the axial sequence code containing the excavation sequence priority mark and the resource allocation association mark is generated for each axial divided block, so as to reasonably arrange the excavation sequence and allocate resources during the construction process.

[0126] Step S144: Perform a composite splicing process on the hierarchical number, the azimuth feature code, and the axial sequence code to generate a set of unique construction block codes that integrate the surrounding rock grade attribute, the spatial coordinate attribute, and the construction parameters, and establish a two-way mapping relationship between the set of unique construction block codes and the 3D model components, so that the hierarchical identifier, the axial identifier, and the block identifier of each block correspond one-to-one with the model spatial coordinates.

[0127] After obtaining the hierarchical number, the azimuth feature code, and the axial sequence code, this embodiment performs a composite splicing process. Taking a specific construction block in this project as an example, assume that its hierarchical number is "Ⅱ-SP", the azimuth feature code is "L-1M-2R", and the axial sequence code is "1-L1". Splicing these three codes, the obtained unique construction block code is "Ⅱ-SP-L-1M-2R-1-L1".

[0128] In this way, all construction blocks are coded. The code of each construction block integrates the surrounding rock grade attribute (reflected by the special support mark in the hierarchical number), the spatial coordinate attribute (reflected by the position information in the azimuth feature code and the axial sequence code), and the construction parameters (reflected by the excavation sequence priority mark and the resource allocation association mark in the axial sequence code).

[0129] To achieve precise management and positioning of the construction blocks, this embodiment establishes a two-way mapping relationship between the set of unique construction block codes and the 3D model components. In the 3D geological model, each construction block corresponds to a specific 3D spatial area with clear spatial coordinates. This embodiment associates the unique code of each construction block with the spatial coordinates of this block in the 3D model.

[0130] For example, in a 3D model, the spatial coordinate range of the construction block "Ⅱ-SP-L-1M-2R-1-L1" is determined by a coordinate measurement tool to be between 5 and 8 meters for the x coordinate, between 3 and 6 meters for the y coordinate, and between 4 and 6 meters for the z coordinate. Bind this spatial coordinate range to the code of this construction block to establish a two-way mapping relationship. In this way, when this embodiment inputs this code into the construction management system, it can quickly locate the corresponding construction block in the 3D model; conversely, when this embodiment selects a certain construction block in the 3D model, its corresponding unique code can also be obtained. Through this two-way mapping relationship, the hierarchical identifier, axial identifier, and block identifier of each block correspond one-to-one with the model space coordinates, facilitating construction personnel to carry out construction planning, progress management, and quality control.

[0131] Step S145: When it is detected that the surrounding rock grade attribute in the construction method database is updated or the lateral partition parameter set and the axial block parameter set are adjusted, dynamically correct the support marking parameter in the azimuth feature code and the resource configuration association marking in the axial sequence code, and synchronously update the associated fields in the unique construction block code set.

[0132] In this embodiment, during the construction process, the construction method database can be updated according to real-time geological exploration data and construction conditions. When it is detected that the surrounding rock grade attribute is updated, for example, the surrounding rock grade of a certain construction block changes from class Ⅲ to class Ⅳ, indicating that the geological conditions of this block have changed and the stability has decreased, and stronger support is required.

[0133] For the support marking parameter in the azimuth feature code, originally this block may correspond to "conventional support", and now it needs to be corrected to "enhanced support". At the same time, due to the change in the surrounding rock grade, the excavation difficulty of this block may increase, and the required construction resources will also change accordingly. For the resource configuration association marking in the axial sequence code, originally it may be "light equipment", and now it needs to be adjusted to "heavy equipment".

[0134] When the lateral partition parameter set or the axial block parameter set is adjusted, for example, due to the discovery of a new fault fracture zone, the lateral partition width parameter and the axial block length parameter are synchronously reduced in equal proportion. This will cause changes in the spatial range and construction conditions of the construction block. For the azimuth feature code, it may be necessary to re-determine the spatial topological relationship and the mechanical movement path association parameters of the left area, middle area, and right area, and accordingly adjust the code. For the axial sequence code, the excavation sequence priority marking and the resource configuration association marking may also need to be re-evaluated and corrected according to the new block situation.

[0135] After completing the correction of the azimuth feature encoding and the axial sequence encoding, synchronously update the associated fields in the set of unique construction block codes. For example, the original unique construction block code was "Ⅱ-SP-L-1M-2R-1-L1", and after correction, it becomes "Ⅱ-SP-L-1M-2R-1-H1", where "H1" indicates that the resource allocation association mark is adjusted to heavy equipment. Through the above dynamic correction and update mechanism, the set of unique construction block codes can always reflect the latest information of the construction blocks.

[0136] Step S150: Perform construction sequence optimization processing on the set of unique construction block codes through a multi-objective optimization algorithm to generate a Pareto optimal solution set including a construction period optimization path, a cost optimization path, and a stability optimization path, and dynamically adjust the construction sequence parameters in the Pareto optimal solution set according to the real-time monitored surrounding rock deformation data.

[0137] In this underground chamber project, the optimization of the construction sequence is crucial for improving construction efficiency, reducing costs, and ensuring construction safety. By processing the set of unique construction block codes through a multi-objective optimization algorithm, multiple optimization paths can be obtained, and dynamic adjustment is made according to the real-time monitored data to ensure the smooth progress of the construction process.

[0138] Step S151: Extract the surrounding rock grade parameters, resource allocation parameters, and spatial position parameters of each block according to the hierarchical identifier and azimuth feature encoding in the set of unique construction block codes. Among them, the surrounding rock grade parameters are derived from the surrounding rock grade field in the construction method database, and the resource allocation parameters include the mechanical operation radius parameter and the support time sequence parameter.

[0139] In this embodiment, in the set of unique construction block codes, the hierarchical identifier and azimuth feature encoding contain rich information. Taking the code "Ⅱ-SP-L-1M-2R-1-H1" as an example, the hierarchical identifier "Ⅱ-SP" indicates that this block is located on the second layer and requires special support, and the corresponding surrounding rock grade parameter of this layer can be found from the construction method database to be class Ⅲ.

[0140] The azimuth feature encoding "L-1M-2R" provides the position information of this block in the horizontal partition. Combining with the three-dimensional geological model, its spatial position parameters can be determined. For example, through the coordinate system in the model, the coordinate range of this block on the x-y plane can be determined.

[0141] In terms of resource allocation parameters, "1-H1" in the axial sequence coding indicates that the excavation order priority of this block is 1, and the resource allocation association mark is "heavy equipment". The mechanical operation radius parameter required for this block can be obtained from the construction method database. Suppose the operation radius of the corresponding three-boom rock drilling jumbo is 6 meters. At the same time, according to the surrounding rock grade and construction requirements, the support time sequence parameters are determined. For example, the initial support is carried out within 24 hours after excavation is completed.

[0142] In this way, each code in the set of unique construction block codes is analyzed, and the surrounding rock grade parameters, resource allocation parameters, and spatial position parameters of each block are extracted to provide basic data for subsequent multi-objective optimization.

[0143] Step S152: Based on the extracted surrounding rock grade parameters, resource allocation parameters, and spatial position parameters, establish a multi-objective optimization function set including a construction period objective function, an economic cost objective function, a resource balance objective function, and a surrounding rock deformation objective function.

[0144] In this embodiment, the construction period objective function mainly considers the excavation time and support time of each construction block. The excavation time is related to the volume of the block, the surrounding rock grade, and the efficiency of the construction equipment. For example, for a block with a surrounding rock grade of class Ⅲ, according to construction experience and equipment performance, the excavation time per cubic meter is about 0.5 hours. The support time is determined according to the support time sequence parameters and the support method. Suppose this block needs bolt support, and the installation time of each bolt is 0.2 hours. This block needs to install 100 bolts, then the support time is 20 hours. Add up the excavation time and support time of all construction blocks to obtain the value of the construction period objective function.

[0145] The economic cost objective function includes mechanical equipment rental costs, material costs, and labor costs. The mechanical equipment rental costs are calculated according to the type of equipment used and the usage time. For example, the rental cost of a three-boom rock drilling jumbo is 500 yuan per hour, and it is used for 100 hours, then the rental cost is 50,000 yuan. The material costs include the procurement costs of materials such as explosives, bolts, and concrete. The labor costs are calculated according to the number of construction workers and the working time. Add up the above costs to obtain the value of the economic cost objective function.

[0146] The resource balance objective function aims to make the use of various resources in the construction process as balanced as possible, avoiding the situation of excessive concentration or idleness of resources. For example, by calculating the number of mechanical equipment used and the number of construction workers in each time period, evaluate whether the resource allocation is reasonable. If too many mechanical equipment are used in a certain time period while idle in other time periods, it means that the resource allocation is unbalanced and needs to be adjusted. The resource balance objective function can be measured by calculating the variance of resource use. The smaller the variance, the more balanced the resource allocation.

[0147] The surrounding rock deformation objective function mainly considers the deformation of the surrounding rock during the construction process. By arranging multiple displacement sensors in the tunnel, the deformation data of the surrounding rock is monitored in real time. According to the surrounding rock grade and construction experience, an allowable deformation threshold is set. When the deformation of the surrounding rock exceeds the threshold, it will affect the construction safety and the stability of the tunnel. The surrounding rock deformation objective function can be expressed as the absolute value of the difference between the actual deformation amount and the allowable deformation threshold. The smaller the difference, the better the control of the surrounding rock deformation.

[0148] Combining these four objective functions together forms a multi-objective optimization function set. In the actual optimization process, it is necessary to comprehensively consider these four objectives to find an optimal construction sequence plan.

[0149] Step S153: Use the NSGA-II algorithm to perform parallel optimization processing on the multi-objective optimization function set to generate a non-dominated solution set containing multiple Pareto front solutions.

[0150] In this embodiment, the NSGA-II algorithm is a commonly used multi-objective optimization algorithm that can find a set of optimal solutions, that is, Pareto front solutions, among multiple objectives. In this project, this embodiment uses the NSGA-II algorithm to perform parallel optimization processing on the multi-objective optimization function set.

[0151] First, initialize a population. Each individual in the population represents a construction sequence plan. Each plan contains the construction sequence of each block in the unique construction block coding set. For example, a construction sequence plan may be the arrangement of blocks such as "Ⅱ-SP-L-1M-2R-1-H1", "Ⅰ-SP-R-1M-2L-2-L1" in a certain order.

[0152] During the iterative process of the algorithm, new individuals are generated through crossover and mutation operations. The crossover operation is to exchange part of the genes of two parent individuals to generate offspring individuals. The mutation operation is to randomly change a certain gene of an individual. Through continuous iteration, the individuals in the population can be continuously optimized so that each individual can reach the optimal as much as possible on multiple objective functions.

[0153] After each iteration ends, non-dominated sorting is performed on the individuals in the population. Non-dominated sorting is to divide the individuals in the population into different levels. The higher the level, the better the performance of the individuals on multiple objective functions. Through non-dominated sorting, the non-dominated individuals in the current population are selected to form Pareto front solutions. After multiple iterations, a non-dominated solution set containing multiple Pareto front solutions is finally generated. The above solutions achieve a balance among multiple objectives such as construction period, cost, resource balance, and surrounding rock deformation, and no solution is superior to other solutions in all objectives.

[0154] Step S154: Optimize the mechanical movement path for the Pareto front solutions through the Traveling Salesman Problem model, calculate the comprehensive scores of each Pareto front solution, and generate a priority ranking result.

[0155] The Traveling Salesman Problem model is a classic optimization problem used to solve the problem of finding the shortest path among multiple locations. In this project, in this embodiment, the construction blocks are regarded as the locations that the traveling salesman needs to visit, and the mechanical movement path is regarded as the itinerary of the traveling salesman.

[0156] For each Pareto front solution, determine the order of the construction blocks that the machine needs to visit according to the construction sequence. Then, use the Traveling Salesman Problem model, combined with the spatial position parameters of each block in the three-dimensional geological model, to calculate the length of the mechanical movement path between different blocks. For example, by calculating the Euclidean distance between two points, determine the shortest path for the machine to move from one block to another.

[0157] After calculating the mechanical movement path lengths of each Pareto front solution, combine the values of the construction period objective function, economic cost objective function, resource balance objective function, and surrounding rock deformation objective function to conduct a comprehensive evaluation of each solution. A weight can be assigned to each objective function. For example, the weight of the construction period objective function is 0.3, the weight of the economic cost objective function is 0.3, the weight of the resource balance objective function is 0.2, and the weight of the surrounding rock deformation objective function is 0.2. Multiply the value of each objective function by the corresponding weight and then sum them up to obtain the comprehensive score of each solution.

[0158] Rank the Pareto front solutions according to the comprehensive scores. The higher the score of a solution, the better its comprehensive performance on multiple objectives and the higher its priority. In this way, obtain the priority ranking result of the Pareto front solutions to provide a reference for the selection of the construction sequence.

[0159] Step S155: Real-time collect the surrounding rock deformation rate data monitored by the fiber optic sensor. When it is detected that the deformation amount exceeds the threshold in the construction method database, trigger the dynamic adjustment mechanism to recalculate the construction sequence parameters in the Pareto front solution.

[0160] Arrange multiple fiber optic sensors in the cavern to collect the surrounding rock deformation rate data in real time. The above sensors can accurately measure the deformation of the surrounding rock in different directions. Different deformation thresholds corresponding to different surrounding rock grades are set in the construction method database. For example, the deformation threshold for class III surrounding rock is 5 millimeters per day.

[0161] During the construction process, continuously monitor the surrounding rock deformation rate data. Once it is detected that the deformation amount of a certain construction block exceeds the threshold, it indicates that there is a problem with the surrounding rock stability of this block, and the construction sequence needs to be adjusted in a timely manner.

[0162] After the dynamic adjustment mechanism is triggered, first update the surrounding rock deformation objective function values of each construction block according to the real-time monitoring data. Since the surrounding rock deformation situation has changed, the value of this objective function will also change accordingly. Then, substitute the updated objective function values into the multi-objective optimization function set and use the NSGA-II algorithm for optimization again.

[0163] During the re-optimization process, new Pareto front solutions can be generated based on the new objective function values. At the same time, use the traveling salesman problem model to optimize the mechanical movement path of the new Pareto front solutions again, calculate the comprehensive score and generate a new priority ranking result. According to the new priority ranking result, adjust the construction sequence parameters in the Pareto front solutions to ensure that the construction process can continue to be efficient and economical on the premise of ensuring the stability of the surrounding rock.

[0164] Step S1551: Real-time collect the surrounding rock deformation rate data and displacement data of each construction block through fiber optic sensors, and compare them with the deformation threshold parameters in the construction method database.

[0165] In this underground chamber project, in order to accurately grasp the surrounding rock state of each construction block, fiber optic sensors are arranged at each key position in the chamber. The above sensors have the characteristics of high precision and high sensitivity, and can collect the surrounding rock deformation rate data and displacement data in real time and accurately. For example, corresponding fiber optic sensors are installed at different areas of the crown arch layer, each block of the rock anchor beam layer, and the side wall of the chamber.

[0166] After the sensors collect the data, they can transmit it to the data processing center, and then organize and analyze the above data in the data processing center. Taking a specific construction block as an example, assume that this block is located on the second layer of the rock anchor beam layer. At a certain moment, the fiber optic sensor collects that the deformation rate of this block in the past hour is 0.5 mm per hour, and the cumulative displacement data is 3 mm.

[0167] The construction method database sets clear deformation threshold parameters for different surrounding rock grades. For the class III surrounding rock where this block is located, the deformation rate threshold is set at 0.8 mm per hour, and the displacement threshold is set at 5 mm. Compare the collected deformation rate data and displacement data with the above threshold parameters, and find that the current deformation rate and displacement data of this block do not exceed the threshold, indicating that the current surrounding rock stability of this block is in a controllable state.

[0168] However, for another construction block near the fault fracture zone, the deformation rate collected by the sensor is 1.2 millimeters per hour, and the displacement data is 6 millimeters. This area belongs to Class Ⅳ surrounding rock, with a deformation rate threshold of 1 millimeter per hour and a displacement threshold of 4 millimeters. Obviously, both the deformation rate and displacement data of this block exceed the thresholds, indicating that there are problems with the surrounding rock stability of this block and timely measures need to be taken.

[0169] Step S1552: When the surrounding rock deformation rate data exceeds the deformation threshold parameter, extract the layer identifier and azimuth feature code of the affected construction block.

[0170] Once it is found that the surrounding rock deformation rate data of a certain construction block exceeds the deformation threshold parameter, it is necessary to immediately accurately locate and identify the affected construction block. In this project, the set of unique construction block codes provides an effective identification method for this embodiment.

[0171] Taking the construction block near the fault fracture zone mentioned above with an excessive deformation rate as an example, through the system of the data processing center, according to the corresponding relationship between the position information of the sensor and the code of the construction block, extract the unique code of this construction block. Suppose the code is "Ⅲ-SP-R-1M-2L-3-H1", where "Ⅲ-SP" is the layer identifier, indicating that this block is located on the third layer and requires special support; "R-1M-2L" is the azimuth feature code, reflecting the position of this block in the horizontal partition and the associated information of the mechanical movement path.

[0172] The purpose of extracting the layer identifier and azimuth feature code is to be able to accurately find the specific position of this construction block in the three-dimensional geological model in the future, and to understand its construction conditions and related parameters. The layer identifier can help this embodiment determine the layer where this block is located, and then understand the geological conditions and construction requirements of this layer; the azimuth feature code can enable this embodiment to clarify the specific position of this block in the horizontal partition and the movement of the machine near this block.

[0173] Step S1553: According to the layer identifier and azimuth feature code, locate the corresponding surrounding rock grade parameter and support time sequence parameter in the three-dimensional geological model.

[0174] After obtaining the layer identifier and azimuth feature code of the affected construction block, use the above coding information to accurately search and locate in the three-dimensional geological model. The three-dimensional geological model details the geological information and construction parameters of each area of the cavern.

[0175] Taking the code "Ⅲ-SP-R-1M-2L-3-H1" as an example, through the hierarchical identifier "Ⅲ-SP", the corresponding area of the third layer is found in the three-dimensional geological model. Since there is a special support mark at this level, it indicates that the geological conditions in this area are relatively complex and the surrounding rock stability is poor. Further view the geological data of this area and determine that the surrounding rock grade parameter is Class Ⅳ.

[0176] According to the azimuth feature code "R-1M-2L", find the specific location of this block in the horizontal partition of the third layer. Combine the geological information and construction plan of this location to find the corresponding support timing parameters. For example, after the excavation of this block is completed, the initial support should be carried out within 12 hours and the secondary support should be completed within 24 hours according to the original plan.

[0177] Accurately obtaining the surrounding rock grade parameters and support timing parameters is very important for subsequent processing. The surrounding rock grade parameters determine the geological strength and stability of this block, providing a basis for adjusting the construction plan and support measures; the support timing parameters are related to the construction progress and quality, ensuring that the support operation is carried out at the appropriate time to ensure the stability of the surrounding rock.

[0178] Step S1554: Based on the updated surrounding rock grade parameters and support timing parameters, recalculate the weight of the stability objective function in the multi-objective optimization function set.

[0179] The multi-objective optimization function set includes multiple objective functions such as the construction period objective function, economic cost objective function, resource balance objective function, and surrounding rock deformation objective function. The stability objective function mainly reflects the stability of the surrounding rock during the construction process, and the size of its weight directly affects the degree of emphasis on the surrounding rock stability in the entire optimization process.

[0180] When the surrounding rock grade parameters and support timing parameters of the affected construction block are updated, it is necessary to recalculate the weight of the stability objective function. Taking the construction block with Class Ⅳ surrounding rock as an example, since its surrounding rock grade is low, the stability is poor, and the deformation rate has exceeded the threshold, it indicates that the surrounding rock stability problem in this block is relatively prominent.

[0181] When recalculating the weight, considering the importance of the surrounding rock stability to the entire construction process, it is necessary to increase the weight of the stability objective function. The original weight of the stability objective function may be 0.2. According to the actual situation of this block, adjust its weight to 0.4. At the same time, correspondingly reduce the weights of other objective functions. For example, adjust the weight of the construction period objective function from 0.3 to 0.2, the weight of the economic cost objective function from 0.3 to 0.2, and the weight of the resource balance objective function from 0.2 to 0.2.

[0182] Through such weight adjustment, in the subsequent multi-objective optimization process, more attention can be paid to the stability of the surrounding rock, and the construction sequence plan that can ensure the stability of the surrounding rock is preferentially selected, thereby reducing construction risks and ensuring construction safety.

[0183] Step S1555: Use the incremental learning algorithm to dynamically update the non-dominated solution set, generate a new Pareto front plan, and trigger a real-time adjustment instruction for the construction sequence parameters.

[0184] After recalculating the weights of the stability objective functions in the multi-objective optimization function set, it is necessary to use the incremental learning algorithm to dynamically update the existing non-dominated solution set. The incremental learning algorithm can efficiently adjust the model when new data is added, without having to retrain the entire model.

[0185] First, integrate the recalculated weights of the stability objective functions, as well as the updated surrounding rock grade parameters, support timing parameters, etc. of the affected construction blocks into new input data. The above data reflects the changes in the current construction situation, especially the latest information of the construction blocks affected by the deformation of the surrounding rock.

[0186] Then select a suitable incremental learning algorithm. Considering the actual engineering situation, the online gradient descent method is selected. Take the existing non-dominated solution set as the initial model state, and each individual (construction sequence plan) in the solution set corresponds to a set of model parameters.

[0187] In the stage of iteratively updating the model parameters, use the online gradient descent method to iteratively update the model parameters according to the new input data. In each iteration, calculate the gradient of the objective function, and adjust the model parameters according to the direction and magnitude of the gradient to gradually reduce the value of the objective function. Specifically, for each individual, calculate its gradient in the multi-objective space according to the new weight of the stability objective function and other objective function values. Then, according to the principle of gradient descent, update the construction sequence parameters of this individual to make it better in multi-objective optimization.

[0188] After each iterative update, perform non-dominated sorting on the updated solution set. Non-dominated sorting divides the individuals in the solution set into different levels. The higher the level, the better the performance of the individual in multiple objective functions. Screen out the non-dominated individuals to form a new Pareto front plan. The above Pareto front plan reaches a new balance among multiple objectives such as construction period, cost, resource balance, and surrounding rock deformation, and takes into account the importance of surrounding rock stability.

[0189] After multiple iterative updates and non-dominated sorting, a new Pareto front plan is finally generated. The above Pareto front plan reflects the optimal construction sequence selection after considering the changes in the surrounding rock deformation situation. Each plan contains the construction sequence of each block in the unique construction block coding set and performs well in multi-objective optimization.

[0190] Finally, the newly generated Pareto front solutions are sent to the construction management system. The construction management system selects the optimal construction sequence solution according to the priority sorting result of the above Pareto front solutions. Then, it triggers a real-time adjustment instruction for the construction sequence parameters and sends the adjusted construction sequence parameters to the on-site construction equipment and personnel. For example, the construction management system sends the new construction sequence information to the control system of the three-boom rock drill through a wireless network, enabling the drill to operate according to the new sequence. At the same time, the construction personnel will also receive the new construction plan and task arrangement to ensure that the construction process can be carried out efficiently and safely according to the adjusted plan.

[0191] Figure 2 FIG. shows a schematic diagram of exemplary hardware and software components of an underground cavern intelligent construction sequence optimization system 100 that can implement the ideas of the present application provided by some embodiments of the present application. For example, the processor 120 can be used on the underground cavern intelligent construction sequence optimization system 100 and is used to execute the functions in the present application.

[0192] The underground cavern intelligent construction sequence optimization system 100 can be a general-purpose server or a special-purpose server, both of which can be used to implement the underground cavern intelligent construction sequence optimization method of the present application. Although only one server is shown in the present application, for convenience, the functions described in the present application can be implemented in a distributed manner on multiple similar platforms to balance the processing load.

[0193] For example, the underground cavern intelligent construction sequence optimization system 100 can include a network port 110 connected to the network, one or more processors 120 for executing program instructions, a communication bus 130, and different forms of storage media 140, such as disks, ROMs, or RAMs, or any combination thereof. Exemplarily, the underground cavern intelligent construction sequence optimization system 100 can also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The method of the present application can be implemented according to the above program instructions. The underground cavern intelligent construction sequence optimization system 100 also includes an input / output (I / O) interface 150 between the computer and other input / output devices.

[0194] For ease of description, only one processor is described in the intelligent construction sequence optimization system 100 for underground chambers. However, it should be noted that the intelligent construction sequence optimization system 100 in this application may also include multiple processors. Therefore, the steps performed by one processor described in this application may also be jointly performed or separately performed by multiple processors. For example, if the processor of the intelligent construction sequence optimization system 100 for underground chambers performs step A and step B, it should be understood that step A and step B may also be jointly performed by two different processors or separately performed in one processor. For example, the first processor performs step A, the second processor performs step B, or the first processor and the second processor jointly perform steps A and B.

[0195] In addition, an embodiment of the present invention further provides a readable storage medium, in which computer-executable instructions are preset. When the processor executes the computer-executable instructions, the above intelligent construction sequence optimization method for underground chambers is implemented.

[0196] It should be noted that, in order to simplify the description of the present invention disclosure and thus help the understanding of one or more embodiments of the invention, in the previous description of the embodiments of the present invention, sometimes multiple features are merged into one embodiment, drawing or description thereof.

Claims

1. An intelligent construction sequence optimization method for underground chambers, characterized in that, The method includes: Integrating a set of geological parameters, a set of construction machinery parameters, and a set of historical construction cases to construct a construction method database; Based on the spatial topological relationship between the geomechanical index in the construction method database and the 3D geological model, determining a set of layer height parameters through a dynamic matching algorithm; According to the set of construction machinery parameters and the set of layer height parameters, using an octree spatial dissection algorithm to generate a set of lateral partition parameters and a set of axial block parameters; Based on the spatial coordinate attributes, surrounding rock grade attributes, the set of lateral partition parameters, and the set of axial block parameters of the 3D geological model, performing composite coding processing to generate a unique construction block coding set including a layer identifier, an axial identifier, and a block identifier; Performing construction sequence optimization processing on the unique construction block coding set through a multi-objective optimization algorithm to generate a Pareto optimal solution set including a construction period optimization path, a cost optimization path, and a stability optimization path, and dynamically adjusting the construction sequence parameters in the Pareto optimal solution set according to the surrounding rock deformation data monitored in real time.

2. The intelligent construction sequence optimization method for underground caverns according to claim 1, wherein The integrating a set of geological parameters, a set of construction machinery parameters, and a set of historical construction cases to construct a construction method database includes: Obtaining the surrounding rock integrity coefficient and in-situ stress distribution data in the set of geological parameters, and establishing a first mapping relationship set between the surrounding rock grade and the layer height, where the first mapping relationship set is obtained by performing non-linear relationship fitting on the layer height records in the set of historical construction cases through a machine learning algorithm; Analyzing the operation radius parameter of the three-boom rock drilling jumbo in the set of construction machinery parameters, and establishing a second mapping relationship set between the mechanical operation range and the lateral partition width, where the second mapping relationship set includes a maximum lateral partition width constraint condition and a minimum safe operation spacing constraint condition; Extracting the fault fracture zone treatment cases and support time sequence cases in the set of historical construction cases, and constructing a dynamic rule base through an incremental learning algorithm, where the dynamic rule base includes a layer height adjustment rule and a block size reduction rule under geological mutation conditions; Performing relational data fusion on the first mapping relationship set, the second mapping relationship set, and the dynamic rule base to generate a construction method database including a surrounding rock grade field, a mechanical parameter field, and a construction constraint field; Updating the surrounding rock grade field in the construction method database through the on-site geological exploration data collected in real time, and triggering the layer height adjustment rule in the dynamic rule base to recalculate the top arch layer height parameter, the rock anchor beam layer height parameter, and the traffic tunnel associated layer height parameter in the set of layer height parameters.

3. The intelligent construction sequence optimization method for underground chambers according to claim 1, wherein The determining a set of layer height parameters through a dynamic matching algorithm based on the spatial topological relationship between the geomechanical index in the construction method database and the 3D geological model includes: Performing spatial coordinate analysis on the 3D geological model to identify the upper inflection point coordinates and the lower inflection point coordinates of the rock anchor beam, where the upper inflection point coordinates are located in the plane rectangular coordinate system through the surrounding rock integrity coefficient in the construction method database and the maximum span comparison algorithm; Calculate the initial height parameter of the crown arch layer according to the surrounding rock grade field and geomechanical index in the construction method database, wherein the initial height parameter of the crown arch layer is obtained by combining a dynamic calculation formula with the bolt installation height constraint condition; Measure the vertical distance between the upper inflection point coordinate of the rock anchor beam and the bottom of the crown arch layer. When the vertical distance exceeds the preset threshold in the construction method database, trigger the layer division decision rule to generate the rock anchor beam layer height parameter including the double-layer division mark; Combine the traffic tunnel elevation parameter and the installation bay position parameter in the construction method database to determine the height parameter of the lower layer of the rock anchor beam through spatial topological analysis, wherein the height parameter of the lower layer of the rock anchor beam satisfies the blasting vibration safety distance constraint and the mechanical operation radius constraint in the construction machinery parameter set; Dynamically correct the initial height parameter of the crown arch layer, the rock anchor beam layer height parameter and the rock anchor beam lower layer height parameter according to the historical layer height record in the construction method database to generate a set of layer height parameters meeting the surrounding rock stability requirements.

4. The intelligent construction sequence optimization method for underground chambers according to claim 3, characterized in that The obtaining the initial height parameter of the crown arch layer by combining a dynamic calculation formula with the bolt installation height constraint condition includes: Obtain the tensile strength parameter of the bolt material and the cohesion parameter of the surrounding rock grade field in the construction method database; Generate the bolt spacing constraint range according to the ratio relationship between the tensile strength parameter and the cohesion parameter; Combine the curvature radius parameter of the crown arch area in the three-dimensional geological model to calculate the coupling coefficient between the bolt installation height and the crown arch layer thickness; Based on the coupling coefficient and the bolt spacing constraint range, dynamically match the optimal solution of the support density in the historical construction case set; Perform an intersection operation on the matching result and the blasting vibration safety distance constraint in the construction method database to generate the feasible region interval of the initial height parameter of the crown arch layer.

5. The intelligent construction sequence optimization method for underground chambers according to claim 1, characterized in that, The generating the transverse partition parameter set and the axial block parameter set by using the octree space subdivision algorithm according to the construction machinery parameter set and the layer height parameter set includes: Based on the crown arch layer height parameter and the rock anchor beam layer height parameter in the layer height parameter set, determine the horizontal projection contour data of each layer of the three-dimensional geological model; Calculate the maximum allowable transverse partition width corresponding to each layer according to the trolley width parameter and the operation radius parameter in the construction machinery parameter set; Perform spatial grid division on the horizontal projection contour data through the octree space subdivision algorithm to generate the initial transverse partition width parameter and the initial axial block length parameter, wherein the initial transverse partition width parameter is less than or equal to the maximum allowable transverse partition width; Dynamically reduce the initial transverse partition width parameter according to the maximum span standard corresponding to the surrounding rock grade field in the construction method database to generate the transverse partition width parameter meeting the surrounding rock stability constraint; Iteratively optimize the initial axial block length parameter based on the daily maximum footage parameter and the mechanical utilization rate parameter in the construction method database to generate the axial block length parameter meeting the mechanical operation cycle constraint. When there is a fault fracture zone area in the construction method database, the lateral partition width parameter and the axial block length parameter are synchronously reduced in equal proportion to generate the adjusted lateral partition width parameter and axial block length parameter in the fault zone; The lateral partition width parameter and the adjusted lateral partition width parameter in the fault zone are combined into a lateral partition parameter set, and the axial block length parameter and the adjusted axial block length parameter in the fault zone are combined into an axial block parameter set, and a support marking parameter is inserted into the combined parameter set.

6. The intelligent construction sequence optimization method for underground chambers according to claim 5, characterized in that, The spatial grid division of the horizontal projection contour data by the octree spatial division algorithm to generate the initial lateral partition width parameter and the initial axial block length parameter includes: Extract the set of irregular boundary vertex coordinates in the horizontal projection contour data; Set the recursive depth threshold of the octree division algorithm according to the spatial distribution density of the vertex coordinate set; During each recursive division process, detect the mutation mark of the surrounding rock grade field contained in the grid cell; When the mutation mark is detected, terminate the recursive division of the current grid cell and lock the grid boundary; Continue to perform spatial division on the grid cells that have not triggered the mutation mark until the recursive depth threshold is reached; Compare the grid cell width parameter in the final division result with the maximum allowable lateral partition width level by level, and generate the initial lateral partition width parameter after removing the over-limit cells; And, according to the axial extension length of the horizontal projection contour data, extract the reference direction vector of the tunnel axis in the three-dimensional geological model; Based on the robotic arm telescopic stroke parameter in the construction machinery parameter set, determine the minimum cutting step length of the axial block; During the octree spatial division process, axially extend and cut the horizontal projection contour along the reference direction vector to generate the initial grid cell containing the axial cutting plane; According to the extension length parameter of the initial grid cell in the reference direction, extract the candidate value of the axial block length that meets the minimum cutting step length constraint; Combined with the axial block records of historical cases in the construction method database, screen out the axial block length parameter that matches the current surrounding rock grade field as the initial axial block length parameter.

7. The intelligent construction sequence optimization method for underground chambers according to claim 1, characterized in that The composite coding process based on the spatial coordinate attributes, surrounding rock grade attributes, the lateral partition parameter set and the axial block parameter set of the three-dimensional geological model to generate a unique construction block coding set containing a hierarchical identifier, an axial identifier and a block identifier includes: Analyze the spatial coordinate attribute data of each layer in the three-dimensional geological model, and combine the surrounding rock category parameter in the surrounding rock grade attribute to generate a hierarchical number containing a Roman numeral sequence and a special support mark, where the special support mark is dynamically generated according to the stability threshold in the surrounding rock grade attribute; Based on the left zone width parameter, middle zone width parameter and right zone width parameter in the lateral partition parameter set, and combined with the azimuth coordinate data in the spatial coordinate attributes of the three-dimensional geological model, generate an azimuth feature code, and the azimuth feature code contains the spatial topological relationship and mechanical movement path association parameters of the left zone, middle zone and right zone; Generate an axial sequence code containing excavation sequence priority marks and resource allocation association marks according to the daily footage block length parameter and volume constraint parameter in the axial block parameter set, in combination with the axial spatial coordinate attributes of the three-dimensional geological model; Perform composite splicing processing on the hierarchical number, azimuth feature code, and axial sequence code to generate a set of unique construction block codes that integrate surrounding rock grade attributes, spatial coordinate attributes, and construction parameters, and establish a two-way mapping relationship between the set of unique construction block codes and the three-dimensional model components, so that the hierarchical identifier, axial identifier, and block identifier of each block correspond one-to-one with the model spatial coordinates; When it is detected that the surrounding rock grade attribute in the construction method database is updated or the lateral partition parameter set and axial block parameter set are adjusted, dynamically correct the support mark parameter in the azimuth feature code and the resource allocation association mark in the axial sequence code, and synchronously update the associated fields in the set of unique construction block codes.

8. The intelligent construction sequence optimization method for underground chambers according to claim 1, wherein Perform construction sequence optimization processing on the set of unique construction block codes through a multi-objective optimization algorithm to generate a Pareto optimal solution set containing construction period optimization paths, cost optimization paths, and stability optimization paths, and dynamically adjust the construction sequence parameters in the Pareto optimal solution set according to the surrounding rock deformation data monitored in real time, including: Extract the surrounding rock grade parameters, resource allocation parameters, and spatial position parameters of each block according to the hierarchical identifier and azimuth feature code in the set of unique construction block codes, where the surrounding rock grade parameters are derived from the surrounding rock grade field in the construction method database, and the resource allocation parameters include mechanical operation radius parameters and support time sequence parameters; Based on the extracted surrounding rock grade parameters, resource allocation parameters, and spatial position parameters, establish a multi-objective optimization function set including a construction period objective function, an economic cost objective function, a resource balance objective function, and a surrounding rock deformation objective function; Use the NSGA-II algorithm to perform parallel optimization processing on the multi-objective optimization function set to generate a non-dominated solution set containing multiple Pareto front solutions; Optimize the mechanical movement path of the Pareto front solutions through the traveling salesman problem model, calculate the comprehensive score of each Pareto front solution, and generate a priority ranking result; Collect the surrounding rock deformation rate data monitored by the fiber optic sensor in real time. When it is detected that the deformation amount exceeds the threshold in the construction method database, trigger a dynamic adjustment mechanism to recalculate the construction sequence parameters in the Pareto front solutions.

9. The intelligent construction sequence optimization method for underground caverns according to claim 8, characterized in that Dynamically adjust the construction sequence parameters in the Pareto optimal solution set according to the surrounding rock deformation data monitored in real time, including: Collect the surrounding rock deformation rate data and displacement data of each construction block in real time through the fiber optic sensor, and compare them with the deformation threshold parameters in the construction method database; When the surrounding rock deformation rate data exceeds the deformation threshold parameters, extract the hierarchical identifier and azimuth feature code of the affected construction block; Locate the corresponding surrounding rock grade parameters and support time sequence parameters in the three-dimensional geological model according to the hierarchical identifier and azimuth feature code. Based on the updated surrounding rock grade parameters and support timing parameters, recalculate the weight of the stability objective function in the set of multi-objective optimization functions; Use the incremental learning algorithm to dynamically update the non-dominated solution set, generate a new Pareto front solution, and trigger a real-time adjustment instruction for the construction sequence parameters.

10. An intelligent construction sequence optimization system for underground chambers, characterized in that, The intelligent construction sequence optimization system for underground caverns includes a processor and a memory, the memory is connected to the processor, the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to implement the intelligent construction sequence optimization method for underground caverns according to any one of claims 1-9 above.

Images