Blasting parameter selection model establishment method and system and parameter selection method

Abstract

The application discloses a tunnel smooth blasting parameter selection method, aiming at the problem of tunnel overbreak and underbreak, constructs a blasting parameter optimization model by taking the minimization of overbreak and underbreak as an objective function, and taking the tunnel section size, lithology grade and geological condition as constraint conditions, and uses an improved genetic algorithm with dynamic constraints, hierarchical coding and multi-stage fitness evaluation to screen out the optimized blasting parameters under the minimum overbreak and underbreak through continuous selection, crossover and mutation operations. Compared with the traditional genetic algorithm, the application improves the global convergence and parameter practicability, and avoids the problems of premature convergence and invalid solution of the traditional genetic algorithm.

Description

Technical Field

[0001] This invention relates to the fields of tunnel blasting and computer intelligent algorithm technology, and in particular to a method for establishing a blasting parameter optimization model, a method for selecting blasting parameters, and their applications. Background Technology

[0002] According to the "2024 China Tunnel Engineering Yearbook," the drill-and-blast method still accounts for 68.3% of mountain tunnel engineering. Due to its strong adaptability to Class V-VI surrounding rock and its 30%-50% lower operating cost compared to TBMs, the drill-and-blast method remains the mainstream construction method. However, its characteristic of "instantaneous impact load" (peak stress wave of 100-300 MPa, duration <10 ms) makes excavation profile control a common challenge in the industry. Factors causing over- and under-excavation in tunnels mainly include geological control, process control, and management factors. Over-excavation increases the amount of excavated material and backfill concrete, creating difficulties for subsequent operations, leading to excessive consumption of manpower, machinery, and materials, and affecting project quality and progress. Under-excavation requires additional manpower or machinery for clearing debris; if supplementary blasting is added, over-excavation is more likely, affecting construction progress, increasing tunnel engineering costs, and, in severe cases, impacting the acceptance and stability of the surrounding rock.

[0003] Therefore, both over-excavation and under-excavation will increase construction costs, delay the construction period, and may even affect tunnel operation safety. Thus, accurately determining the smooth blasting parameters for tunnels to improve tunnel structural safety and subsequent construction efficiency is an urgent problem to be solved. Summary of the Invention

[0004] To address the shortcomings of the existing technologies, this invention proposes a method for selecting tunnel smooth blasting parameters. The method takes minimizing over- and under-excavation as the objective function and tunnel cross-sectional dimensions, lithology grade, and geological conditions as constraints. It adds a dynamic constraint-guided crossover and mutation strategy, a hierarchical coding mechanism, and a multi-stage fitness evaluation method to the traditional genetic algorithm. Through continuous selection, crossover, and mutation operations, the blasting parameters are optimized, and the optimal tunnel smooth blasting parameters are finally obtained.

[0005] The technical solution adopted in this invention includes the following steps:

[0006] The method includes determining the optimization variables, objective function, and constraints, wherein:

[0007] The optimization variables are the spacing between the peripheral holes and the thickness of the light burst layer;

[0008] The constraints are that the error of the actual diameter of the tunnel after blasting is within the allowable error range, the spacing between the surrounding holes is less than the limit value, and the thickness of the blasting layer is less than the maximum allowable thickness.

[0009] The objective function is to minimize the over-excavation and under-excavation after smooth blasting of the tunnel.

[0010] Therefore, the present invention employs the above-described method for selecting tunnel smooth blasting parameters, which has the following beneficial effects:

[0011] First, this invention, based on the traditional genetic algorithm framework, introduces dynamic constraint embedding, dynamic repair, and multi-stage evaluation, optimizing the burst parameters through continuous selection, crossover, and mutation operations. Compared with the traditional genetic algorithm, the optimized algorithm improves global convergence and parameter usability, while avoiding the premature convergence and invalid solution problems of the traditional genetic algorithm.

[0012] Second, the present invention adopts a layered hybrid coding mechanism, which distinguishes primary and secondary variables through coding, determines the peripheral hole spacing and the thickness of the light burst layer as core control parameters, and determines the charge density and detonation sequence as auxiliary parameters, prioritizes the optimization of core parameters that have a significant impact on the objective function, and directly embeds them into the feasible interval through a dynamic constraint embedding mechanism, thereby reducing invalid searches and saving time;

[0013] Third, this invention uses a multi-stage evaluation mechanism of "coarse screening - fine calculation" to perform targeted repair on individuals that violate constraints during genetic operations such as crossover and mutation, which can ensure that the population always evolves within the feasible domain; and when calculating fitness, the constraint violation degree is weighted to the objective function through a penalty function, which forces the algorithm to prioritize meeting the hard requirements of engineering and generate only parameter combinations that conform to the current lithology.

[0014] Fourth, this invention generates local offspring to replace low-fit individuals in the population by perturbing the neighborhood of the selected optimal individuals. By perturbing randomly, a certain degree of randomness is preserved, and the population is prevented from falling into a single solution due to forced correction, thereby improving the convergence accuracy and solving the problems of premature convergence and insufficient practicality of solution sets in traditional genetic algorithms.

[0015] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method proposed in this invention.

[0017] Figure 2 This is a flowchart of a traditional genetic algorithm.

[0018] Figure 3 This is a flowchart of the improved genetic algorithm of the present invention.

[0019] Figure 4 The results are ANSYS simulations shown in the examples.

[0020] Figure 5The present invention is compared with the SGA algorithm and the PSO algorithm in the embodiments.

[0021] Figure 6 The optimal blasting parameters and fitness relationship in the example Figure 3 Dimensional distribution diagram. Detailed Implementation

[0022] In the description of this invention, it should also be noted that, unless otherwise expressly specified and limited, these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0023] like Figure 1 As shown, the method for selecting tunnel smooth blasting parameters of the present invention includes the following steps:

[0024] 1. Constructing an optimization model for blasting parameters

[0025] (1) Determine the objective function

[0026] The most direct evaluation indicator for determining the rationality of selected tunnel blasting parameters is the smooth blasting effect. The evaluation indicators for smooth blasting effect include various factors such as the size of the rock fragments after blasting and the amount of over- and under-excavation. Among these, the amount of over- and under-excavation directly reflects the degree of deviation between the blasted profile and the designed cross-section, and directly affects the tunnel structural safety, support costs, and subsequent construction efficiency. It is a core indicator for measuring excavation quality and can be used to quantify the blasting effect and guide parameter optimization. Therefore, this invention selects the amount of over- and under-excavation as the evaluation indicator. Based on this evaluation indicator, the optimization objective of the objective function is determined to minimize the amount of over- and under-excavation after smooth blasting of the tunnel. The amount of over- and under-excavation is the overall deviation between the actual tunnel profile and the designed profile after blasting. Thus, the objective function is defined as:

[0027]

[0028] in, This represents the sum of over-excavation and under-excavation. Let i be the elevation of the i-th measured contour point after the blast. For the design elevation, n represents the total number of measuring points on the tunnel cross section;

[0029] (2) Determine the constraints

[0030] Before determining the constraints, a detailed analysis of each influencing factor in the tunnel blasting process is conducted.

[0031] First, the dimensional error after tunnel blasting cannot be too large or too small, otherwise it will affect structural safety or increase repair costs. For example, a large dimensional error may prevent the subsequent installation of support structures or require additional concrete filling, which is both costly and time-consuming. Second, the hardness of different rocks directly affects the selection of borehole spacing. For example, hard rocks can withstand larger borehole spacing, while soft rocks require denser drilling; otherwise, the surface will be uneven after blasting. Finally, when blasting in areas with poor geological conditions, the thickness of each blast must be reduced because the rocks in these areas are inherently unstable, to avoid excessive blast thickness that could trigger a collapse. Therefore, this invention selects the peripheral borehole spacing and the thickness of the smooth blast layer as constraints for the objective function.

[0032] Before designing the blasting parameter scheme, the tunnel cross-section dimensions are determined first, and then the spacing between surrounding holes and the thickness of the blasting layer are constrained based on the cross-section dimensions.

[0033] The allowable error range for the actual diameter of the tunnel after blasting is:

[0034]

[0035] In the formula, This is the actual diameter of the tunnel after blasting. The design diameter of the tunnel, As the allowable error ratio, the embodiments of the present invention are set as follows: .

[0036] ① Spacing constraints of peripheral holes

[0037] Since the spacing between peripheral boreholes is affected by the lithology grade, the hardness of the surrounding rock is first quantified using the Protodyakonov coefficient. The maximum allowable spacing between peripheral boreholes increases with the increase of the Protodyakonov coefficient. Therefore, the range of peripheral borehole spacing is dynamically limited based on the Protodyakonov coefficient value:

[0038]

[0039] In the formula, The Protodyakonov coefficient is obtained through geological exploration, and It is soft rock. It is a medium-hard rock. It is hard rock; The spacing between the peripheral holes; This is the minimum value of the spacing between the peripheral holes; The maximum value of the spacing between the peripheral holes can be determined by the Protodyakonov coefficient; Determined by the minimum drilling requirements and with reference to engineering experience, it can be... Set to 0.3 ;

[0040] The method of dynamically adjusting the spacing of peripheral holes according to the Protodyakonov coefficient can avoid collapse caused by excessive spacing of peripheral holes in soft rock, or waste of resources caused by excessive spacing of peripheral holes in hard rock.

[0041] in:

[0042]

[0043]

[0044] In the formula, For correction factor, ;

[0045] Based on engineering specifications and historical data, a mapping relationship table of lithology, Protodyakonov coefficient and surrounding borehole spacing was constructed, namely the lithology-parameter mapping rule shown in Table 1.

[0046]

[0047] ② Constraints on the thickness of the flash-blast layer under different geological conditions

[0048] Under normal geological conditions, based on the actual diameter of the tunnel before blasting Based on the lithological grade, the maximum allowable thickness is determined. Among them, conventional geological conditions specifically refer to stable geological areas with intact rock masses and no obvious adverse geological structures (such as faults, fracture zones, karst caves, etc.).

[0049] The determined maximum allowable thickness is:

[0050]

[0051] in, This represents the maximum permissible thickness of the flash-blast layer under conventional geological conditions.

[0052] In fault or fracture zone areas, the thickness of the flash-blast layer can be determined during the geological exploration stage before construction using methods such as ground-penetrating radar, borehole scanning, and CT scanning.

[0053]

[0054] In the formula, The safety factor is set based on the geological risk assessment report. The thickness of the photoblast layer under the fault or fracture zone.

[0055] For soft rock or high-stress areas, the maximum allowable thickness is adjusted according to the lithology grade using the following formula:

[0056]

[0057] in, This refers to the maximum permissible thickness of the light-burst layer in soft rock or high-stress areas.

[0058] (3) Establish a blasting parameter optimization model

[0059] Based on the above objective function and constraints, the following optimization model for blasting parameters is established:

[0060]

[0061] in, This is the actual diameter of the tunnel before blasting; It is the sum of over-excavation and under-excavation.

[0062] This invention transforms a practical engineering problem into a mathematical optimization model as shown above by defining the objective function and constraints. This model can find the blasting parameters that minimize over- and under-excavation under three types of constraints: cross-sectional dimensions, spacing between peripheral holes, and geological conditions. These blasting parameters include the spacing between peripheral holes S, the thickness of the smooth blasting layer W, and the charge density. The detonation sequence T. Among these, according to the "Technical Specification for Smooth Blasting in Tunnels," the peripheral hole spacing S and the thickness of the smooth blasting layer W are determined to be core parameters that significantly affect the objective function, along with the charge density. The detonation timing T is an auxiliary parameter. Since auxiliary parameters are factors that have a relatively small impact on the blasting effect, they are mostly discrete options or empirical values. That is, they are usually some fixed values, and the size of these fixed values ​​is generally an empirical value.

[0063] 2. Solve the blasting parameter optimization model using a dynamic constraint-guided genetic algorithm.

[0064] Genetic algorithms are heuristic optimization algorithms used to solve optimization problems. They search for the global optimal solution in the solution space by simulating mechanisms such as natural selection, crossover, and mutation. Figure 2 The flowchart of a traditional genetic algorithm shows that it mainly includes chromosome encoding (generating the initial population), fitness function (evaluating individual quality), genetic operators (selection, crossover, mutation), and runtime parameters (termination condition determination). However, traditional genetic algorithms may converge prematurely and have high computational overhead when the population diversity is insufficient. To address these issues, this invention optimizes the traditional genetic algorithm framework, such as... Figure 3 As shown, this invention solves the blasting parameter optimization model by introducing a dynamic constraint-oriented mechanism, multi-stage fitness evaluation, and local refined search (the modules added in this invention are shown in the red box). Under the premise of satisfying dynamic constraints, it can select the blasting parameter combination with the minimum over- and under-excavation amount.

[0065] (1) Genetic algorithm based on dynamic constraint guidance

[0066] The improved genetic algorithm of this invention (based on a dynamic constraint-guided genetic algorithm) has the following improvements compared to the traditional genetic algorithm:

[0067] ① Dynamic constraint guidance mechanism

[0068] During tunnel blasting, the tunnel is typically divided into multiple construction sections to better control the blasting effect and safety. The goal of this invention is to obtain the optimal blasting parameters for each construction section. Within the same construction section, the lithology and geological structure are relatively uniform, with negligible differences. Therefore, the blasting schemes designed for the same construction section are identical. However, when local anomalies exist within a section, such as sudden crossings of faults or lithological changes during tunnel construction, causing the initial feasible region to fail to cover all situations, this invention introduces a dynamic constraint guidance mechanism. During the iteration of the genetic algorithm, changes in geological conditions are monitored in real time. If a change in geological conditions is detected, the dynamic constraint mechanism dynamically increases or decreases the range of values ​​for the peripheral borehole spacing S and the maximum allowable thickness W of the smooth blasting layer based on the new geological data.

[0069] Specifically, this invention flexibly sets the allowable range of blasting parameters (core parameters) based on the physical characteristics of the surrounding rock (current lithology and geological conditions), rather than using fixed values. In the gene generation stage of the improved genetic algorithm, the feasible range of blasting parameters (core parameters) is dynamically set according to the current lithology and geological conditions; that is, the feasible range of the spacing between peripheral holes and the thickness of the smooth blasting layer is set. This process is called constraint embedding. For example, if... If the rock type is large enough, the rock is classified as hard rock, allowing for an increase in the hole spacing; conversely, if the rock type is large enough, the rock is classified as soft rock, allowing for a decrease in the hole spacing. The improved genetic algorithm of this invention automatically limits the search range of subsequent parameters, generating only parameter combinations that match the current lithology. By directly integrating engineering constraints such as lithology and geology into the gene generation and mutation process, it avoids the generation of a large number of invalid solutions due to random search in traditional genetic algorithms.

[0070] ② Multi-stage fitness evaluation

[0071] First, a simplified elastic wave model is used to quickly screen and eliminate individuals with excessive over- or under-excavation. The elastic wave model is a simplified mechanical model used to quickly estimate the disturbance effect of blasting energy transfer on the surrounding rock. Its core formula is:

[0072]

[0073] in, The estimated over- or under-excavation amount is given by A, where A is the charge energy per hole, k is the rock mass attenuation coefficient, and d is the distance from the measuring point to the blast source.

[0074] The single-hole charge energy, rock mass attenuation coefficient, and distance of the measuring point from the blast source are all set when the construction plan is determined.

[0075] Secondly, in the coarse screening stage, the estimated over- and under-excavation amounts can be quickly calculated based on the simplified elastic wave model described above. ,Will Over- and under-dig thresholds If a comparison is made, If the threshold is exceeded, the individual is removed, thus eliminating multiple individuals exceeding the over- or under-mining threshold, preventing them from undergoing the more time-consuming finite element simulation. The over- or under-mining threshold... Determined according to the provisions of the "Technical Specifications for Highway Tunnel Construction";

[0076] Secondly, in the precise calculation stage, finite element blasting simulation is used to obtain the accurate over- and under-excavation amounts for the remaining individuals after the preliminary screening stage. When performing finite element blasting simulation, LS-DYNA software or ANSYS software can be used to perform high-precision numerical simulation of the blasting process.

[0077] Finally, based on the obtained accurate over- and under-excavation amounts Establish a fitness function calculation formula, which can be used to evaluate the quality of individuals;

[0078] The fitness function combines the precise over- or under-excavation amount with the constraint violation degree and introduces a dynamic penalty factor, resulting in the following specific formula for calculating the fitness function:

[0079]

[0080] Where F is the fitness value, the larger the value, the better the solution; It is a dynamic penalty factor; To constrain the degree of violation;

[0081] In the formula, the dynamic penalty factor Able to adjust constraint violation degree The weights used in fitness calculation are expressed as follows:

[0082]

[0083] in, The initial penalty value is set to 0.1 in this embodiment of the invention; The growth coefficient is set to 0.01 in this embodiment of the invention; t is the current iteration number;

[0084] The formula for calculating the constraint violation degree C is:

[0085]

[0086] in, The parameter exceeds the limit. To constrain the upper limit;

[0087] The parameter exceeding the limit refers to the difference between the actual value of the parameter and the upper limit of the constraint, which is the maximum peripheral hole spacing and the maximum allowable thickness of the light burst layer.

[0088] In summary, this invention improves computational efficiency and optimization accuracy through a multi-stage evaluation mechanism of "coarse screening - fine calculation".

[0089] ③ Local fine-grained search

[0090] After selecting high-quality individuals through a multi-stage evaluation mechanism, the best individuals are then subjected to neighborhood perturbation to generate local offspring that replace low-fit individuals in the population. After neighborhood perturbation, crossover, mutation, evaluation, and perturbation continue until the termination condition is met.

[0091] When dealing with neighborhood perturbations, based on the dynamic constraints of the previous step—namely, the upper limits of the maximum peripheral hole spacing and the maximum allowable thickness of the light burst layer—are further adjusted. Adjustments were made within the specified range to obtain new maximum peripheral hole spacing limits and maximum allowable thickness limits for the light burst layer. Within the new maximum peripheral hole spacing limits and maximum allowable thickness limits for the light burst layer, high-quality individuals were re-selected to generate local offspring.

[0092] This invention addresses the problems of premature convergence and insufficient practicality of solution sets in traditional genetic algorithms by reducing invalid searches through constraint embedding, lowering computational costs through multi-stage evaluation, and improving convergence accuracy through local perturbation. The final output is a combination of blasting parameters that both meets the constraints of dynamic geological conditions and minimizes over- and under-excavation.

[0093] (2) Solving the blasting parameter optimization model

[0094] In the initialization phase of a genetic algorithm, its core task is to generate an initial population that meets engineering constraints, thereby transforming actual engineering parameters (such as the spacing between peripheral holes, the thickness of the light burst layer, etc.) into genes that the genetic algorithm can operate on.

[0095] Traditional genetic algorithms initialize the population through chromosome encoding. This invention, however, distinguishes between primary and secondary variables during chromosome encoding, prioritizing the optimization of core parameters (primary variables) and then auxiliary parameters (secondary variables). The optimized core parameters are directly embedded into the feasible solution interval (e.g., the spacing between surrounding holes does not exceed 60% of the tunnel diameter) to reduce invalid searches and save time. Encoding is divided into main chain encoding and auxiliary chain encoding, performed in parallel. The main chain encodes the core parameters using real-number encoding, while the auxiliary chain encodes the auxiliary parameters using binary encoding. The specific steps include:

[0096] ① Main chain processing flow

[0097] S1: Initial core parameter population is generated using Latin hypercube sampling (LHS);

[0098] First, the range of values ​​for each core parameter (peripheral hole spacing, light burst layer thickness) is divided into N sub-intervals (N is the population size). Then, a value is randomly selected from each sub-interval of each core parameter to ensure that each sub-interval is sampled only once. Finally, different sampled values ​​are randomly combined to generate N individuals. Each parameter is evenly distributed within its value range and there is no overlapping sub-interval coverage.

[0099] Next, in the gene generation stage (i.e. the initial population generation stage), the present invention dynamically sets the feasible range of the main chain parameters according to the current lithology and geological conditions, so that the selected genes (individuals) are high-quality genes, thereby avoiding the generation of a large number of invalid solutions due to random search in traditional genetic algorithms;

[0100] Based on historical blasting data and similar engineering cases, N is usually set to 50~200. However, due to limited computing resources, in order to balance accuracy and speed, N is set to 100 in this embodiment of the invention.

[0101] S2: Perform crossover and selection operations on the generated initial parameter population to generate preliminary offspring;

[0102] First, simulated binary crossover is used as the crossover operation. By mixing the parent gene values, the offspring gene values ​​are determined. For example, if the parent AS = 5.2m and the parent BS = 5.8m, the average of the two parent genes is taken, and then the offspring S = 5.5m.

[0103] Then, through selection operations, high-quality parents are selected from the current population, the position of gene crossover is randomly determined, and some genes are exchanged to generate preliminary offspring. The specific selection operation refers to: within the determined feasible interval, the core parameters (S, W) are screened, and the core parameters that are within the feasible interval and close to the middle value of the feasible interval are selected as high-quality parents.

[0104] S3: Perform mutation operations on the initial offspring to obtain the corrected parameters;

[0105] For parameters of the initial offspring that exceed the dynamic constraints, a hard truncation method is used for correction, directly adjusting the parameters to the nearest boundary value. For example, taking the core parameter S as an example, if... offspring Then the child parameter Directly corrected to ;

[0106] This step, by forcibly pulling the out-of-limit parameters back to the current dynamic constraint range, ensures that all parameter combinations meet engineering safety requirements and prevents invalid solutions from entering the next generation;

[0107] S4: Perform a local fine-grained search to select the optimal combination of core parameters;

[0108] First, through the multi-stage fitness evaluation described above, the fitness value is calculated using the fitness function described above. In each generation, the core parameter combination with the smallest precise over-mining and under-mining (i.e. the highest fitness value) is selected as the candidate core parameters. Then, by introducing a small random perturbation into the selected candidate core parameters, local offspring are generated in their neighborhood to replace individuals with low fitness.

[0109] When performing neighborhood perturbation, taking the adjustment of the peripheral hole spacing S as an example, the parameters are adjusted using the following formula:

[0110]

[0111] in, For the disturbance amplitude, These are random numbers from a standard normal distribution. The peripheral hole spacing is corrected using the hard cut-off method. The new peripheral hole spacing after applying neighborhood perturbation;

[0112] For example, in Based on, according to It is possible to get ;

[0113] Similarly, when adjusting the thickness W of the light burst layer, a formula similar to the one above is used, for example... The thickness W of the light burst layer is also perturbed in the neighborhood, replacing individuals with low fitness in the original population.

[0114] If the parameter exceeds the specified dynamic constraint range again after the disturbance, the hard truncation method is used again for secondary correction to correct the parameter to the boundary value until the corrected parameter is within the dynamic constraint range.

[0115] This dynamic repair method ensures that the final parameters strictly meet the dynamic constraints, and retains a certain degree of randomness through random perturbation, thus preventing the population from falling into a single solution due to forced correction.

[0116] Repeat steps S1-S4 until the parameters converge to obtain the optimal combination of core parameters.

[0117] ② Secondary chain processing flow

[0118] Since the auxiliary parameters are mostly discrete options or empirical values, that is, generally fixed values, the chromosome coding stage in the auxiliary chain still uses the traditional genetic algorithm.

[0119] For auxiliary parameters (charge density, detonation timing), the auxiliary chain uses binary encoding to achieve discretization optimization of the parameters.

[0120] S1: Population initialization;

[0121] First, the number of binary bits is determined based on the number of discrete options for the auxiliary parameter. Specifically, for an auxiliary parameter with M discrete values, the number of binary bits n in the generated binary encoded string should satisfy... ;

[0122] Secondly, the population is initialized by randomly generating binary strings, with each segment of the binary string corresponding to different parameters;

[0123] The binary encoding mapping rules corresponding to the detonation timing and charge density in this invention are shown in Table 2:

[0124]

[0125] S2: Perform crossover operation and select parent individuals with higher fitness;

[0126] High-quality parents are selected through crossover, that is, parents with high fitness (also referring to the gene values ​​of offspring). Then, a portion of binary fragments are exchanged through single-point crossover. For example, if the charge density is parent A=01 and parent B=10, the offspring will be 00 after crossover.

[0127] S3: Perform mutation operations on the screened individuals;

[0128] The binary code of a certain bit in the binary string is flipped, turning 0 into 1 and 1 into 0, thereby introducing randomness to maintain population diversity;

[0129] S4: Binary encoding conversion;

[0130] By using the mapping rules between the binary encoding of auxiliary parameters and engineering parameters (Table 2), the binary encoding is converted into actual engineering parameters.

[0131] ③ Screening of blasting parameter combinations

[0132] By merging the actual engineering parameters obtained from the auxiliary chain with the continuous parameters obtained from the main chain, the optimal combination of blasting parameters can be obtained.

[0133] Once the optimal combination of blasting parameters is obtained, the fitness value F is calculated using the fitness function formula mentioned above. The quality of the blasting parameter combination is evaluated based on the size of the fitness value F, ensuring that the algorithm selects the globally optimal combination of blasting parameters while satisfying dynamic constraints.

[0134] ④ Termination conditions

[0135] When the rate of change of optimal fitness over 10 consecutive generations The algorithm terminates when the maximum number of iterations is reached (set to 100 in this scheme). The final output is the parameter combination that minimizes over- and under-mining, i.e., the optimal solution or a near-optimal solution set.

[0136] Example

[0137] Data from the blasting operations of Guangming Tunnel and Mingyue Mountain Tunnel were collected. Over-excavation and under-excavation were determined using cross-sectional scanning measurements of the tunnel profile after blasting. A total of 443 data sets were collected during the blasting phase. Due to incomplete data during the blasting process, missing records and data that clearly did not conform to patterns were removed to ensure data integrity and accuracy, resulting in a final selection of 324 data sets. The selected datasets, as shown in Table 3, were collected for each blasting operation.

[0138]

[0139] Before determining the surrounding hole spacing and the thickness of the blasting layer, the tunnel cross-sectional dimensions, including the diameter of this section of the tunnel, must be determined. The thickness is 12m, the surrounding rock is classified as Class III, corresponding to a Protodyakonov coefficient of 4, and there are local fault fracture zones. Therefore, the thickness of the flash-blasted layer is limited. .

[0140] Based on the tunnel smooth blasting parameter selection method proposed in this invention, the specific steps for obtaining the optimal blasting parameters include:

[0141] ① Initial population generation: Using engineering experience-guided Latin hypercube sampling (LHS) and combining it with the distribution of historical blasting data shown in Table 3, 100 initial individuals were generated. Statistical analysis was performed on these 100 initial individuals, and 80% of them met the constraints. The initial population generated is shown in Table 4:

[0142]

[0143] A dynamic constraint mechanism is used to further screen individuals in the initial population (containing 100 initial individuals) generated above, forcing... and restrictions By using a hard truncation method, W and S are corrected to the nearest boundary value. Examples of the filtered parameters are shown in Table 5.

[0144]

[0145] The main chain uses simulated binary crossover and Gaussian mutation, while the auxiliary chain uses single-point crossover and bit flipping. Excessive parameters are handled according to... The corrected parameters after crossover and mutation are shown in Table 6.

[0146]

[0147] ② Multi-stage fitness evaluation: Elastic wave model is used to calculate in the coarse screening stage. ,disuse Of the individuals, 60% are eliminated in this stage. In the actuarial stage, finite element blasting simulations are performed on the remaining individuals to calculate accuracy. The fitness evaluation results are shown in Table 7:

[0148]

[0149] ③ Finally, for the optimal individual in each generation... Fine-tuned offspring were generated within the specified range, and the adjusted individuals are shown in Table 8:

[0150]

[0151] Ultimately, under the engineering scenario of a tunnel diameter of 12m, surrounding rock grade III, and local fault fracture zones, the optimal parameter combination obtained through the improved genetic algorithm is: peripheral hole spacing S = 5.83m, blasting layer thickness W = 0.76m, and charge density. With an initiation time T=50ms, this blasting scheme can reduce the over-excavation and under-excavation to 7.6cm while fully satisfying the dynamic constraints. Simulations of the optimized parameter combination were performed using ANSYS, and the simulation results are as follows: Figure 4 As shown, the total deviation between the actual blasting profile and the designed cross-section, calculated based on the profile comparison, is E = 7.6 cm, significantly lower than the standard limit. Therefore, it can be considered the theoretically optimal blasting scheme under the current geological conditions. To verify the effectiveness of the model improvement, this model was compared with the SGA (Standard Genetic Algorithm) and PSO (Particle Swarm Optimization) models. The experimental comparison indicators and results are shown in Table 9.

[0152]

[0153] As shown in Table 9, the E=8.2cm of this invention is 35% lower than that of SGA and 22% lower than that of PSO. This is mainly due to the dynamic constraints that avoid invalid solutions and the improved accuracy of the algorithm through local search. SGA, on the other hand, suffers from a surge in over- and under-digging due to parameter exceeding limits. Furthermore, this invention, through multi-stage evaluation, reduces the number of finite element simulations to 40 per generation, with a total time of 45 minutes. PSO, lacking a coarse screening mechanism, requires simulation of all 100 individuals, resulting in a longer processing time. Figure 5 The graph compares the convergence curves of the three algorithms. It shows that the solid blue line (DCGA) converges the fastest, achieving an F-value of 0.15 after 35 generations and having the lowest over- and under-mining rates. The dashed red line (SGA) converges slowly with a final F-value of 0.08, exhibiting premature convergence. The dotted red line (PSO) shows mid-term oscillations, with a final F-value of 0.12. Figure 5It can be concluded that the improved genetic algorithm (DCGA) is significantly better than the traditional algorithm in terms of convergence speed and solution quality. Figure 6 The figure shows the three-dimensional distribution of parameters and fitness. The x-axis represents the spacing between the surrounding holes, the y-axis represents the thickness of the light burst layer, and the z-axis represents the fitness. The fitness value gradually decreases from yellow to purple. It can be seen from the figure that the red marked point is the global optimal solution, which is located in the high fitness region. The fitness of parameter combinations far away from the optimal region decreases significantly.

[0154] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A blasting parameter selection system, characterized in that, The system includes a blasting parameter selection module, which contains a blasting parameter selection model. The expression of the blasting parameter selection model is: ； in, It is the sum of over-excavation and under-excavation. The elevation of the actual contour points after the blast. For design elevation, This is the Protodyakonov coefficient. The spacing between the peripheral holes, This represents the minimum value of the spacing between the surrounding holes. This represents the maximum value of the spacing between the surrounding holes. This represents the maximum permissible thickness under normal geological conditions. For safety reasons, The thickness of the photoblast layer under the fault or fracture zone. This represents the diameter of the tunnel before blasting.

2. A method for selecting blasting parameters, characterized in that, Based on the blasting parameter selection model in the blasting parameter selection system as described in claim 1, the method includes: The core parameters are the spacing between peripheral holes and the thickness of the light burst layer, while the charge density and detonation sequence are auxiliary parameters. By treating each core parameter and auxiliary parameter as an individual, an improved genetic algorithm is used to solve the blasting parameter optimization model, resulting in the combination of core parameters and the combination of auxiliary parameters. The core parameter combination and the auxiliary parameter combination are combined to obtain the globally optimal blasting parameter combination.

3. The method for selecting blasting parameters as described in claim 2, characterized in that, The improved genetic algorithm includes parallel main chain processing branches and auxiliary chain processing branches, wherein: In the main chain processing branch, based on the constraints of the peripheral hole spacing and the thickness of the blasting layer, the optimal peripheral hole spacing and the thickness of the blasting layer that minimize the over- and under-excavation under the constraints are determined. In the auxiliary chain processing branch, binary encoding is used to discretize the auxiliary parameters and generate individual optimal charge density and detonation timing.

4. The method for selecting blasting parameters as described in claim 3, characterized in that, On the main chain processing branch, based on the constraints of the peripheral hole spacing and the thickness of the blasting layer, the optimal peripheral hole spacing and blasting layer thickness that minimize over- and under-excavation under the constraints are determined, including: Obtain the surrounding rock grade and geological conditions of the current tunnel construction section, and dynamically determine the first and second feasible solution intervals for the spacing between peripheral holes and the maximum allowable thickness of the blasting layer, respectively. By using Latin hypercube sampling, N initial individuals are generated, resulting in an initial population of size N. Perform crossover and selection operations on the generated initial parameter population to generate preliminary offspring; In the mutation operation, the parameters of the offspring that are not in the first feasible solution interval or the second feasible solution interval in the initial offspring are corrected respectively; The modified offspring parameters of each generation are coarsely screened and finely screened to select the best individuals as candidate core parameters. For each offspring, the candidate core parameters are perturbed in the neighborhood, and the perturbed candidate core parameters are then corrected. Repeat the above steps until the candidate core parameters converge, and then take the candidate core parameters at this point as the optimal combination of core parameters.

5. The method for selecting blasting parameters as described in claim 4, characterized in that, In the mutation operation, the parameters of the offspring that are not in the first feasible solution interval or the second feasible solution interval in the initial offspring are corrected, including: The offspring parameters that exceed the first or second feasible solution interval are hard-truncated and corrected to the boundary value closest to the upper limit of the first or second feasible solution interval to obtain the first corrected offspring parameters; wherein, the offspring parameters include the peripheral hole spacing and the light burst layer thickness.

6. The method for selecting blasting parameters as described in claim 5, characterized in that, The process of performing coarse and fine screening on the corrected parameters of each generation of offspring to select the optimal individuals as candidate core parameters includes: Calculate the estimated over- or under-mining amount for each offspring individual, and perform a coarse screening by comparing the estimated over- or under-mining amount with the over- or under-mining threshold, and remove individuals that exceed the over- or under-mining threshold; The remaining individuals after coarse screening are subjected to finite element blasting simulation to obtain the precise over- and under-excavation amounts. A comprehensive fitness function is established based on the precise over- and under-excavation amounts, and the comprehensive fitness value of the remaining individuals is calculated through the comprehensive fitness function. The p individuals with the highest fitness values ​​are selected as candidate core parameters.

7. The method for selecting blasting parameters as described in claim 6, characterized in that, The process of perturbing the candidate core parameters of each offspring within their neighborhood and then correcting the perturbed candidate core parameters includes: A small random perturbation is added to the candidate core parameters generated for each offspring to obtain perturbed candidate core parameters; Individuals in the perturbation candidate core parameters that exceed the current first feasible solution interval or second feasible solution interval are hard-truncated and then corrected to the boundary value closest to the upper limit value of the first feasible solution interval or second feasible solution interval.

8. A device for selecting blasting parameters, characterized in that, The method for selecting blasting parameters according to any one of claims 2-7 is implemented, including: The parameter determination module is used to determine the core parameters such as the spacing between peripheral holes and the thickness of the light burst layer, and the auxiliary parameters such as the charge density and the detonation sequence. The calculation module treats each core parameter and auxiliary parameter as an individual and uses an improved genetic algorithm to solve the blasting parameter optimization model to obtain the core parameter combination and auxiliary parameter combination. The selection module is used to merge the core parameter combination and the auxiliary parameter combination to obtain the globally optimal blasting parameter combination.

9. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the blasting parameter selection method as described in any one of claims 2-7.

Images