A BIM-based intelligent construction process monitoring method and system

Abstract

The present invention discloses a BIM-based intelligent construction process monitoring method and system, belonging to the field of intelligent construction technology. By constructing a BIM corresponding to the underground excavation construction site of the subway station conversion section, and performing finite element analysis on the underground excavation construction site of the subway station conversion section based on the BIM, stress concentration areas are determined, and then safety influencing factors corresponding to the stress concentration areas are collected, and a pre-deployed machine learning model is used to identify the safety influencing factors, and support structure optimization data is determined. Finally, the support structure optimization data is transmitted to on-site workers, so that the on-site workers can carry out construction according to the support structure optimization data, thereby realizing dynamic optimization of support structure parameters, being able to adapt to the surrounding rock stress redistribution process, and effectively ensuring construction safety.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent construction, and in particular relates to a BIM-based intelligent construction process monitoring method and system. Background Art

[0002] Subway station transfer section tunneling is a construction method that involves underground excavation, primarily used for the construction of transfer sections. This construction method requires underground excavation, necessitating a series of safety measures to ensure the safety of construction workers and the surrounding environment. Subway station transfer section tunneling is typically performed using a shield machine, which can precisely excavate and advance underground. During excavation, geological conditions and groundwater levels must be continuously monitored to ensure construction safety. Furthermore, a series of support measures, such as shotcrete and steel frame installation, are also required to prevent soil collapse. Subway station transfer section tunneling has the advantage of minimizing the impact on surface traffic and the surrounding environment while also improving construction efficiency. However, subway station transfer section tunneling also carries certain risks, such as fluctuating groundwater levels and complex geological conditions. Therefore, strict safety and monitoring measures are required to ensure construction safety. In short, subway station transfer section tunneling is a complex underground engineering construction method that requires stringent safety and monitoring measures. Existing technologies often set support parameters (such as anchor length and steel frame spacing) based on empirical experience or static geological reports, without real-time adjustment based on excavation disturbances. For example, during the anchor construction process, the anchor preload is fixed and cannot adapt to the stress redistribution process of the surrounding rock, resulting in anchor failure in the local stress concentration area (such as breaking or slipping). The stiffness gradient transition cannot be achieved in the layer-biased stratum, exacerbating the stress mutation in the cross-channel connection section. Summary of the Invention

[0003] The present invention provides a BIM-based intelligent construction process monitoring method and system to solve the problem of low safety caused by setting support parameters based on experience or static geological reports in the prior art.

[0004] In one aspect, the present invention provides a BIM-based intelligent construction process monitoring method, comprising:

[0005] Construct a BIM corresponding to the underground excavation construction site of the subway station conversion section, and conduct a finite element analysis of the underground excavation construction site of the subway station conversion section based on the BIM to identify stress concentration areas;

[0006] Collecting safety influencing factors corresponding to the stress concentration area, and using a pre-deployed machine learning model to identify the safety influencing factors, and determining support structure optimization data;

[0007] The support structure optimization data is transmitted to on-site workers, so that the on-site workers can perform construction according to the support structure optimization data, thereby completing BIM-based intelligent construction process monitoring.

[0008] In one possible implementation, a BIM corresponding to the underground excavation construction site of the subway station conversion section is constructed, and a finite element analysis is performed on the underground excavation construction site of the subway station conversion section based on the BIM to determine the stress concentration area, including:

[0009] Construct a BIM corresponding to the underground excavation construction site of the subway station conversion section, and perform finite element analysis on the construction site based on the BIM to obtain stress data corresponding to the underground excavation construction site of the subway station conversion section;

[0010] The BIM corresponding to the underground excavation construction site of the subway station conversion section is evenly divided into multiple different areas. According to the stress data corresponding to the underground excavation construction site of the subway station conversion section, the areas where the stress data exceeds the preset stress threshold and is greater than the preset percentage are determined as stress concentration areas.

[0011] In a possible implementation, collecting safety influencing factors corresponding to the stress concentration area includes:

[0012] Dividing the stress concentration area into a plurality of sub-stress concentration areas, and for each sub-stress concentration area, using the stress, strain, displacement and lithology corresponding to the center of the sub-stress concentration area to obtain the safety influencing factor corresponding to the sub-stress concentration area;

[0013] The safety influencing factors corresponding to all the sub-stress concentration areas are combined to form the safety influencing factor corresponding to the stress concentration area.

[0014] In one possible implementation, a pre-deployed machine learning model is used to identify the safety influencing factors and determine support structure optimization data, including:

[0015] The safety influencing factors corresponding to the stress concentration areas are organized into a data matrix, and the data matrix is ​​input into a pre-deployed machine learning model to obtain support structure optimization data.

[0016] In one possible implementation, the pre-deployment method of the machine learning model includes:

[0017] Constructing a machine learning model capable of data matrix recognition, and initializing and encoding hyperparameters of the machine learning model to obtain multiple hyperparameter encodings;

[0018] Obtain pre-stored historical security influencing factors and supporting structure optimization data labels corresponding to the historical security influencing factors, and obtain the fitness corresponding to each hyperparameter encoding based on the historical security influencing factors and the supporting structure optimization data labels corresponding to the historical security influencing factors;

[0019] According to the fitness corresponding to the hyperparameter encoding, determining the hyperparameter encoding with the largest fitness as the optimal encoding;

[0020] A random chain search strategy is used to perform a joint chain search on the hyperparameter encoding to obtain the hyperparameter encoding after the joint chain search;

[0021] Based on the optimal code, a probabilistic decision-making balanced search strategy is used to perform a global and local balanced search on the hyperparameter code after the joint chain search to obtain the hyperparameter code after the balanced search;

[0022] Adopting an adaptive random chaos search strategy to perform a global search on the hyperparameter encoding after the equilibrium search, and obtaining the hyperparameter encoding after the global search;

[0023] Repeatedly execute the random chain search strategy, the probabilistic decision equilibrium search strategy, and the adaptive random chaos search strategy until the training end condition is met and the optimal code is obtained again;

[0024] The machine learning model is deployed based on the retrieved optimal encoding.

[0025] In one possible implementation, initializing and encoding the hyperparameters of the machine learning model to obtain multiple hyperparameter encodings includes:

[0026] Performing random initialization between upper and lower limits corresponding to hyperparameters of the machine learning model, and encoding the initialized hyperparameters into a vector to obtain an initial hyperparameter encoding;

[0027] Based on the initial hyperparameter encoding, multiple other hyperparameter encodings are obtained as follows:

[0028]

[0029] in, Indicates the k Hyperparameter encoding, and when k=1, is the initial hyperparameter encoding, Indicates the k +1 hyperparameter encoding, represents the remainder function, represents pi, sin represents the sine function, represents the first constant term and is set to 0.5; b represents the second constant term and is set to 0.2.

[0030] In one possible implementation, a random chain search strategy is used to perform a joint chain search on the hyperparameter encoding to obtain the hyperparameter encoding after the joint chain search, including:

[0031] For any hyperparameter code, randomly match the hyperparameter code with another hyperparameter code to obtain a joint search code corresponding to each hyperparameter code;

[0032] According to the fitness corresponding to all hyperparameter codes, the hyperparameter codes are arranged in descending order of fitness to obtain the arranged hyperparameter codes;

[0033] For the hyperparameter codes after permutation, a joint chain search is performed on the hyperparameter codes according to the joint search codes corresponding to the hyperparameter codes, and the hyperparameter codes after the joint chain search are obtained as follows:

[0034]

[0035] in, Indicates the t The first training i The sorted hyperparameter encoding d dimensional hyperparameters, i =1,2,..,M, where M represents the total number of hyperparameter encodings, d =1,2,..,D, where D represents the total dimension of hyperparameters in the hyperparameter encoding. Indicates the i The hyperparameter encoding after the joint chain search d dimensional hyperparameter, sin represents the sine function, cos represents the cosine function, represents pi, represents the first random number between (0,1), represents the second random number between (0,1), Indicates the i The first order of the joint search encoding corresponding to the hyperparameter encoding d dimensional hyperparameters.

[0036] In one possible implementation, based on the optimal encoding, a probabilistic decision-making balanced search strategy is used to perform a global and local balanced search on the hyperparameter encoding after the joint chain search to obtain the hyperparameter encoding after the balanced search, including:

[0037] Based on the current number of training times, the adaptive decision probability is obtained as:

[0038]

[0039]

[0040] in, represents the adaptive decision probability, Indicates the preset probability basic value, represents the Weibull distribution control factor, W represents the Weibull distribution function, T Indicates the preset maximum number of training times. t Indicates the current number of training times; represents the shape parameter of the Weibull distribution, represents the scale parameter of Weibull distribution, and e represents the natural constant;

[0041] Generate a third random number between (0, 1), and determine whether the third random number is less than the adaptive decision probability. If so, perform a Levy flight search on the hyperparameter code after the joint chain search to obtain a hyperparameter code after the equilibrium search. Otherwise, perform an optimal direction search on the hyperparameter code after the joint chain search based on the optimal code to obtain a hyperparameter code after the equilibrium search.

[0042] Perform Lévy flight search on the hyperparameter encoding after the joint chain search to obtain the hyperparameter encoding after the balanced search:

[0043]

[0044]

[0045]

[0046]

[0047]

[0048] in, Indicates the t The first training j The hyperparameter encoding after the joint chain search d dimensional hyperparameters, Indicates the j The first hyperparameter encoding after the equilibrium search d dimensional hyperparameters, j =1,2,..,M, where M represents the total number of hyperparameter encodings, Indicates the t +1 training session j The update speed of the hyperparameter encoding corresponding to the joint chain search is d Dimensional elements, represents the first inertia weight, Indicates the t The first trainingj The update speed of the hyperparameter encoding corresponding to the joint chain search is d Dimensional elements, represents the first learning factor, represents the second learning factor, represents the search range control factor set between (1,3), Indicates the j Hyperparameter encoding after joint chain search with the first random hyperparameter encoding The Euclidean distance between Indicates the j Hyperparameter encoding after joint chain search with the second random hyperparameter encoding The Euclidean distance between represents the first Lévy flight factor between (0,1), represents the second Lévy flight factor between (0,1);

[0049] Based on the optimal code, the optimal direction search is performed on the hyperparameter code after the joint chain search, and the hyperparameter code after the balanced search is obtained as follows:

[0050]

[0051]

[0052] in, The optimal coding d dimensional hyperparameters.

[0053] In a possible implementation, an adaptive random chaos search strategy is used to perform a global search on the hyperparameter encoding after the equilibrium search to obtain the hyperparameter encoding after the global search, including:

[0054] Get the chaos search control factor as:

[0055]

[0056] in, Indicates the t The chaos search control factor during the training, and at the initial moment, the chaos search control factor is randomly generated between (0,1), Indicates the t +1 chaotic search control factor during training, represents the remainder function, represents pi;

[0057] According to the chaos search control factor, a global search is performed on the hyperparameter encoding after the equilibrium search, and the hyperparameter encoding after the global search is obtained as follows:

[0058]

[0059]

[0060]

[0061]

[0062] in, Indicates the t The first training n The first hyperparameter encoding after the equilibrium search d dimensional hyperparameters, n =1,2,..,M, where M represents the total number of hyperparameter encodings, Indicates the n The first hyperparameter encoding after the global search d dimensional hyperparameters, Indicates the t +1 training session n The update speed of the hyperparameter encoding corresponding to the equilibrium search is d Dimensional elements, Indicates the t The first training n The update speed of the hyperparameter encoding corresponding to the equilibrium search is d Dimensional elements, represents the second inertia weight, represents the third learning factor, Represents the randomly generated new hyperparameter encoding d dimensional hyperparameters, represents the fourth random number uniformly distributed between (-1,1), Indicates the preset maximum value of the second inertia weight, Indicates the preset minimum value of the second inertia weight, Indicates the preset maximum value of the third learning factor, represents the preset minimum value of the third learning factor, Represents the fifth random number uniformly distributed between (0,1).

[0063] On the other hand, the present invention provides a BIM-based intelligent construction process monitoring system, comprising: a BIM data analysis module, a machine learning module, and a data feedback module;

[0064] The BIM data analysis module is used to construct a BIM corresponding to the underground excavation construction site of the subway station conversion section, and perform finite element analysis on the underground excavation construction site of the subway station conversion section based on the BIM to determine the stress concentration area;

[0065] The machine learning module is used to collect safety influencing factors corresponding to the stress concentration area, and use a pre-deployed machine learning model to identify the safety influencing factors and determine support structure optimization data;

[0066] The data feedback module is used to transmit the support structure optimization data to on-site workers, so that the on-site workers can perform construction according to the support structure optimization data and complete BIM-based intelligent construction process monitoring.

[0067] The present invention provides a BIM-based intelligent construction process monitoring method and system. The method constructs a BIM corresponding to the underground excavation construction site of the subway station conversion section, and performs finite element analysis on the underground excavation construction site of the subway station conversion section based on the BIM to determine the stress concentration area. Then, the safety influencing factors corresponding to the stress concentration area are collected, and a pre-deployed machine learning model is used to identify the safety influencing factors, and support structure optimization data is determined. Finally, the support structure optimization data is transmitted to on-site workers so that the on-site workers can perform construction according to the support structure optimization data. This realizes dynamic optimization of support structure parameters, can adapt to the surrounding rock stress redistribution process, and effectively ensures construction safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0069] Figure 1 A flowchart of a BIM-based intelligent construction process monitoring method provided in an embodiment of the present invention.

[0070] Figure 2 A structural diagram of a BIM-based intelligent construction process monitoring system provided in an embodiment of the present invention.

[0071] Among them, 201-BIM data analysis module, 202-machine learning module, and 203-data feedback module.

[0072] The above drawings illustrate specific embodiments of the present invention, which will be described in more detail below. These drawings and the accompanying description are not intended to limit the scope of the present invention in any way, but rather to illustrate the concept of the present invention to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0073] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0074] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0075] like Figure 1 As shown, an embodiment of the present invention provides a BIM-based intelligent construction process monitoring method, comprising:

[0076] S101: Construct a BIM (Building Information Modeling) corresponding to the underground excavation construction site of the subway station transfer section, and perform a finite element analysis on the underground excavation construction site of the subway station transfer section based on the BIM to determine stress concentration areas;

[0077] BIM is an overall working method based on intelligent models and digital technology. It integrates data from multiple fields to achieve information management of the entire life cycle of building assets. It not only includes the building's geometric shape, location information, physical properties, material information, etc., but also supports all participants to share data based on a unified model to improve collaboration efficiency. Therefore, the embodiment of the present invention constructs a BIM corresponding to the subway station conversion section dark excavation construction site, which can achieve comprehensive monitoring of the subway station conversion section dark excavation construction site. Then, based on the BIM, a finite element analysis is performed on the subway station conversion section dark excavation construction site, which can achieve stress analysis, thereby determining the stress concentration area and providing data support for optimizing the support structure parameters.

[0078] S102: collecting safety influencing factors corresponding to the stress concentration area, and using a pre-deployed machine learning model to identify the safety influencing factors, and determining support structure optimization data;

[0079] Safety factors can include stress, rock properties, strain, and displacement at the subway station transfer section tunnel construction site. These parameters are crucial for optimizing support structure parameters. Therefore, a pre-deployed machine learning model can be used to identify these safety factors and determine support structure optimization data. This support structure optimization data refers to the optimized support structure parameters.

[0080] S103: Transmitting the support structure optimization data to on-site workers, so that the on-site workers perform construction according to the support structure optimization data, thereby completing BIM-based intelligent construction process monitoring.

[0081] For example, after the support structure optimization data is transmitted to on-site workers, the workers can optimize the on-site support structure parameters (such as optimizing the anchor rod length and / or steel frame spacing) based on the support structure optimization data.

[0082] It is worth noting that, in order to improve the optimization accuracy, the support structure optimization data is a parameter, and the above technical solutions are implemented for different types of parameters to achieve the optimization of individual parameters.

[0083] In one possible implementation, a BIM corresponding to the underground excavation construction site of the subway station conversion section is constructed, and a finite element analysis is performed on the underground excavation construction site of the subway station conversion section based on the BIM to determine the stress concentration area, including:

[0084] Construct a BIM corresponding to the underground excavation construction site of the subway station conversion section, and perform finite element analysis on the construction site based on the BIM to obtain stress data corresponding to the underground excavation construction site of the subway station conversion section;

[0085] The BIM corresponding to the underground excavation construction site of the subway station conversion section is evenly divided into multiple different areas. According to the stress data corresponding to the underground excavation construction site of the subway station conversion section, the areas where the stress data exceeds the preset stress threshold and is greater than the preset percentage are determined as stress concentration areas.

[0086] For stress concentration areas with greater stress, stronger support is required. Therefore, the embodiments of the present invention need to optimize the support structure parameters of the stress concentration areas.

[0087] In a possible implementation, collecting safety influencing factors corresponding to the stress concentration area includes:

[0088] Dividing the stress concentration area into a plurality of sub-stress concentration areas (which may be evenly divided), and for each sub-stress concentration area, using the stress, strain, displacement and lithology corresponding to the center of the sub-stress concentration area, obtaining the safety influencing factor corresponding to the sub-stress concentration area;

[0089] The safety influencing factors corresponding to all the sub-stress concentration areas are combined to form the safety influencing factor corresponding to the stress concentration area.

[0090] By dividing the stress concentration areas into sub-areas, the data matrix can be collected, which is more conducive to subsequent data identification.

[0091] In one possible implementation, a pre-deployed machine learning model is used to identify the safety influencing factors and determine support structure optimization data, including:

[0092] The safety influencing factors corresponding to the stress concentration areas are organized into a data matrix, and the data matrix is ​​input into a pre-deployed machine learning model to obtain support structure optimization data.

[0093] Optionally, the data matrix can be normalized before being input into a pre-deployed machine learning model to reduce data complexity and improve data recognition efficiency. However, it is worth noting that if data is normalized before input, the training data also needs to be normalized during the pre-deployment of the machine learning model.

[0094] In one possible implementation, the pre-deployment method of the machine learning model includes:

[0095] A1. Construct a machine learning model capable of data matrix recognition, and initialize and encode hyperparameters of the machine learning model to obtain multiple hyperparameter codes;

[0096] For example, a convolutional neural network can be used to build a machine learning model, and then the hyperparameters of the machine learning model can be initialized and encoded using a random initialization method.

[0097] A2. Obtain pre-stored historical security influencing factors and corresponding supporting structure optimization data labels, and obtain the fitness corresponding to each hyperparameter encoding based on the historical security influencing factors and corresponding supporting structure optimization data labels;

[0098] The support structure optimization data labels corresponding to historical safety influencing factors can be labels pre-set by staff after professional analysis. Then, after data learning, the staff no longer need to carry out complex and repetitive analysis processes, and can quickly adjust the support structure parameters to ensure construction safety.

[0099] After applying hyperparameter encoding to the machine learning model, the historical safety influencing factors can be used as input data, and the supporting structure optimization data labels corresponding to the historical safety influencing factors can be used as the expected output data. Then, the loss function value corresponding to the hyperparameter encoding can be obtained, and after taking the negative of the loss function value, the fitness corresponding to the hyperparameter encoding can be obtained.

[0100] A3. According to the fitness corresponding to the hyperparameter encoding, determine the hyperparameter encoding with the largest fitness as the optimal encoding;

[0101] A4. Perform a joint chain search on the hyperparameter encoding using a random chain search strategy to obtain the hyperparameter encoding after the joint chain search.

[0102] A5. Based on the optimal code, a probabilistic decision-making balanced search strategy is used to perform a global and local balanced search on the hyperparameter code after the joint chain search to obtain a hyperparameter code after the balanced search;

[0103] A6. Using an adaptive random chaos search strategy to perform a global search on the hyperparameter encoding after the equilibrium search, to obtain the hyperparameter encoding after the global search;

[0104] A7. Repeat the random chain search strategy, the probabilistic decision equilibrium search strategy, and the adaptive random chaos search strategy until the training end condition is met and the optimal code is obtained again;

[0105] A8. Deploy the machine learning model based on the re-acquired optimal encoding, i.e., use the hyperparameters contained in the optimal encoding as the final hyperparameters of the machine learning model, and deploy the machine learning model.

[0106] Optionally, after executing the random chain search strategy, the probabilistic decision equilibrium search strategy, and the adaptive random chaos search strategy, the hyperparameters may be subjected to limit crossing processing to ensure that the hyperparameters are always within their corresponding upper and lower limits.

[0107] Prior art techniques for training machine learning models suffer from poor training results and difficulty finding a global optimal solution, resulting in an inability to accurately optimize support structure parameters. Therefore, embodiments of the present invention propose an algorithm for optimizing machine learning models to improve training speed, training results, and global training capabilities, thereby finding a global optimal solution, improving the output accuracy of the machine learning model, and effectively optimizing support structure parameters, thereby enhancing construction safety.

[0108] In one possible implementation, initializing and encoding the hyperparameters of the machine learning model to obtain multiple hyperparameter encodings includes:

[0109] Performing random initialization between upper and lower limits corresponding to hyperparameters of the machine learning model, and encoding the initialized hyperparameters into a vector to obtain an initial hyperparameter encoding;

[0110] Based on the initial hyperparameter encoding, multiple other hyperparameter encodings are obtained as follows:

[0111]

[0112] in, Indicates the k Hyperparameter encoding, and when k=1, is the initial hyperparameter encoding, Indicates the k +1 hyperparameter encoding, represents the remainder function, represents pi, sin represents the sine function, represents the first constant term and is set to 0.5; b represents the second constant term and is set to 0.2.

[0113] The initialization encoding process provided by the embodiment of the present invention can be a process in which multiple hyperparameters are encoded at the beginning of the algorithm, which can be more evenly distributed in the high-dimensional solution space, which is not only conducive to improving the training speed of the algorithm, but also can help the algorithm find the global optimal solution.

[0114] In one possible implementation, a random chain search strategy is used to perform a joint chain search on the hyperparameter encoding to obtain the hyperparameter encoding after the joint chain search, including:

[0115] For any hyperparameter code, randomly match the hyperparameter code with another hyperparameter code to obtain a joint search code corresponding to each hyperparameter code;

[0116] According to the fitness corresponding to all hyperparameter codes, the hyperparameter codes are arranged in descending order of fitness to obtain the arranged hyperparameter codes;

[0117] For the hyperparameter codes after permutation, a joint chain search is performed on the hyperparameter codes according to the joint search codes corresponding to the hyperparameter codes, and the hyperparameter codes after the joint chain search are obtained as follows:

[0118]

[0119] in, Indicates the t The first training i The sorted hyperparameter encoding d dimensional hyperparameters, i =1,2,..,M, where M represents the total number of hyperparameter encodings, d =1,2,..,D, where D represents the total dimension of hyperparameters in the hyperparameter encoding. Indicates the i The hyperparameter encoding after the joint chain search d dimensional hyperparameter, sin represents the sine function, cos represents the cosine function, represents pi, represents the first random number between (0,1), represents the second random number between (0,1), Indicates the iThe first order of the joint search encoding corresponding to the hyperparameter encoding d dimensional hyperparameters.

[0120] The embodiment of the present invention adopts a random chain search strategy, which can enable all hyperparameter encodings to be searched in a chain in the solution space while adopting a sine-cosine route for search, which is not only conducive to escaping the local optimal solution, but also can improve the search precision of the later stage of the algorithm.

[0121] In one possible implementation, based on the optimal encoding, a probabilistic decision-making balanced search strategy is used to perform a global and local balanced search on the hyperparameter encoding after the joint chain search to obtain the hyperparameter encoding after the balanced search, including:

[0122] Based on the current number of training times, the adaptive decision probability is obtained as:

[0123]

[0124]

[0125] in, represents the adaptive decision probability, Indicates the preset probability basic value, represents the Weibull distribution control factor, W represents the Weibull distribution function, T Indicates the preset maximum number of training times. t Indicates the current number of training times; represents the shape parameter of the Weibull distribution, represents the scale parameter of Weibull distribution, and e represents the natural constant;

[0126] In the early stages of the algorithm, The value is large, and the algorithm is mainly based on global search, which strengthens the global search ability and makes the hyperparameter encoding in the group closer to the optimal solution; as the number of iterations increases, The value gradually decreases, the algorithm mainly performs local search, implements refined search, and the optimization accuracy is improved; in the late iteration of the algorithm, The value has increased again, allowing the algorithm to escape from the local optimum.

[0127] Generate a third random number between (0, 1), and determine whether the third random number is less than the adaptive decision probability. If so, perform a Levy flight search on the hyperparameter code after the joint chain search to obtain a hyperparameter code after the equilibrium search. Otherwise, perform an optimal direction search on the hyperparameter code after the joint chain search based on the optimal code to obtain a hyperparameter code after the equilibrium search.

[0128] Perform Lévy flight search on the hyperparameter encoding after the joint chain search to obtain the hyperparameter encoding after the balanced search:

[0129]

[0130]

[0131]

[0132]

[0133]

[0134] in, Indicates the t The first training j The hyperparameter encoding after the joint chain search d dimensional hyperparameters, Indicates the j The first hyperparameter encoding after the equilibrium search d dimensional hyperparameters, j =1,2,..,M, where M represents the total number of hyperparameter encodings, Indicates the t +1 training session j The update speed of the hyperparameter encoding corresponding to the joint chain search is d Dimensional elements, represents the first inertia weight, Indicates the t The first training j The update speed of the hyperparameter encoding corresponding to the joint chain search is d Dimensional elements, represents the first learning factor, represents the second learning factor, represents the search range control factor set between (1,3), Indicates the j Hyperparameter encoding after joint chain search with the first random hyperparameter encoding The Euclidean distance between Indicates the j Hyperparameter encoding after joint chain search with the second random hyperparameter encoding The Euclidean distance between represents the first Lévy flight factor between (0,1), represents the second Lévy flight factor between (0,1);

[0135] Based on the optimal code, the optimal direction search is performed on the hyperparameter code after the joint chain search, and the hyperparameter code after the balanced search is obtained as follows:

[0136]

[0137]

[0138] in, The optimal coding d dimensional hyperparameters.

[0139] This embodiment of the present invention employs a probabilistic decision-making equilibrium search strategy, which can be performed in two different search modes. In the early stages of the algorithm, the Levy flight search is the primary method, which performs adaptive learning and, in combination with Levy flights, provides a certain degree of global search capability. In the later stages of the algorithm, the optimal direction search is the primary method, which enables rapid algorithm convergence and improves convergence accuracy.

[0140] In a possible implementation, an adaptive random chaos search strategy is used to perform a global search on the hyperparameter encoding after the equilibrium search to obtain the hyperparameter encoding after the global search, including:

[0141] Get the chaos search control factor as:

[0142]

[0143] in, Indicates the t The chaos search control factor during the training, and at the initial moment, the chaos search control factor is randomly generated between (0,1), Indicates the t +1 chaotic search control factor during training, represents the remainder function, represents pi;

[0144] According to the chaos search control factor, a global search is performed on the hyperparameter encoding after the equilibrium search, and the hyperparameter encoding after the global search is obtained as follows:

[0145]

[0146]

[0147]

[0148]

[0149] in, Indicates the t The first training nThe first hyperparameter encoding after the equilibrium search d dimensional hyperparameters, n =1,2,..,M, where M represents the total number of hyperparameter encodings, Indicates the n The first hyperparameter encoding after the global search d dimensional hyperparameters, Indicates the t +1 training session n The update speed of the hyperparameter encoding corresponding to the equilibrium search is d Dimensional elements, Indicates the t The first training n The update speed of the hyperparameter encoding corresponding to the equilibrium search is d Dimensional elements, represents the second inertia weight, represents the third learning factor, Represents the randomly generated new hyperparameter encoding d dimensional hyperparameters, represents the fourth random number uniformly distributed between (-1,1), Indicates the preset maximum value of the second inertia weight, Indicates the preset minimum value of the second inertia weight, Indicates the preset maximum value of the third learning factor, represents the preset minimum value of the third learning factor, Represents the fifth random number uniformly distributed between (0,1).

[0150] The adaptive random chaotic search strategy provided by the embodiment of the present invention makes the algorithm in a chaotic state during the search, which is more conducive to jumping out of the local optimal solution, improving the powerful global search capability, and setting adaptive search parameters, which will not excessively affect the convergence of the algorithm in the later stage.

[0151] Optionally, an annealing simulation algorithm or a greedy strategy may be used to control the search process of the adaptive random chaos search strategy to increase the training speed of the algorithm.

[0152] The present invention provides a BIM-based intelligent construction process monitoring method and system. The method constructs a BIM corresponding to the underground excavation construction site of the subway station conversion section, and performs finite element analysis on the underground excavation construction site of the subway station conversion section based on the BIM to determine the stress concentration area. Then, the safety influencing factors corresponding to the stress concentration area are collected, and a pre-deployed machine learning model is used to identify the safety influencing factors, and support structure optimization data is determined. Finally, the support structure optimization data is transmitted to on-site workers so that the on-site workers can perform construction according to the support structure optimization data. This realizes dynamic optimization of support structure parameters, can adapt to the surrounding rock stress redistribution process, and effectively ensures construction safety.

[0153] like Figure 2 As shown, the present invention provides a BIM-based intelligent construction process monitoring system, comprising: a BIM data analysis module 201, a machine learning module 202, and a data feedback module 203;

[0154] The BIM data analysis module 201 is used to construct a BIM corresponding to the underground excavation construction site of the subway station conversion section, and perform finite element analysis on the underground excavation construction site of the subway station conversion section based on the BIM to determine the stress concentration area;

[0155] The machine learning module 202 is used to collect safety influencing factors corresponding to the stress concentration area, and use a pre-deployed machine learning model to identify the safety influencing factors and determine support structure optimization data;

[0156] The data feedback module 203 is used to transmit the support structure optimization data to the on-site workers, so that the on-site workers can perform construction according to the support structure optimization data and complete the BIM-based intelligent construction process monitoring.

[0157] An embodiment of the present invention provides a BIM-based intelligent construction process monitoring system that can implement the above-mentioned method and technical solution. Its principles and beneficial effects are similar and will not be repeated here.

[0158] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0159] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0160] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0161] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0162] Those skilled in the art will understand that all or part of the steps in implementing the above facts and methods can be completed by instructing relevant hardware through a program, and the program involved or the program can be stored in a computer-readable storage medium. When the program is executed, it includes the following steps: the corresponding method steps are then brought out, and the storage medium can be ROM / RAM, a disk, an optical disk, etc.

[0163] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A BIM-based intelligent construction process monitoring method, characterized in that: include: Construct a BIM corresponding to the underground excavation construction site of the subway station conversion section, and conduct a finite element analysis of the underground excavation construction site of the subway station conversion section based on the BIM to identify stress concentration areas; Collecting safety influencing factors corresponding to the stress concentration area, and using a pre-deployed machine learning model to identify the safety influencing factors, and determining support structure optimization data; Transmitting the support structure optimization data to on-site workers so that they can perform construction according to the support structure optimization data and complete BIM-based intelligent construction process monitoring; The pre-deployment method of the machine learning model includes: Constructing a machine learning model capable of data matrix recognition, and initializing and encoding hyperparameters of the machine learning model to obtain multiple hyperparameter encodings; Obtain pre-stored historical security influencing factors and supporting structure optimization data labels corresponding to the historical security influencing factors, and obtain the fitness corresponding to each hyperparameter encoding based on the historical security influencing factors and the supporting structure optimization data labels corresponding to the historical security influencing factors; According to the fitness corresponding to the hyperparameter encoding, determining the hyperparameter encoding with the largest fitness as the optimal encoding; A random chain search strategy is used to perform a joint chain search on the hyperparameter encoding to obtain the hyperparameter encoding after the joint chain search; Based on the optimal code, a probabilistic decision-making balanced search strategy is used to perform a global and local balanced search on the hyperparameter code after the joint chain search to obtain the hyperparameter code after the balanced search; Adopting an adaptive random chaos search strategy to perform a global search on the hyperparameter encoding after the equilibrium search, and obtaining the hyperparameter encoding after the global search; Repeatedly execute the random chain search strategy, the probabilistic decision equilibrium search strategy, and the adaptive random chaos search strategy until the training end condition is met and the optimal code is obtained again; deploying the machine learning model based on the retrieved optimal encoding; Initialize and encode the hyperparameters of the machine learning model to obtain multiple hyperparameter codes, including: Performing random initialization between upper and lower limits corresponding to hyperparameters of the machine learning model, and encoding the initialized hyperparameters into a vector to obtain an initial hyperparameter encoding; Based on the initial hyperparameter encoding, multiple other hyperparameter encodings are obtained as follows: in, Indicates the k Hyperparameter encoding, and when k=1, it is Initial hyperparameter encoding, Indicates the k +1 hyperparameter encoding, represents the remainder function, represents pi, sin represents the sine function, represents the first constant term and is set to 0.5; b represents the second constant term and is set to 0.2; A random chain search strategy is used to perform a joint chain search on the hyperparameter encoding, and the hyperparameter encoding after the joint chain search is obtained, including: For any hyperparameter code, randomly match the hyperparameter code with another hyperparameter code to obtain a joint search code corresponding to each hyperparameter code; According to the fitness corresponding to all hyperparameter codes, the hyperparameter codes are arranged in descending order of fitness to obtain the arranged hyperparameter codes; For the hyperparameter codes after permutation, a joint chain search is performed on the hyperparameter codes according to the joint search codes corresponding to the hyperparameter codes, and the hyperparameter codes after the joint chain search are obtained as follows: in, Indicates the t The first training i The sorted hyperparameter encoding d dimensional hyperparameters, i =1,2,..,M, where M represents the total number of hyperparameter encodings, d =1,2,..,D, where D represents the total dimension of hyperparameters in the hyperparameter encoding. Indicates the i The hyperparameter encoding after the joint chain search d dimensional hyperparameter, sin represents the sine function, cos represents the cosine function, represents pi, represents the first random number between (0,1), represents the second random number between (0,1), Indicates the i The first order of the joint search encoding corresponding to the hyperparameter encoding d dimensional hyperparameters; Based on the optimal encoding, a probabilistic decision-making balanced search strategy is used to perform global and local balanced searches on the hyperparameter encoding after the joint chain search, and the hyperparameter encoding after the balanced search is obtained, including: Based on the current number of training times, the adaptive decision probability is obtained as: in, represents the adaptive decision probability, Indicates the preset probability basic value, represents the Weibull distribution control factor, W represents the Weibull distribution function, T Indicates the preset maximum number of training times. t Indicates the current number of training times; represents the shape parameter of the Weibull distribution, represents the scale parameter of Weibull distribution, and e represents the natural constant; Generate a third random number between (0, 1), and determine whether the third random number is less than the adaptive decision probability. If so, perform a Levy flight search on the hyperparameter code after the joint chain search to obtain a hyperparameter code after the equilibrium search. Otherwise, perform an optimal direction search on the hyperparameter code after the joint chain search based on the optimal code to obtain a hyperparameter code after the equilibrium search. Perform Lévy flight search on the hyperparameter encoding after the joint chain search, and obtain the hyperparameter encoding after the balanced search as follows: in, Indicates the t The first training j The hyperparameter encoding after the joint chain search d dimensional hyperparameters, Indicates the j The first hyperparameter encoding after the equilibrium search d dimensional hyperparameters, j =1,2,..,M, where M represents the total number of hyperparameter encodings, Indicates the t +1 training session j The update speed of the hyperparameter encoding corresponding to the joint chain search is d Dimensional elements, represents the first inertia weight, Indicates the t The first training j The update speed of the hyperparameter encoding corresponding to the joint chain search is d Dimensional elements, represents the first learning factor, represents the second learning factor, represents the search range control factor set between (1,3), Indicates the j Hyperparameter encoding after joint chain search with the first random hyperparameter encoding The Euclidean distance between Indicates the j Hyperparameter encoding after joint chain search with the second random hyperparameter encoding The Euclidean distance between represents the first Lévy flight factor between (0,1), represents the second Lévy flight factor between (0,1); Based on the optimal code, the optimal direction search is performed on the hyperparameter code after the joint chain search, and the hyperparameter code after the balanced search is obtained as follows: in, The optimal coding d dimensional hyperparameters.

2. The BIM-based intelligent construction process monitoring method according to claim 1, characterized in that: Construct a BIM corresponding to the underground excavation construction site of the subway station conversion section, and conduct a finite element analysis of the underground excavation construction site of the subway station conversion section based on the BIM to identify stress concentration areas, including: Construct a BIM corresponding to the underground excavation construction site of the subway station conversion section, and perform finite element analysis on the construction site based on the BIM to obtain stress data corresponding to the underground excavation construction site of the subway station conversion section; The BIM corresponding to the underground excavation construction site of the subway station conversion section is evenly divided into multiple different areas. According to the stress data corresponding to the underground excavation construction site of the subway station conversion section, the areas where the stress data exceeds the preset stress threshold and is greater than the preset percentage are determined as stress concentration areas.

3. The BIM-based intelligent construction process monitoring method according to claim 1, characterized in that: Collect the safety influencing factors corresponding to the stress concentration area, including: Dividing the stress concentration area into a plurality of sub-stress concentration areas, and for each sub-stress concentration area, using the stress, strain, displacement and lithology corresponding to the center of the sub-stress concentration area to obtain the safety influencing factor corresponding to the sub-stress concentration area; The safety influencing factors corresponding to all the sub-stress concentration areas are combined to form the safety influencing factor corresponding to the stress concentration area.

4. The BIM-based intelligent construction process monitoring method according to claim 3 is characterized in that: Use pre-deployed machine learning models to identify the safety impact factors and determine supporting structure optimization data, including: The safety influencing factors corresponding to the stress concentration areas are organized into a data matrix, and the data matrix is ​​input into a pre-deployed machine learning model to obtain support structure optimization data.

5. The BIM-based intelligent construction process monitoring method according to claim 1, characterized in that: An adaptive random chaos search strategy is used to perform a global search on the hyperparameter encoding after the equilibrium search, and the hyperparameter encoding after the global search is obtained, including: Get the chaos search control factor as: in, Indicates the t The chaos search control factor during the training, and at the initial moment, the chaos search control factor is randomly generated between (0,1), Indicates the t +1 chaotic search control factor during training, represents the remainder function, represents pi; According to the chaos search control factor, a global search is performed on the hyperparameter encoding after the equilibrium search, and the hyperparameter encoding after the global search is obtained as follows: in, Indicates the t The first training n The first hyperparameter encoding after the equilibrium search d dimensional hyperparameters, n =1,2,..,M, where M represents the total number of hyperparameter encodings, Indicates the n The first hyperparameter encoding after the global search d dimensional hyperparameters, Indicates the t +1 training session n The update speed of the hyperparameter encoding corresponding to the equilibrium search is d Dimensional elements, Indicates the t The first training n The update speed of the hyperparameter encoding corresponding to the equilibrium search is d Dimensional elements, represents the second inertia weight, represents the third learning factor, Represents the randomly generated new hyperparameter encoding d dimensional hyperparameters, represents the fourth random number uniformly distributed between (-1,1), Indicates the preset maximum value of the second inertia weight, Indicates the preset minimum value of the second inertia weight, Indicates the preset maximum value of the third learning factor, represents the preset minimum value of the third learning factor, Represents the fifth random number uniformly distributed between (0,1).

6. A BIM-based intelligent construction process monitoring system, which is capable of executing the BIM-based intelligent construction process monitoring method according to any one of claims 1 to 5, characterized in that: include: BIM data analysis module, machine learning module, and data feedback module; The BIM data analysis module is used to construct a BIM corresponding to the underground excavation construction site of the subway station conversion section, and perform finite element analysis on the underground excavation construction site of the subway station conversion section based on the BIM to determine the stress concentration area; The machine learning module is used to collect safety influencing factors corresponding to the stress concentration area, and use a pre-deployed machine learning model to identify the safety influencing factors and determine support structure optimization data; The data feedback module is used to transmit the support structure optimization data to on-site workers, so that the on-site workers can perform construction according to the support structure optimization data and complete BIM-based intelligent construction process monitoring.

Images