Tunnel surrounding rock displacement evolution prediction method and system based on improved ensemble learning

Abstract

The invention provides a tunnel surrounding rock displacement evolution prediction method and system based on improved ensemble learning, and the method comprises the steps: inputting obtained tunnel geological data into n base learners, fusing the output results of the n base learners based on the reputation value corresponding to each base learner, obtaining a fusion result, and obtaining a prediction result; and obtaining a tunnel surrounding rock displacement prediction result according to the fusion result. According to the method, the output results of a plurality of base learners are fused, fusion is carried out based on the reputation value as the weight of each base learner, and compared with single-feature and single-model diagnosis, the accuracy and reliability of the diagnosis result are higher.

Description

Technical Field

[0001] The present invention belongs to the technical field related to tunnel surrounding rock displacement prediction, and in particular to a tunnel surrounding rock displacement evolution prediction method and system based on improved ensemble learning. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] During tunnel construction, the stability of the surrounding rock is a key factor in ensuring the safety and success of the project. The evolution of surrounding rock displacement directly affects the stability of the tunnel structure and construction safety. However, the prediction of surrounding rock displacement is a complex issue, which is affected by a variety of geological and construction factors, and its dynamic changes are highly nonlinear and uncertain. Traditional prediction methods, such as linear regression, often have difficulty accurately capturing these complex evolution laws, resulting in insufficient accuracy of the prediction results. Summary of the invention

[0004] In order to overcome the shortcomings of the above-mentioned prior art, the present invention provides a method and system for predicting the displacement evolution of tunnel surrounding rock based on improved ensemble learning. The present invention integrates the output results of multiple base learners, and integrates them based on reputation value as the weight of each base learner. Compared with single model diagnosis, the accuracy and reliability of the diagnosis result are higher.

[0005] To achieve the above object, the first aspect of the present invention provides a method for predicting the displacement evolution of surrounding rock of a tunnel based on improved ensemble learning, comprising:

[0006] Obtain tunnel geological data;

[0007] The acquired tunnel geological data are respectively input into n pre-built basis learners to obtain output results of the n basis learners, where n is a positive integer;

[0008] The reputation values ​​corresponding to the n base learners are used as coefficients and fused with the output results of the n base learners. The prediction results of tunnel surrounding rock displacement are obtained according to the fusion results.

[0009] A second aspect of the present invention provides a tunnel surrounding rock displacement evolution prediction system based on improved ensemble learning, comprising:

[0010] An acquisition unit, used for acquiring tunnel geological data;

[0011] A processing unit, used for inputting the acquired tunnel geological data into n pre-built basis learners respectively, to obtain output results of the n basis learners, where n is a positive integer;

[0012] The fusion unit is used to use the reputation values ​​corresponding to the n base learners as coefficients to fuse the output results of the n base learners, and obtain the tunnel surrounding rock displacement prediction result according to the fusion result.

[0013] The third aspect of the present invention provides a computer device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the computer device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, a method for predicting the evolution of tunnel surrounding rock displacement based on improved ensemble learning is executed.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, a method for predicting the displacement evolution of tunnel surrounding rock based on improved ensemble learning is executed.

[0015] A fifth aspect of the present invention provides a computer program product, comprising a computer program, which, when executed by a processor, implements a method for predicting the evolution of tunnel surrounding rock displacement based on improved ensemble learning as described in the first aspect.

[0016] One or more of the above technical solutions have the following beneficial effects:

[0017] In the present invention, the acquired tunnel geological data are respectively input into n base learners, and the output results of the n base learners are fused based on the reputation values ​​corresponding to the base learners to obtain a fusion result, and the tunnel surrounding rock displacement prediction result is obtained according to the fusion result. The present invention fuses the output results of multiple base learners, and fuses them based on the reputation value as the coefficient of each base learner. Compared with the single model diagnosis, the accuracy and reliability of the diagnosis result are higher.

[0018] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0020] Figure 1 This is a flow chart of prediction of tunnel surrounding rock displacement evolution law based on improved ensemble learning in Embodiment 1 of the present invention;

[0021] Figure 2 This is an overall block diagram of the prediction of the evolution law of tunnel surrounding rock displacement based on improved ensemble learning in the first embodiment of the present invention. DETAILED DESCRIPTION

[0022] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0023] It should be noted that the terms used herein are for describing specific embodiments only and are not intended to be limiting of exemplary embodiments according to the present invention.

[0024] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0025] Embodiment 1

[0026] like Figure 1-Figure 2 As shown, this embodiment discloses a method for predicting the displacement evolution of tunnel surrounding rock based on improved ensemble learning, comprising:

[0027] Obtain tunnel geological data;

[0028] The acquired tunnel geological data are respectively input into n pre-built basis learners to obtain output results of the n basis learners, where n is a positive integer;

[0029] The reputation values ​​corresponding to the n base learners are used as coefficients and fused with the output results of the n base learners. The prediction results of tunnel surrounding rock displacement are obtained according to the fusion results.

[0030] In this embodiment, m base learners are established, and the training set is used as the input of each base learner. For each base learner, the output result is compared with the actual value. If the absolute value of the error is less than K, the base learner is retained. If the absolute value of the error is greater than or equal to K, the base learner is eliminated. Finally, n base learners are retained. The available training data is evenly divided into n+1 subsets, and each subset is selected in turn as a verification set in an independent cycle. The remaining n subsets are mapped one by one to form a data set for training n base learners to obtain a base learner feature matrix. The reputation method is introduced to operate with the prediction feature matrix output by the base learner to obtain new features; the clustering algorithm is introduced to set T labels, and the clustering algorithm is combined with the base learner to obtain a reinforcement learner; the new features are input to predict the final result. After that, a user prediction information interaction interface is developed based on Unity, and the parameters submitted by the user of the interaction interface are input into the agent prediction model. The model prediction results are output to the prediction information interaction interface, and engineers can observe the displacement evolution law more intuitively.

[0031] Combine the following Figure 2The tunnel surrounding rock displacement evolution prediction method based on improved ensemble learning proposed in this embodiment is described in detail, including:

[0032] Step 1: Construct a displacement prediction model.

[0033] Step 1-1: Obtain tunnel geological information data, including stratum distribution data, construction location information data, support information data, obtain monitoring information data, including time and space displacement data, perform numerical simulation on the acquired field data based on FLAC, and obtain surrounding rock displacement data, including tunnel burial depth, excavation progress, etc.

[0034] The base learner is trained according to the tunnel address data and the surrounding rock displacement data obtained by simulation, and the training data is divided into a training set for model training and a test set for evaluating model performance.

[0035] Step 1-2: Establish m basis learners L = {L1, L2, ..., L m}.

[0036] Step 1-3: Evenly divide the available training data into n+1 subsets S = {S1, S2, ..., S n+1}. Select each subset S in turn k As the validation set, the remaining n subsets constitute the training set as follows:

[0037] D train (k)=S1∪S2∪...∪S k-1 ∪S k+1 ∪...∪S n (1)

[0038] Using the training set D train(k) Train n retained base learners as follows:

[0039] L={L1,L2,...,L n}

[0040] Get the base learner training result matrix M = {M1, M2, ..., M n}.

[0041] For each base learner, calculate its prediction result The absolute value of the error between the actual value y is as follows:

[0042]

[0043] Where N is the number of samples in the training set, y j is the actual value of the jth sample, is the base learner L iis the predicted value of the jth sample.

[0044] Step 1-4: Set the threshold K, if (E i <K), then retain the base learner L i , otherwise it is removed. Finally, n base learners are retained.

[0045] Step 2: Obtain on-site tunnel geological data, geological information data, including stratum distribution data, construction location information data, support information data, monitoring information data, including time and space displacement data, and make predictions based on the trained base learner.

[0046] Step 2-1: Introduce reputation-weighted majority voting method, using accuracy as performance indicator and reputation value R i Defined as:

[0047]

[0048] Among them, TP is the number of true positives, TN is the number of true negatives, and Total Samples is the total number of samples in the validation set.

[0049] Step 2-2: Normalize the reputation value to ensure that the sum of the reputation value is 1. The formula is as follows:

[0050]

[0051] Among them, w i is the normalized reputation value, and n is the number of base learners.

[0052] Step 2-3: For the acquired tunnel geological data to be predicted, based on the trained n basis learners, obtain the prediction result matrix of the n basis learners, according to the obtained prediction result matrix.

[0053] During the training process of the base learner, the cluster center, ie, the threshold, determined according to each category label determined below is used as the output result of the base learner corresponding to the training data.

[0054] Set the category label c, including c1 for no displacement, c2 for slight displacement, c3 for moderate displacement, and c4 for severe displacement, and standardize the data:

[0055]

[0056] Among them, x is the original data, μ is the mean of the data, and σ is the standard deviation of the data.

[0057] Determine the number of clusters to be 4, randomly select 4 data points as the starting cluster centers, and assign each data point to the nearest cluster center A j , the formula is as follows:

[0058]

[0059] Among them, x i is a data point, c j，i is the i-th eigenvalue of the j-th cluster center.

[0060] For each data point, it is classified according to the distance from each cluster center according to the Euclidean distance formula, and the center of each cluster is updated to the mean of the cluster members. The above process is repeated until the cluster center no longer changes, and the threshold is defined according to the cluster center.

[0061] Among the four cluster centers, the value between the first and second centers can be used as the upper limit of "slight displacement", and the value between the second and third centers can be used as the upper limit of "moderate displacement".

[0062] The specific threshold settings are as follows: no displacement: c1 = T1; slight displacement: c2 = T2; moderate displacement: c3 = T3; severe displacement: c4 = T4.

[0063] Step 2-4: Use the normalized reputation value w i , perform dot product operation on the prediction result matrix obtained by the base learner to obtain the new feature w m , input reinforcement learning machine G(x), the formula is as follows:

[0064]

[0065] in, is an indicator function. If the mth base learner predicts that the sample x belongs to category c4, the function value is 0, otherwise it is 1.

[0066] Step 4-5: Predict the tunnel surrounding rock displacement based on G(x).

[0067] Step 5: Establish a two-way communication mechanism.

[0068] Step 5-1: Create an HTTP request, create a service based on the POST protocol in Python, ensure that the Python service listens to and processes the POST request, returns the corresponding data, and generates an executable file on the Python side, which is started in the background when running this system.

[0069] Step 5-2: Unity stores the data in a JSON file and sends it to the Python server as part of an HTTP request.

[0070] Step 5-3: On the Python server side, use the Flask framework to receive HTTP requests and obtain data from the requests. Python receives the data from the Unity side and uses this data as the input of the tunnel surrounding rock displacement deep learning agent model to make predictions. The predicted result data is returned to the Unity client as the required response data.

[0071] Step 6: Use the Unity engine to design prediction information and interactive interface. Use the Dropdown component of the GUI in Unity to create a custom selection box, customize the excavation footage, use the Input Field component of the GUI in Unity to create an input box, enter the tunnel depth, enter the surrounding rock parameters, including cohesion, internal friction angle and elastic modulus, use the Button component of the GUI in Unity to create a confirmation option, click the confirmation option, and transfer the parameters of tunnel depth, excavation footage, cohesion, internal friction angle and elastic modulus to the established Python tunnel surrounding rock displacement prediction model to obtain the final tunnel surrounding rock displacement prediction result, and finally transfer it to Unity and display it in the interactive interface. The surrounding rock displacement prediction model interactive interface is shown as follows: Figure 2 shown.

[0072] Embodiment 2

[0073] The purpose of this embodiment is to provide a tunnel surrounding rock displacement evolution prediction system based on improved ensemble learning, including:

[0074] An acquisition unit, used for acquiring tunnel geological data;

[0075] A processing unit, used for inputting the acquired tunnel geological data into n pre-built basis learners respectively, to obtain output results of the n basis learners, where n is a positive integer;

[0076] The fusion unit is used to use the reputation values ​​corresponding to the n base learners as coefficients to fuse the output results of the n base learners, and obtain the tunnel surrounding rock displacement prediction result according to the fusion result.

[0077] Embodiment 3

[0078] The purpose of this embodiment is to provide a computing device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the program.

[0079] Embodiment 4

[0080] The purpose of this embodiment is to provide a computer-readable storage medium.

[0081] A computer-readable storage medium stores a computer program, which executes the steps of the above method when executed by a processor.

[0082] Embodiment 5

[0083] The purpose of this embodiment is to provide a computer program product, including a computer program, which implements the method described in Embodiment 1 when executed by a processor.

[0084] The steps involved in the apparatuses of the above embodiments 2, 3 and 4 correspond to the method embodiment 1, and the specific implementation methods can refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0085] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0086] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A tunnel surrounding rock displacement evolution prediction method based on improved ensemble learning, characterized in that: include: Obtain tunnel geological data; The acquired tunnel geological data are respectively input into n pre-built basis learners to obtain output results of the n basis learners, where n is a positive integer; The reputation values ​​corresponding to the n base learners are used as coefficients and fused with the output results of the n base learners. The prediction results of tunnel surrounding rock displacement are obtained according to the fusion results.

2. The method for predicting the displacement evolution of surrounding rock of a tunnel based on improved ensemble learning according to claim 1, characterized in that: The method also includes: extracting training samples from historical tunnel geological data, inputting the extracted training samples into m base learners, comparing the error between the output result of each base learner and the actual value, eliminating the base learners corresponding to the error values ​​greater than k, and obtaining n base learners.

3. The method for predicting the displacement evolution of surrounding rock of a tunnel based on improved ensemble learning according to claim 1, characterized in that: The accuracy of the prediction of the base learner is used as the reputation value, and the calculated reputation value is normalized and used as the coefficient of the result predicted by the corresponding base learner.

4. The method for predicting displacement evolution of surrounding rock of a tunnel based on improved ensemble learning according to claim 1, characterized in that: The geological data include: stratum distribution data, construction location information data, and support information data.

5. The tunnel surrounding rock displacement evolution prediction method based on improved ensemble learning as claimed in claim 1, wherein the reputation values ​​corresponding to the n base learners are used as coefficients, and the output results of the n base learners are fused, and the tunnel surrounding rock displacement prediction result is obtained according to the fusion result, which is specifically: in, G(x) represents the reinforcement learner, ω m is the dot product result of the output of n base learners and the corresponding coefficients, is the indicator function, G m (x) indicates that the predicted sample x of the mth base learner belongs to the category label c.

6. A tunnel surrounding rock displacement evolution prediction method based on improved ensemble learning as described in claim 5, determining category labels and corresponding cluster centers, classifying by calculating the distance between each data point and each cluster center according to Euclidean method, updating each cluster center to the mean of the cluster members, repeating the above process until the cluster center no longer changes, and training the base learner according to the determined cluster center as the output result corresponding to the training data.

7. A tunnel surrounding rock displacement evolution prediction system based on improved ensemble learning, characterized in that: include: An acquisition unit, used for acquiring tunnel geological data; A processing unit, used for inputting the acquired tunnel geological data into n pre-built basis learners respectively, to obtain output results of the n basis learners, where n is a positive integer; The fusion unit is used to use the reputation values ​​corresponding to the n base learners as coefficients to fuse the output results of the n base learners, and obtain the tunnel surrounding rock displacement prediction result according to the fusion result.

8. A computer device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the computer device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, a method for predicting the displacement evolution of surrounding rock of a tunnel based on improved ensemble learning as described in any one of claims 1 to 6 is executed.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the method for predicting the displacement evolution of tunnel surrounding rock based on improved ensemble learning as described in any one of claims 1 to 6 is executed.

10. A computer program product, characterized in that It includes a computer program, which, when executed by a processor, implements a tunnel surrounding rock displacement evolution prediction method based on improved ensemble learning as described in any one of claims 1 to 6.

Images