Intelligent grouting pressure adjusting method and system based on deep learning

Abstract

The application discloses a grouting pressure intelligent adjusting method and system based on deep learning, relates to the technical field of grouting pressure adjusting, and obtains geological geophysical prospecting data of a grouting area before grouting construction; parameters of a surrounding environment influence, an active regulation and control pressure pulse and a grouting equipment itself pressure pulse are taken as a state S describing grouting diffusion, each parameter in the state S is input into a pre-trained grouting pressure influence prediction model, and corresponding grouting effects are obtained; according to the geological geophysical prospecting data, a required grouting diffusion threshold state is calculated, and each parameter in the state S, the corresponding grouting effects and the grouting diffusion threshold state are input into a pre-trained grouting pressure regulation and control model; the model takes the grouting diffusion threshold state as a target of the grouting effects, and outputs a grouting pressure adjusting operation; according to real-time feedback of actual grouting conditions and actual grouting parameters, the grouting pressure regulation and control model is updated, and the grouting pressure is adjusted based on the model, so that automatic and accurate adjustment of the grouting pressure is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of grouting pressure regulation, and particularly relates to a grouting pressure intelligent regulation method and system based on deep learning. BACKGROUND

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

[0003] In the process of underground engineering construction, many complex geological sections such as extremely water-rich or high-pressure conditions are often encountered, and the regional rock-soil mass often exhibits relatively weak and disturbed local collapse properties, which will threaten the safety of construction personnel, equipment and normal production of the project. At present, the disasters in the unfavorable geological section of underground engineering are often treated by grouting, and the grout is injected into the stratum through the grouting pipe to cement the original loose structure or fracture into a whole, so as to achieve the effect of plugging water gushing and reinforcing surrounding rock.

[0004] In the actual grouting process, the grouting pressure is often regulated by manual control due to the influence of grouting equipment and environment. Manual control is strong, and it is difficult to accurately determine the required grouting pressure according to the actual stratum conditions, which makes it difficult to control the diffusion range of the grout and the actual grouting effect. That is, at present, the pressure regulation in the grouting process is often realized by manually controlling part of the grouting equipment and manually adjusting the change of the grouting pressure according to the grouting conditions. However, in the actual grouting process, in addition to the influence of the pressure fluctuation caused by manual regulation (i.e. active regulation of pressure pulse), due to the limitation of the grouting equipment itself, the delivery of the grout is not stable and continuous, and the pressure pulse of the grouting equipment itself will also affect the pressure regulation in the grouting process. The existing manual pressure regulation method cannot accurately consider the influence of the pressure pulse of the grouting equipment itself, and does not comprehensively consider the comprehensive influence of the equipment and the automatic regulation of the pressure pulse in the grouting process, so it is difficult to realize the automatic and intelligent regulation of the grouting pressure considering the stratum environment. SUMMARY

[0005] To solve the above problems of the prior art, the present application provides a grouting pressure intelligent regulation method and system based on deep learning, which realizes the automatic, real-time and accurate regulation of the grouting pressure by constructing a grouting pressure influence prediction model and a grouting pressure regulation model, and then feeding back the real-time data of the actual geological conditions and grouting according to the actual geological conditions and grouting data.

[0006] In a first aspect, the present application provides a grouting pressure intelligent regulation method based on deep learning.

[0007] A grouting pressure intelligent regulation method based on deep learning, comprising:

[0008] Obtaining geological geophysical data of a region to be grouted before grouting construction;

[0009] The state S of the surrounding environment influence, the active regulation pressure pulse, and the parameters of the pressure pulse of the grouting equipment itself are taken as a description of grouting diffusion, and each parameter in the state S is input into a pre-trained grouting pressure influence prediction model to obtain a corresponding grouting effect.

[0010] According to the geological geophysical data, the required grouting diffusion threshold state is calculated, and each parameter in the state S, the corresponding grouting effect, and the grouting diffusion threshold state are input into a pre-trained grouting pressure regulation model.

[0011] According to the real-time feedback of the actual grouting situation and the actual grouting parameters, the grouting pressure regulation model is updated, and the updated model is used for grouting pressure regulation.

[0012] In a second aspect, the present application provides a grouting pressure intelligent regulation system based on deep learning.

[0013] A grouting pressure intelligent regulation system based on deep learning comprises:

[0014] A data acquisition module is configured to acquire geological geophysical data of a grouting area before grouting construction.

[0015] A grouting effect feedback module is configured to take the surrounding environment influence, the active regulation pressure pulse, and the parameters of the pressure pulse of the grouting equipment itself as a description of grouting diffusion, and input each parameter in the state S into a pre-trained grouting pressure influence prediction model to obtain a corresponding grouting effect.

[0016] A grouting pressure regulation module is configured to calculate a required grouting diffusion threshold state according to the geological geophysical data, and input each parameter in the state S, the corresponding grouting effect, and the grouting diffusion threshold state into a pre-trained grouting pressure regulation model.

[0017] A grouting pressure regulation update module is configured to update the grouting pressure regulation model according to the real-time feedback of the actual grouting situation and the actual grouting parameters, and use the updated model for grouting pressure regulation.

[0018] In a third aspect, the present application further provides an electronic device comprising a memory for storing executable instructions and a processor for executing the executable instructions stored in the memory to implement the above-mentioned grouting pressure intelligent regulation method based on deep learning.

[0019] In a fourth aspect, the present application also provides a computer readable storage medium storing executable instructions for causing a processor to implement the deep learning-based grouting pressure intelligent adjustment method described above when the executable instructions are executed by the processor.

[0020] In a fifth aspect, the present application also provides a computer program product comprising executable instructions stored in a computer readable storage medium, wherein a processor of an electronic device reads the executable instructions from the computer readable storage medium and executes the executable instructions to implement the deep learning-based grouting pressure intelligent adjustment method described above.

[0021] The above one or more technical solutions have the following beneficial effects:

[0022] 1. The present application provides a deep learning-based grouting pressure intelligent adjustment method and system, which considers the active adjustment in the actual grouting process and the pressure pulse of the grouting equipment itself, analyzes the grout diffusion affected by the two pressure pulse conditions and the surrounding energy exchange, establishes a grouting pressure influence prediction model; at the same time, considering the limiting relationship between active adjustment and pressure pulse of the grouting equipment itself, ensuring the efficiency of grouting diffusion and backflow process, realizing the orifice pipe plugging in the grouting process, establishing a grouting pressure regulation model; at the same time, considering the connectivity relationship between different grouting holes, adjusting the active regulation parameters of the grouting pressure regulation model to reduce the loss of grout backflow and grouting material, and at the same time, according to the real-time feedback of the field data, updating the pressure threshold of the grouting pressure regulation model in real time; the present application realizes the automatic real-time and accurate adjustment of grouting pressure by constructing the grouting pressure influence prediction model and the grouting pressure regulation model, and according to the real-time feedback of the actual geological conditions and grouting data on site.

[0023] 2. The present application establishes a grouting database based on geological data, grouting parameters and pressure pulse information, trains the grouting pressure influence prediction model using the grouting database, considers the corresponding grout diffusion range, grout flow rate and grout backflow under each pressure pulse during the training process, sets the multivariate nonlinear variation of grout diffusion range and grout backflow under the conditions of surrounding environmental influence, active regulation of pressure pulse and multiple frequency superposition of grouting equipment itself pressure pulse, and at the same time, considers the influence of the active regulation pressure pulse on the periodic time of the grouting equipment itself pressure pulse, so as to ensure the accuracy of the grouting pressure influence prediction model constructed and trained.

[0024] 3、The application pre-processes the data before actual grouting construction, such as geological geophysical data, grouting parameter data, etc., considers the limiting relationship of active adjustment and grouting equipment pressure pulse, ensures the efficiency of grouting diffusion and backflow process, realizes the orifice pipe plugging in the grouting process, establishes a grouting pressure regulation and control model; at the same time, the connectivity relationship between different grouting holes is considered, the grouting pressure regulation and control model is adjusted to actively control the pressure pulse parameters, so as to reduce the slurry backflow and the loss of grouting materials, according to the real-time feedback of the field data, the pressure threshold of the grouting pressure regulation and control model is updated in real time, so as to obtain a real-time accurate grouting pressure regulation value closer to the actual situation.

[0025] Advantages of the additional aspects of the application will be partially given in the following description, partially will become obvious from the following description, or will be understood by the practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0026] The drawings accompanying the specification of this application form a part thereof, serve to provide further understanding of the application, and together with the description of the exemplary embodiments of the application, serve to explain the application, and do not constitute improper limitations on the application.

[0027] Figure 1 A flow chart of the grouting pressure intelligent adjustment method based on deep learning in the embodiment of the application is shown in the figure.

[0028] Figure 2 A cycle framework diagram of the single time step multi-task learning with weight update in the embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0029] It should be noted that the following detailed description is exemplary only, is only for describing specific embodiments, and is intended to provide further explanation of the application, and is not intended to limit the exemplary embodiments according to the application. Unless otherwise indicated, all technical and scientific terms used herein have the same meaning as generally understood by those skilled in the art to which the application belongs. In addition, it should be understood that when the terms "comprise" and / or "include" are used in the specification, they mean that the features, steps, operations, devices, components and / or their combinations exist.

[0030] Embodiment one

[0031] The embodiment provides a grouting pressure intelligent adjustment method based on deep learning, which specifically comprises the following steps:

[0032] Step S1, obtaining geological geophysical data of a grouting area before grouting construction;

[0033] Step S2, the surrounding environment, active control pressure pulse, grouting equipment itself pressure pulse parameters as a description of grouting diffusion state S, the parameters in state S are input into the pre-trained grouting pressure influence prediction model, and the corresponding grouting effect is obtained;

[0034] Step S3, according to the geological geophysical data, the required grouting diffusion threshold state is calculated, and the parameters in state S, the corresponding grouting effect and the grouting diffusion threshold state are input into the pre-trained grouting pressure control model, the model takes the grouting diffusion threshold state as the target of the grouting effect, and outputs the grouting pressure regulation operation, which includes the change value of each parameter in state S;

[0035] Step S4, according to the real-time feedback of the actual grouting parameters and the actual grouting parameters, the grouting pressure control model is updated, and the updated model is used for grouting pressure regulation.

[0036] The grouting pressure intelligent adjustment method based on deep learning proposed in the embodiment is described in more detail as follows.

[0037] Firstly, before grouting construction in the grouting area to be grouted, the grouting pressure influence prediction model and the grouting pressure control model are constructed and pre-trained, and the specific process is as shown in Figure 1 , including:

[0038] (1) Grouting pressure influence prediction model

[0039] (1.1) Obtain grouting information in the historical grouting construction process, including: geological geophysical data and grouting parameter data and pressure pulse information.

[0040] Specifically, through geological exploration and other ways, the geological geophysical data of the grouting area is obtained, that is, the geological characteristics of the area are collected, including geological minerals, geological elements, voids, and fracture related information, through data analysis, the distribution and connectivity relationship of voids and fractures space are obtained; at the same time, the corresponding grouting parameter data is obtained, that is, the flow rate, diffusion range, filling effect and other information of slurry under different medium conditions and grouting pressure, the pressure change curve in the grouting process is collected, and the fluctuation trend under different geological conditions and different slurry material concentration is recorded; in addition, the pressure pulse information generated by each manual adjustment during the grouting process (referred to as active adjustment pressure pulse) and the pressure pulse information generated by the grouting equipment itself (referred to as grouting equipment itself pressure pulse) are collected at the same time, which includes time, frequency, amplitude and duration.

[0041] Furthermore, based on the different degrees of influence of each parameter data on the final grouting, the above-mentioned parameters are divided into key parameters and non-key parameters. The key parameters include the active adjustment of pressure pulse, the pressure pulse of the grouting equipment itself, the flow rate of grout under different media conditions and grouting pressure, diffusion range, filling effect, etc.; the non-key parameters include grouting volume, geological minerals, geological elements, etc.

[0042] (1.2) The acquired data is preprocessed to construct the grouting database.

[0043] Specifically, firstly, an autoencoder learns the change patterns in the data to identify outliers and fill in missing values. Outliers in non-critical parameters can be directly removed, while outliers in critical parameters are marked to determine construction problems that may exist during the actual grouting process.

[0044] Furthermore, numerical data of different dimensions and units are standardized and normalized to ensure data quality and consistency.

[0045] After the above preprocessing, the surrounding environmental influences (i.e., the actual geological environment influences after grouting, reflected in real-time geological data, such as fracture spatial distribution and connectivity), actively controlled pressure pulses, and the pressure pulse parameters of the grouting equipment itself are used as the state S describing the grouting diffusion. The data of this state S and its corresponding grouting diffusion (including grout diffusion range, grout flow rate, and grout return range) data are used as training data to construct a grouting database, and the preprocessed data is stored in the database.

[0046] (1.3) Construct and train a prediction model for the influence of grouting pressure

[0047] First, a predictive model for the impact of grouting pressure is constructed. Specifically, such as... Figure 2 As shown, a multi-task learning model framework considering weights is adopted. Different branch networks are set up to correspond to the grout diffusion range, grout velocity, and grout return range of grouting. The input layer of each branch network is used as a shared layer. The influence of the surrounding environment under different medium conditions and grout material concentrations, the active control pressure pulse, and the pressure pulse of the grouting equipment itself are used as shared input values ​​E1 to E3 of the neural network shared layer into the model. Then, the shared layer data is mapped to the fully connected layers based on nonlinear activation functions in multiple branch networks. In this way, the relationship between various input data and the grout diffusion range, grout velocity, and grout return range (i.e., outputs 1 to 3) of each branch network is learned. In addition, in each branch network, the initial gain weights W corresponding to the parameters of the surrounding environment, the active control pressure pulse, and the pressure pulse of the grouting equipment itself in different branch networks are added. 1n ~W 3n, each weight is updated by the loss error in the training process.

[0048] Secondly, the data in the database is selected into a training set and a validation set. The data in the training set is used to train the above-mentioned grouting pressure influence prediction model. The training process is as follows: at each time step, the corresponding slurry diffusion range, slurry flow rate, and slurry backflow condition under each pressure pulse are considered. The multi-element nonlinear variation of the slurry diffusion range, slurry flow rate, and slurry backflow range under the conditions of the surrounding environmental influence, the active control pressure pulse, and the multi-frequency superposition of the pressure pulse of the grouting equipment itself is set. Meanwhile, the influence of the active control pressure pulse on the periodic time of the pressure pulse of the grouting equipment itself is considered. The training process of the above-mentioned grouting pressure influence prediction model can be expressed as:

[0049] F(y1,y2,y3)=W(ε(i,t),X i,T ,x(t,X i,T ))^b(T,t);

[0050] wherein y1, y2, and y3 represent the diffusion range of the grouting slurry, the slurry flow rate, and the slurry backflow condition respectively, ε(i,t) represents the energy exchange coefficient affected by the surrounding environment, X i,T represents the parameter values of the i-th frequency active control pressure pulse within T time variation, x(t,X i,T represents the parameter values of the pressure pulse of the grouting equipment itself affected by the i-th frequency active control pressure pulse X within t time, b(T,t) represents the gain value of each pressure pulse parameter value within t time, W=[W 1n ,W 2n ,W 3n ] represents the gain weight of the parameter values of the surrounding environmental influence, the active control pressure pulse, and the pressure pulse of the grouting equipment itself, and n represents the number of next-level parameters of each parameter value. Taking the pulse parameter as an example, it includes time, frequency, amplitude, and duration, etc.

[0051] In the training process, based on the initial values of the input initial surrounding environmental influence, active control pressure pulse, and pressure pulse of the grouting equipment itself, and the initial gain weight and gain value of the three parameters corresponding to different branch networks, the deviation between the predicted value and the true value of the model output is taken as the loss (i.e. losses 1-3) for continuous iterative training, and the gain weight and gain value on different branch networks are updated, and the bias ε(i,t) and b(T,t) coefficient values are continuously adjusted until the error is within the allowable range. The accuracy of the model is verified by the validation set, and the finally trained grouting pressure influence prediction model is obtained.

[0052] (2) Grouting pressure control model

[0053] (2.1) Constructing a grouting pressure control model and its training dataset. Specifically, the surrounding environment, actively controlled pressure pulses, and the pressure pulse parameters of the grouting equipment itself are used as the state S describing grout diffusion. Changes in the parameters of state S are taken as continuous grouting operations A. The grouting effect feedback under different grouting operations A can be obtained through the pre-trained grouting pressure influence prediction model. This grouting effect includes the grout diffusion range, grout flow rate, and grout backflow range. Furthermore, through the above method, the grouting effect corresponding to different states S under different grouting operations A can be obtained, thereby constructing a training dataset.

[0054] (2.2) Obtain geological and geophysical data of the area to be grouted before actual grouting construction, and preprocess the obtained data. Analyze the information of voids and fissures and their connectivity based on the preprocessed geological and geophysical data.

[0055] (2.3) Based on the recorded information such as void size, porosity, crack aperture, and extension length, calculate the required grouting diffusion threshold state S. i This state S i Including the grouting diffusion radius threshold R m and grouting diffusion time threshold T m The formulas for calculating each threshold are as follows:

[0056]

[0057] Among them, R m α1 and α2 are the grout diffusion radius thresholds, respectively; V1 and V2 are the corresponding weights for pores and fractures, respectively; φ is the porosity; h is the rock layer thickness; ω is the fracture aperture; L f This represents the length of the crack extension.

[0058]

[0059] Among them, T m denoted as the grouting diffusion time threshold, μ as the viscosity of the injected grout, and k as the permeability.

[0060] (2.4) Combine the preprocessed geological geophysical data with the required grouting diffusion threshold state S iAs initial data, high pressure and low pressure alternating pulses are set on active regulation, high pressure and low pressure threshold values are set according to the grouting pressure range required by construction, and the initial cycle change of high and low pressure is set, so that the pressure pulse of the grouting equipment itself has different cycle pressure changes under each active regulation pressure pulse; the grouting pressure regulation model based on the basic architecture of deep Q network is trained by using the initial data and the training data set, and the parameter values of the two pressure pulses are continuously adjusted, and the pressure pulse of the second half of the active regulation is appropriately adjusted to realize the promotion and hindering effect of the pressure pulse of the grouting equipment itself, by taking the active regulation as a limiter of the pressure pulse of the equipment itself, it is ensured that the grout backflow in the grouting process does not exceed the boundary of the grouting hole, and the hole pipe plugging in the grouting process is realized.

[0061] Specifically, the initial values of the parameters in state S are taken as the input of the model, and the geological geophysical data and the grouting diffusion threshold state S are input at the same time i , the grouting effect is taken as the output of the model, and the model is trained until the final output grouting effect reaches the required grouting diffusion threshold state S i . The training process of the grouting pressure regulation model can be represented as:

[0062] X=w X (m j ,T1),x=w x (n j ,T2,X),z=f(X,x);

[0063] Where X is the action value of the active regulation pressure pulse, x is the action value of the grouting equipment itself pressure pulse, w X and w x are the reward coefficients under the conditions of active regulation and grouting equipment itself pressure pulse, m j is the parameter value of the pressure pulse of the active regulation in the jth cycle, n j is the parameter value of the pressure pulse of the grouting equipment itself in the jth cycle of active regulation, T1 and T2 are the adjustable cycle parameters of the active regulation and the equipment itself pressure pulse respectively, f is the limiting relationship between the active regulation and the equipment itself pressure pulse, z is the output result of the model, and the output result is the grouting effect, including the diffusion range of the grouting slurry, the slurry speed and the slurry backflow.

[0064] Finally, according to the above initial data, the parameter values of the two pressure pulses are continuously adjusted to make the grouting effect obtained from the grouting pressure influence prediction model reach S iThe matching is a process of normalizing and subtracting the values of the slurry diffusion range, slurry speed and slurry backflow range of the two parts of grouting, and then performing weighted average on the differences in the three states. When the final weighted average value is within the set threshold range, it means that the matching is completed.

[0065] After the above model is built and trained, based on the pre-trained grouting pressure influence prediction model and grouting pressure regulation model, real-time pressure intelligent regulation of the grouting area is performed, including:

[0066] First, the geological geophysical data of the grouting area before grouting construction is obtained. Second, the surrounding environmental impact, active regulation pressure pulse, and grouting equipment pressure pulse parameters are used as the state S to describe the grouting diffusion. The parameters in state S are input into the pre-trained grouting pressure influence prediction model to obtain the corresponding grouting effect. Then, according to the geological geophysical data, the required grouting diffusion threshold state is calculated, and the parameters in state S, the corresponding grouting effect, and the grouting diffusion threshold state are input into the pre-trained grouting pressure regulation model. The model takes the grouting diffusion threshold state as the target of the grouting effect and automatically outputs the grouting pressure regulation operation, which includes the change value of each parameter in state S, mainly the change value of the active regulation pressure pulse and the grouting equipment pressure pulse, i.e. the grouting pressure regulation value. On this basis, the automatic adjustment of the grouting operation is performed according to the change value of the parameters, so as to realize the intelligent regulation of the grouting pressure.

[0067] Preferably, when the grouting pressure regulation value is output by the grouting pressure regulation model, the grouting pressure regulation model needs to be updated according to the actual situation of the region and the real-time feedback of the actual grouting parameters on site, and the updated model is used for grouting pressure regulation, so that the output grouting pressure regulation value is more accurate. Specifically, the following situations are included:

[0068] ① According to the geological geophysical data of the region, the connectivity between multiple holes is considered, i.e. the connectivity between multiple different grouting holes is considered. In the region with high connectivity, the frequency of active regulation in the grouting pressure regulation model is adjusted, and the amplitude of the active regulation is adjusted, so that the grouting slurry can penetrate and diffuse in a certain area at an appropriate speed, and can still achieve a high cumulative reward under high active regulation pressure pulse and meet the matching with the grouting diffusion threshold state S i

[0069] ② According to the grouting situation of each grouting hole, the diffusion situation of the slurry in the front grouting area is obtained, and the setting time and bonding strength of the slurry are combined to update and adjust the front grouting area, i.e. the grouting diffusion threshold state S i ​The updated numerical value is imported into the grouting pressure regulation model to perform real-time regulation of the grouting pressure.

[0070] ③For multi-hole combined grouting, the grout diffusion range and flow condition of multiple grouting holes are considered, that is, the grouting diffusion threshold state S1, S2,..., S n The different grout diffusion threshold states (mainly the grout diffusion radius) are determined in the spatial dimension, the flow speed of the grout is obtained in the time dimension, the frequency of active regulation of the grouting pressure regulation model is adjusted, the amplitude is adjusted, the time when the grout of each hole flows to the overlapping area of the grout diffusion radius of each hole is accurately obtained, and the loss of grout backflow and grouting material is reduced.

[0071] That is, under the complex geological structure, the overall diffusion under different grouting pressures is predicted under the initial threshold condition to achieve the goal of uniform filling, and then the pressure range is predicted according to the real-time feedback of the field data, the pressure threshold of the grouting pressure regulation model is updated, and the real-time regulation of the grouting pressure based on the actual situation is realized to ensure the accuracy of the final grouting.

[0072] In the above manner, real-time and accurate grouting pressure regulation closer to the actual situation can be realized to achieve a better grouting effect.

[0073] Embodiment Two

[0074] The embodiment provides a grouting pressure intelligent regulation system based on deep learning, which comprises:

[0075] A data acquisition module is configured to acquire geological geophysical data of a grouting area before grouting construction.

[0076] A grouting effect feedback module is configured to use the parameters of the surrounding environment influence, the active regulation pressure pulse and the grouting equipment pressure pulse as the state S describing the grouting diffusion, input the parameters in the state S into a pre-trained grouting pressure influence prediction model, and obtain the corresponding grouting effect.

[0077] A grouting pressure regulation module is configured to calculate the required grouting diffusion threshold state according to the geological geophysical data, and input the parameters in the state S, the corresponding grouting effect and the grouting diffusion threshold state into a pre-trained grouting pressure regulation model.

[0078] A grouting pressure regulation updating module is configured to update the grouting pressure regulation model according to the real-time feedback of the grouting actual situation and the actual grouting parameters, and regulate the grouting pressure by using the updated model.

[0079] Embodiment three

[0080] The embodiment provides an electronic device, comprising: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement the method provided by the embodiment.

[0081] Embodiment four

[0082] The embodiment also provides a computer readable storage medium storing executable instructions, which, when executed by a processor, cause the processor to perform the method provided by the embodiment.

[0083] Embodiment five

[0084] The embodiment provides a computer program product, which comprises executable instructions, the executable instructions being computer instructions; the executable instructions are stored in a computer readable storage medium. When the processor of the electronic device reads the executable instructions from the computer readable storage medium, the processor executes the executable instructions, so that the electronic device executes the method provided by the embodiment.

[0085] The steps and methods in the above embodiments two to five correspond to the method embodiment one, and the specific embodiments can refer to the related description part of the embodiment one. The term "computer readable storage medium" should be understood as including a single medium or multiple media of one or more instruction sets; it should also be understood as including any medium capable of storing, encoding or carrying instruction sets for execution by a processor and causing the processor to perform any method in the present application.

[0086] Those skilled in the art should understand that the above modules or steps of the present application can be realized by a general computer device, alternatively, they can be realized by program codes executable by a computing device, so that they can be stored in a storage device for execution by a computing device, or they can be respectively manufactured into individual integrated circuit modules, or a plurality of modules or steps among them can be manufactured into a single integrated circuit module to realize. The present application is not limited to any specific combination of hardware and software.

[0087] The above only describes the preferred embodiments of the present application, and the specific embodiments of the present application are described in conjunction with the drawings, but are not limited to the protection scope of the present application. Those skilled in the art should understand that various modifications or changes made on the basis of the technical solutions of the present application without creative labor are still within the protection scope of the present application.

Claims

1. A method for intelligent adjustment of grouting pressure based on deep learning, characterized in that, include: Obtain geological and geophysical data of the area to be grouted before grouting construction; The surrounding environment, actively controlled pressure pulse, and pressure pulse parameters of the grouting equipment itself are used to describe the state of grout diffusion. S , will state S Each parameter is input into a pre-trained grouting pressure influence prediction model to obtain the corresponding grouting effect; Based on geological geophysical data, calculate the required grouting diffusion threshold state, and then set the state... S Each parameter and its corresponding grouting effect, as well as the grouting diffusion threshold state, are input into a pre-trained grouting pressure control model. This model takes the grouting diffusion threshold state as the target of the grouting effect and outputs a grouting pressure adjustment operation, which includes state adjustment. S The changes in various parameters of the pressure pulse actively controlled by the grouting equipment itself; Based on the actual grouting situation and real-time feedback of actual grouting parameters, the grouting pressure control model is updated, and the updated model is used to adjust the grouting pressure.

2. The intelligent grouting pressure adjustment method based on deep learning as described in claim 1, characterized in that, The training process of the grouting pressure-affected prediction model includes: Acquire geological geophysical data, grouting parameter data, and pressure pulse information; The acquired data is preprocessed to construct a grouting database; A grouting pressure influence prediction model based on a weighted multi-task learning architecture is constructed, including: setting up different branch networks corresponding to the diffusion range, velocity, and return range of the grout; using the input layer of each branch network as a shared layer; inputting the surrounding environmental influences under different media conditions and grout material concentrations, the actively controlled pressure pulse, and the pressure pulse of the grouting equipment itself as shared input values ​​into the model; and mapping the shared layer data to fully connected layers based on nonlinear activation functions in multiple branch networks to learn the relationship between various input data and the final output of each branch network for the grout diffusion range, velocity, and return range; and adding initial gain weights for the parameters of the surrounding environmental influences, the actively controlled pressure pulse, and the pressure pulse of the grouting equipment itself in different branch networks in each branch network. The model was trained using the training dataset in the grouting database. Through continuous iterative training and adjustments, a trained model for predicting the impact of grouting pressure was obtained.

3. The intelligent grouting pressure adjustment method based on deep learning as described in claim 2, characterized in that, The training process for the prediction model of the influence of grouting pressure is as follows: In the calculation of each time step, the slurry diffusion range, slurry velocity, and slurry return range corresponding to each pressure pulse are considered. The multivariate nonlinear changes of slurry diffusion range, slurry velocity, and slurry return range are set under the conditions of multiple superposition of surrounding environmental influences, actively controlled pressure pulses, and pressure pulses of the grouting equipment itself. At the same time, the influence of the pressure pulse of the grouting equipment itself on the period time of the actively controlled pressure pulse is also considered. The training process of the prediction model for the influence of grouting pressure is represented as follows: ； in, , , These represent the diffusion range, grout velocity, and grout backflow of the grout, respectively. This represents the energy exchange coefficient influenced by the surrounding environment. express T Within the time change i The parameter values ​​of each frequency of actively controlled pressure pulse. express t Within a time period, affected by the first i Each frequency actively controlled pressure pulse The impact on the pressure pulse parameters of the grouting equipment itself. Indicates in t The gain values ​​of each pressure pulse parameter within a given time period. This indicates the gain weights of the surrounding environmental influences, actively controlled pressure pulses, and the pressure pulse parameter values ​​of the grouting equipment itself. n This indicates the number of next-level parameters for each parameter value.

4. The intelligent grouting pressure adjustment method based on deep learning as described in claim 1, characterized in that, The training process of the grouting pressure control model is as follows: The surrounding environment, actively controlled pressure pulse, and pressure pulse parameters of the grouting equipment itself are used to describe the state of grout diffusion. S , will state S The changes in various parameters are used as a basis for continuous grouting operations. A By using a pre-trained prediction model of grouting pressure influence, the grouting effect under different grouting operations can be obtained; based on different states S Different grouting operations A The grouting effect is analyzed to construct a training dataset; the grouting effect includes grout diffusion range, grout flow rate, and grout backflow range. Obtain geological geophysical data of the area to be grouted before actual grouting construction, and calculate the required grout diffusion threshold state based on the geological geophysical data. ; Based on geological geophysical data and grouting diffusion threshold status As initial data, alternating high and low pressure pulses were set for active regulation. High and low pressure thresholds were set according to the grouting pressure range required for construction, and initial periodic changes of high and low pressure were established. Using the initial data and training dataset, a grouting pressure regulation model based on a deep Q-network architecture was trained. The parameter values ​​of the two pressure pulses were continuously adjusted to ensure that the grouting effect obtained from the grouting pressure influence prediction model was consistent with the grouting diffusion threshold state. The matching was completed to train the grouting pressure control model.

5. The intelligent grouting pressure adjustment method based on deep learning as described in claim 4, characterized in that, The grouting diffusion threshold state Including grouting diffusion radius threshold and grouting diffusion time threshold The formula for calculating the grouting diffusion radius threshold is as follows: ； In the above formula, This is the threshold value for the grouting diffusion radius. and These are the weights corresponding to voids and fissures, respectively. and These represent the volumes of grout available for injection into the voids and fissures, respectively. Porosity For the thickness of the rock strata, For crack aperture, The length of the crack extension; The formula for calculating the grouting diffusion time threshold is: ； In the above formula, The grouting diffusion time threshold, The viscosity of the injected grout, This refers to penetration rate.

6. The intelligent grouting pressure adjustment method based on deep learning as described in claim 4, characterized in that, The training process of the grouting pressure control model is represented as follows: ； in, To actively adjust the action value of pressure pulses, The value of the pressure pulse action of the grouting equipment itself. and These are the reward coefficients for active adjustment and the pressure pulse conditions of the grouting equipment itself, respectively. For the first j The parameters of the actively adjusted pressure pulse within each cycle. To actively adjust the first j The pressure pulse parameters of the grouting equipment itself within each cycle. and These are the adjustable period parameters for the pressure pulses, which are actively adjusted and those for the equipment itself. In order to actively adjust the limiting relationship with the pressure pulse of the equipment itself, This is the output of the model, i.e., the grouting effect.

7. A deep learning-based intelligent grouting pressure regulation system, characterized in that, include: The data acquisition module is used to acquire geological and geophysical data of the area to be grouted before grouting construction. The grouting effect feedback module uses parameters such as the influence of the surrounding environment, actively controlled pressure pulses, and the pressure pulses of the grouting equipment itself to describe the state of grout diffusion. S , will state S Each parameter is input into a pre-trained grouting pressure influence prediction model to obtain the corresponding grouting effect; The grouting pressure control module is used to calculate the required grouting diffusion threshold state based on geological geophysical data, and to set the state accordingly. S Each parameter and its corresponding grouting effect, as well as the grouting diffusion threshold state, are input into a pre-trained grouting pressure control model. This model takes the grouting diffusion threshold state as the target of the grouting effect and outputs a grouting pressure adjustment operation, which includes state adjustment. S The changes in various parameters of the pressure pulse actively controlled by the grouting equipment itself; The grouting pressure control and update module is used to update the grouting pressure control model based on the actual grouting situation and real-time feedback of actual grouting parameters, and to adjust the grouting pressure using the updated model.

8. An electronic device, characterized in that, include: Memory, used to store executable instructions; The processor, when executing executable instructions stored in the memory, implements the deep learning-based intelligent grouting pressure adjustment method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The device stores executable instructions that, when executed by a processor, implement the deep learning-based intelligent grouting pressure adjustment method according to any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product includes executable instructions stored in a computer-readable storage medium; When the processor of the electronic device reads the executable instructions from the computer-readable storage medium and executes the executable instructions, it implements the deep learning-based intelligent adjustment method for grouting pressure as described in any one of claims 1-6.

Images