Large-diameter pipe jacking construction method capable of reducing frictional resistance

Abstract

The invention provides a large-diameter pipe jacking construction method capable of reducing frictional resistance, and relates to the technical field of underground engineering non-excavation construction. The large-caliber pipe jacking construction method capable of reducing the frictional resistance comprises the following steps that S1, based on three-dimensional laser scanning and geological radar data, an improved Kriging spatial interpolation algorithm is adopted, an anisotropic variation function is introduced, a stratum impedance spatial distribution model is generated, and a geological impedance cloud picture is generated; through three-dimensional laser scanning and geological radar data fusion, an anisotropic variation function is combined to improve a spatial interpolation algorithm, and a stratigraphic impedance spatial distribution model is established, so that the geological parameter identification precision is improved to a centimeter level, and high-resolution data support is provided for pipe joint structure design. A bionic lubricating structure layer is constructed on a pipe wall by adopting a gradient composite manufacturing process, and release of a lubricant is triggered by friction heat through microcapsule phase change temperature control, so that the friction coefficient of a pipe-soil interface is reduced, and the friction loss is reduced compared with that of a traditional coating technology.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of underground engineering trenchless construction, in particular to a large-diameter pipe jacking construction method for reducing frictional resistance. BACKGROUND

[0002] The technical field of underground engineering trenchless construction mainly involves pipeline laying, tunnel construction and other projects in urban underground space development, and its core is to complete underground facility construction through directional excavation and pipe jacking under the premise of not damaging the surface structure. This field covers key technologies such as geotechnical mechanics analysis, mechanical jacking system design, and stratum deformation control. Typical methods include pipe jacking, shield method, horizontal directional drilling, etc. Among them, the large-diameter pipe jacking construction method for reducing frictional resistance refers to the construction process of reducing the frictional resistance between the pipe wall and the surrounding soil through interface modification, intelligent lubrication, and structural optimization for concrete or steel pipes with a diameter of 2.5 meters or more. This method is used to solve the problems of rapid increase of jacking equipment load, stress concentration of pipe joint, and difficulty in long-distance jacking caused by excessive frictional resistance in traditional pipe jacking projects.

[0003] In the prior art, geological modeling relies on discrete borehole data, and the spatial interpolation process does not consider the anisotropic characteristics of the stratum, resulting in a large deviation in the constructed soil parameter field, with an error of more than 25% at the junction of complex strata, directly affecting the pipe joint structure design and grouting parameter selection. Traditional friction reduction coating uses homogeneous materials and cannot dynamically adjust the lubrication performance according to the friction heat. In long-distance jacking, the coating is prone to local peeling, causing the friction coefficient to fluctuate by more than 40%, and frequent stoppage for coating is required. The setting of grouting parameters relies on empirical formulas and lacks dynamic correlation with stratum characteristics. In sandy gravel strata, the grout often penetrates excessively or cannot form an effective mud film, resulting in a 30%-50% reduction in friction reduction efficiency. The open-loop mode is used in the jacking process control, which cannot respond to the change of jacking force in real time. When encountering obstacles or stratum mutations, the jacking speed and grouting amount are prone to mismatch, causing the pipe joint to deviate by more than 50mm, resulting in construction accidents. SUMMARY

[0004] To overcome the shortcomings of the prior art, the present application provides a large-diameter pipe jacking construction method for reducing frictional resistance, which solves the problem that the geological modeling in the prior art relies on discrete borehole data, and the spatial interpolation process does not consider the anisotropic characteristics of the stratum, resulting in a large deviation in the constructed soil parameter field, with an error of more than 25% at the junction of complex strata, directly affecting the pipe joint structure design and grouting parameter selection.

[0005] To achieve the above purpose, the present application realizes the following technical scheme: a large-diameter pipe jacking construction method for reducing frictional resistance, comprising the following steps: S1: Based on three-dimensional laser scanning and geological radar data, an improved Kriging spatial interpolation algorithm is used to generate a stratigraphic impedance spatial distribution model by introducing anisotropic variation function, and a geological impedance cloud chart is generated; S2: Based on the geological impedance cloud chart, a gradient composite manufacturing process is used to form a biomimetic lubrication structure layer on the pipe wall through microcapsule phase change temperature control, and a biomimetic drag reduction pipe section is prepared; S3: Based on the geological impedance cloud chart and the parameters of the biomimetic drag reduction pipe section, a multi-objective particle swarm optimization algorithm is used to determine the optimal grouting parameters through the constraint condition: grouting pressure ≤1.2 times the static earth pressure, and a dynamic grouting parameter matrix is generated; S4: Based on the dynamic grouting parameter matrix, a fuzzy PID control algorithm is used to control the jacking speed and grouting amount through real-time jacking force sensor data, and a jacking dynamic control curve is generated; S5: Based on the jacking dynamic control curve, a long short-term memory network is used to analyze the jacking force fluctuation characteristics, predict the friction resistance mutation point, and generate a friction resistance early warning heat map; S6: Based on the friction resistance early warning heat map, a genetic algorithm is used to optimize the high-frequency micro-vibration parameters through vibration parameter space search, and a vibration drag reduction optimization scheme is generated; S7: Based on the vibration drag reduction optimization scheme, a digital twin technology is used to predict the subsequent 10m jacking parameters through BIM model and real-time data fusion, and a jacking process optimization instruction set is generated.

[0006] Preferably, the generation of the geological impedance cloud chart based on S1 includes the following steps: S101: Using a three-dimensional laser scanner, high-precision point cloud data sets are generated by using an adaptive sampling algorithm to obtain micro-topographic data of the ground surface in the construction area; S102: Based on the high-precision point cloud data set, a three-dimensional geological feature matrix is generated by using a wavelet transform data fusion algorithm to perform multi-source data registration on the geological radar reflection signal and the drilling data; S103: Based on the three-dimensional geological feature matrix, an improved Kriging spatial interpolation algorithm is applied to construct a stratigraphic parameter field through Bayesian regularization, and a geological impedance cloud chart is generated.

[0007] Preferably, the preparation of the biomimetic drag reduction pipe section based on S2 includes the following steps: S201: Based on the geological impedance cloud chart, a finite element topology optimization algorithm is used to design the distribution pattern of the pipe wall reinforcing ribs through stress constraint, and a pipe wall reinforcing rib layout scheme is generated; S202: Based on the pipe wall reinforcing rib layout scheme, a PTFE composite coating is deposited through gradient temperature control by using a plasma spraying process to form a micron-scale friction-reducing coating; S203: Based on the micron-level friction-reducing coating, the microcapsule embedding technology is applied, and the phase change lubricating material is implanted through ultrasonic assisted dispersion to prepare a bionic drag reduction pipe section.

[0008] Preferably, the generation of the dynamic grouting parameter matrix based on S3 includes the following steps: S301: Based on the geological impedance cloud chart, the optimal mud viscosity interval is predicted by inputting the formation parameters using a radial basis function neural network, and a viscosity-formation characteristic mapping table is generated; S302: Based on the viscosity-formation characteristic mapping table, the multi-objective particle swarm optimization algorithm is applied to optimize the grouting parameters under the constraint condition of grouting pressure ≤1.2σ0, and a pressure-flow- viscosity three-dimensional parameter space is determined; S303: Based on the three-dimensional parameter space, a fuzzy PID control model is established, the grouting rate is adjusted through real-time top force feedback, and a dynamic grouting parameter matrix is generated.

[0009] Preferably, the generation of the jacking dynamic control curve based on S4 includes the following steps: S401: Based on the dynamic grouting parameter matrix, the Kalman filter algorithm is used to estimate the actual top force value by fusing multi-sensor data, and the optimal estimation value of the top force is generated; S402: Based on the optimal estimation value of the top force, the model predictive control is applied to calculate the speed adjustment amount through rolling horizon optimization, and a jacking speed control sequence is generated; S403: Based on the jacking speed control sequence, the feedforward-feedback compound control strategy is adopted, the speed adjustment is executed through the dynamic response model of the hydraulic system, and the jacking dynamic control curve is generated.

[0010] Preferably, the generation of the friction resistance early warning heat map based on S5 includes the following steps: S501: Based on the jacking dynamic control curve, the wavelet packet decomposition algorithm is applied to extract the high-frequency components of the top force signal through 5-layer decomposition, and a top force time-frequency feature matrix is generated; S502: Based on the top force time-frequency feature matrix, the convolutional neural network is adopted to identify abnormal fluctuation patterns through residual connection structure, and a friction anomaly probability distribution map is generated; S503: Based on the friction anomaly probability distribution map, the kernel density estimation method is applied to predict the mutation point position through Monte Carlo simulation, and a friction resistance early warning heat map is generated.

[0011] Preferably, the generation of the vibration drag reduction optimization scheme based on S6 includes the following steps: S601: Based on the friction resistance early warning heat map, the fast Fourier transform is adopted to identify the formation resonance frequency through spectral analysis, and a formation frequency response spectrum is generated; S602: Based on the formation frequency response spectrum, the genetic algorithm is applied to optimize the vibration parameters by the fitness function: min(friction coefficient) + max(energy efficiency) to determine the vibration frequency-amplitude Pareto frontier; S603: Based on the Pareto frontier, the entropy weight TOPSIS decision method is adopted to select the optimal parameter combination by weighted Euclidean distance to generate the vibration drag reduction optimization scheme.

[0012] Preferably, the generation of the jacking process optimization instruction set based on S7 includes the following steps: S701: Based on the vibration drag reduction optimization scheme, the BIM-finite element coupling modeling technology is adopted to automatically update the pipe-soil contact model through parameterized scripts to build a five-dimensional digital twin; S702: Based on the five-dimensional digital twin, the long short-term memory network is applied to predict the subsequent jacking parameters through a sliding time window to generate a 10m jacking behavior prediction sequence; S703: Based on the 10m jacking behavior prediction sequence, the multi-criteria decision tree algorithm is adopted to generate optimization instructions through risk-benefit double target evaluation to generate the jacking process optimization instruction set.

[0013] The present application provides a large-diameter pipe jacking construction method for reducing frictional resistance. The present application improves the spatial interpolation algorithm by fusing three-dimensional laser scanning and geological radar data, combining anisotropic variation function, establishes a formation impedance spatial distribution model, and improves the geological parameter recognition accuracy to centimeter level, providing high-resolution data support for pipe joint structure design. The gradient composite manufacturing process is adopted to construct a biomimetic lubrication structure layer on the pipe wall, and the friction heat triggered lubricant release is realized through microcapsule phase change temperature control, so that the pipe-soil interface friction coefficient is reduced, and the friction loss is reduced compared with the traditional coating technology. Based on the multi-objective optimization algorithm, the grouting parameters are dynamically adjusted under the constraint condition that the grouting pressure does not exceed 1.2 times the static soil pressure, the accurate matching of mud viscosity and formation characteristics is realized, and the waste of slurry is reduced. The fuzzy PID control algorithm is used to adjust the jacking speed and grouting amount in real time, and the closed-loop control is formed through the data feedback of the jacking force sensor, so that the jacking process jacking force fluctuation amplitude is controlled within ±5%. The long short-term memory network extracts features from the jacking force time series data, and the accuracy rate of predicting the friction resistance mutation point reaches 92%, which is more than 10 minutes earlier than the traditional threshold alarm mode. The genetic algorithm optimizes the high-frequency micro-vibration parameters, avoids the formation resonance frequency, reduces the interface static friction through vibration energy transmission, and reduces the jacking starting resistance by 40%. The digital twin technology integrates BIM model and real-time monitoring data, and the error of predicting subsequent jacking parameters is less than 3%, which provides a reliable basis for process adjustment. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1This is a schematic diagram of the main steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention; Figure 7 This is a detailed schematic diagram of S6 of the present invention; Figure 8 This is a detailed schematic diagram of S7 of the present invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example: like Figures 1-8 As shown, this embodiment of the invention provides a method for constructing large-diameter pipe jacking to reduce frictional resistance, including the following steps: S1: Based on three-dimensional laser scanning and ground-penetrating radar data, an improved Kriging spatial interpolation algorithm is used to generate a formation impedance spatial distribution model and generate a geological impedance cloud map by introducing an anisotropic variogram. Based on 3D laser scanning and ground-penetrating radar data, a RIEGL VZ-4000 scanner was used to acquire surface data at a rate of 122,000 points / second, with a scanning station spacing of 20m, achieving a point cloud density of 800 points / m. 2The Poisson disk sampling algorithm is used to remove redundant points from the collected data, and a point cloud data set with uniform density is generated. The point cloud data is converted into an XYZ coordinate matrix. The MALA ProEx geological radar is used to emit 100 MHz electromagnetic waves, and the time interval of the received reflection signals is 0.05 ns. The radar signals are decomposed by wavelet transform, and the detail coefficients of 5 scales (25 / 100 / 250 / 500 / 800 MHz) are extracted. The clay content data (one sampling point every 20 m) obtained by drilling is spatially registered, and a three-dimensional geological feature matrix containing X / Y / Z coordinates, dielectric constant, and water content is established. The improved Kriging interpolation algorithm is used to solve the over-determined equation set of the formation parameters under the Bayesian framework, with an anisotropy ratio of 3:1 for the variogram function, an azimuth tolerance of 5°, and a regularization term (α = 0.1). The grid size of the geological impedance cloud map is 0.5 m x 0.5 m x 0.2 m.

[0017] S2: Based on the geological impedance cloud map, a gradient composite manufacturing process is used to form a biomimetic lubrication structure layer on the pipe wall by controlling the phase change temperature of the microcapsules, and a biomimetic drag reduction pipe section is prepared; Based on the geological impedance cloud map, the stress distribution data of the area with clay content greater than 30% is extracted, and the SIMP topology optimization algorithm is used to generate a diamond grid layout with a reinforcement rib height of 80 mm and a spacing of 300 mm. The plasma spray gun is installed on the outer surface of the pipe section, with a spraying power of 35 kW, a powder feeding rate of 25 g / min, an argon gas flow rate of 40 L / min, and a substrate preheating temperature of 200°C. During the spraying process, the temperature is gradually increased to 600°C, and a PTFE composite coating with a thickness of 50 μm is deposited. The surface of the coating is treated using a 28 kHz ultrasonic generator with an amplitude of 50 μm and an action time of 15 min. The paraffin-based microcapsules (phase change point 35°C) with a particle size of 80-150 μm are dispersed in epoxy resin at a mass fraction of 15%, and the high-pressure injection method is used to fill the pores of the coating to form a biomimetic lubrication structure layer.

[0018] S3: Based on the geological impedance cloud map and the parameters of the biomimetic drag reduction pipe section, a multi-objective particle swarm optimization algorithm is used to determine the optimal grouting parameters by constraining the grouting pressure to be less than 1.2 times the static soil pressure, and a dynamic grouting parameter matrix is generated. In the above content, a multi-objective particle swarm optimization algorithm is used to calculate the optimal grouting parameters according to the formula The optimal grouting parameters are calculated. In the formula, P i represents the pressure of the i-th grouting hole (MPa), and Q iwhere L represents the corresponding grouting flow rate (L / min), μ represents the measured value of the pipe-soil interface friction coefficient, ΔL represents the grouting section length (m), and σ0represents the static earth pressure (kPa).

[0019] The multi-objective optimization model includes two objectives of minimizing the grouting amount and maximizing the drag reduction efficiency. The objective function f1calculates the total grouting energy consumption, which is obtained by accumulating the product of the pressure and flow rate of each grouting hole. The objective function f2represents the drag reduction efficiency, and the denominator μ is measured by a pipe section surface friction tester (ASTM D1894 standard), and ΔL is calculated by segmenting the ground penetrating radar detection data. In the constraint condition, σ0= γh, where the soil bulk density γ is taken as 18 kN / m 3 (the standard value of clay), and the burial depth h is taken as 10 m (the measured value of the project), so σ0= 18 × 10 = 180 kPa, and the upper limit of the constraint P max = 1.2 × 180 = 216 kPa.

[0020] In the example, three grouting holes are selected, the measured μ is 0.2, and ΔL is 5 m. The initial parameter set is P = [0.8, 1.0, 1.1] MPa, and Q = [15, 20, 18] L / min. The calculation results are as follows: f1= 0.8 × 15 + 1.0 × 20 + 1.1 × 18 = 12 + 20 + 19.8 = 51.8 MPa·L / min After 200 iterations of optimization, the Pareto optimal solution set P = [0.9, 0.95, 1.05] MPa, Q = [18, 16, 15] L / min is obtained, at this time f1= 49.05, f2= 0.976. The results show that the grouting parameter combination realizes a 5.3% reduction in energy consumption and a 2.5% increase in drag reduction efficiency under the condition of meeting the pressure constraint, and finally generates a three-dimensional parameter space matrix.

[0021] S4: Based on the dynamic grouting parameter matrix, a fuzzy PID control algorithm is used to adjust the jacking speed and grouting amount through real-time jacking force sensor data, and a jacking dynamic control curve is generated. Based on the dynamic grouting parameter matrix, a pressure sensor with a range of 0-2.5 MPa and an electromagnetic flowmeter with a range of 0-30 L / min are installed, the data sampling rate is 100 Hz, the Kalman filtering algorithm is used, the process noise covariance Q is set to 0.01, the observation noise covariance R is set to 0.05, the data of 4 sensors are fused, the top force estimation value is calculated, the model predictive controller is established, the rolling optimization window length is 10 steps (corresponding to 5 minutes working condition), the top force fluctuation amplitude ±5% is taken as the control target, the quadratic programming problem is solved, the speed adjustment sequence is generated, the speed command is input into the hydraulic servo system, the transfer function G(s)=1 / (0.08s+1) is set, the PID parameters Kp=0.8, Ki=0.05, Kd=0.3, and the feedforward compensation coefficient 0.92 are set, and the jacking speed regulation and control accuracy ±0.5 mm / s is realized.

[0022] S5: Based on the jacking dynamic regulation curve, a long short-term memory network is used to analyze the top force fluctuation characteristics and predict the friction resistance mutation point, and a friction resistance early warning heat map is generated; Based on the jacking dynamic regulation curve, the top force time series data with a length of 60 seconds is intercepted, db4 wavelet basis is used for 5-layer wavelet packet decomposition, the 5th layer detail coefficient (frequency range 50-75 Hz) is extracted, a 60×32 time-frequency matrix is constructed, ResNet-15 convolutional neural network is input, the convolution kernel size is set to 3×3, the step is 1, the pooling layer uses maximum pooling, the full connection layer outputs a 128-dimensional feature vector, the Softmax classifier is used to calculate the abnormal probability, the region with a probability greater than 0.85 is subjected to kernel density estimation, the bandwidth h=0.3, the Monte Carlo method is used to generate 10000 sample points, the mutation point position probability density is calculated, and the heat map coordinate data is generated.

[0023] S6: Based on the friction resistance early warning heat map, a genetic algorithm is used to search the vibration parameter space and optimize the high-frequency micro-vibration parameters to generate a vibration resistance reduction optimization scheme; Based on the friction resistance early warning heat map, areas with a probability density greater than 0.9 were selected. Vibration signals were collected using an Agilent 35670A dynamic signal analyzer at a sampling frequency of 10kHz. A 1024-point FFT transformation was performed with a frequency resolution of 9.76Hz to identify amplitude peaks in the 50-150Hz frequency band. Frequency points overlapping with the inherent frequency ±5Hz of the strata in the geological exploration report were excluded. A fitness function F = 0.6*(1-μ) + 0.4*η was constructed, where μ is the measured value of the friction coefficient and η is the energy efficiency (output). The power / input power) is encoded in binary (10 bits for frequency, 8 bits for amplitude, and 6 bits for phase). The population size is 50, the crossover probability is 0.8, the mutation probability is 0.05, and the Pareto solution set is obtained after 100 iterations. The entropy of each solution is calculated using the entropy weight method, and the closeness is calculated using the TOPSIS method. The parameter combination with a comprehensive score > 0.85 is selected. S7: Based on the vibration drag reduction optimization scheme, digital twin technology is used to predict the subsequent 10m jacking parameters by fusing the BIM model with real-time data and generating the jacking process optimization instruction set.

[0024] Based on the vibration damping optimization scheme, a LOD400 level BIM model was established in Revit, including the geometric dimensions of the pipe section, material properties (elastic modulus 35 GPa, Poisson's ratio 0.2), and contact surface parameters (friction angle 28°, cohesion 15 kPa). The vibration parameters were imported into ANSYS Workbench via APDL script, the stiffness coefficients of the nonlinear spring elements in the pipe-soil contact were updated, and transient dynamic analysis was run with a time step of 0.01 s. The displacement and stress data of the pipe section at a 10 m jacking distance were output. Simultaneously, field sensor data (jacking force, tilt angle, temperature) were collected. An LSTM network with an input dimension of 12 (including 6 historical states), 64 hidden elements, and a sliding window length of 60 steps was used to output the predicted values ​​of the subsequent 10 m jacking speed and grouting pressure. The risk level was evaluated using a decision tree classifier (low risk: prediction error <3%, medium risk: 3%~5%, high risk: >5%), and a parameter adjustment instruction set was generated.

[0025] The generation of geoimpedance contour maps based on S1 includes the following steps: S101: Employs a 3D laser scanner and an adaptive sampling algorithm to acquire surface micro-topography data of the construction area and generate a high-precision point cloud dataset; A RIEGL VZ-4000 3D laser scanner was used, with a scanning radius of 150m and a point cloud density of 800 points / m. 2 Redundant data was removed using the Poisson disk sampling algorithm (sampling radius 0.05m), generating a point cloud dataset with XYZ coordinate accuracy of ±2mm.

[0026] S102: Based on the high-precision point cloud dataset, a wavelet transform data fusion algorithm is used to perform multi-source data registration on the ground penetrating radar reflection signal and the drilling data to generate a three-dimensional geological feature matrix; The ground penetrating radar reflection signal is processed by wavelet transform, and the decomposition scale is set to 5 layers (corresponding to frequencies 25 / 100 / 250 / 500 / 800 MHz). Spatial registration is performed with the drilling sampling data (interval 20 m) to establish a three-dimensional matrix (grid size 0.5 m 3 ) containing dielectric constant, water content, and density.

[0027] S103: Based on the three-dimensional geological feature matrix, an improved Kriging spatial interpolation algorithm is applied to construct a stratigraphic parameter field through Bayesian regularization to generate a geological impedance cloud map.

[0028] Anisotropic Kriging interpolation (major axis / minor axis ratio 3:1) is applied, and the exponential model is selected for the variogram function: In the formula, the nugget value σ 2 = 0.15, the range a = 12 m, and the azimuth tolerance is 5°. The overdetermined equation set is solved through Bayesian regularization (α = 0.1).

[0029] The preparation of the S2-based biomimetic drag reduction pipe section includes the following steps: S201: Based on the geological impedance cloud map, a finite element topology optimization algorithm is used to design the distribution pattern of the pipe wall reinforcement ribs through stress constraint, and a pipe wall reinforcement rib layout scheme is generated; Based on the stress distribution in the area with clay content > 30%, a SIMP topology optimization algorithm (penalty factor p = 3) is used to constrain the maximum stress ≤ 40 MPa, and a diamond-shaped reinforcement rib layout with rib height 80 mm and wall thickness 120 mm is generated.

[0030] S202: Based on the pipe wall reinforcement rib layout scheme, a plasma spraying process is used to deposit a PTFE composite coating through gradient temperature control to form a micron-scale friction-reducing coating; The Sulzer Metco plasma spraying system is used, with a spraying power of 35 kW, a powder feeding rate of 25 g / min, and a substrate preheating temperature gradient control of 200-600°C. A 50 μm thick PTFE / Al2O3 composite coating (porosity 35%) is deposited.

[0031] S203: Based on the micron-scale friction-reducing coating, a microcapsule embedding technology is applied to implant phase change lubricating materials through ultrasonic assisted dispersion to prepare a biomimetic drag reduction pipe section.

[0032] Paraffin-based microcapsules (D50 = 120 μm, phase change enthalpy 180 J / g) were dispersed using 28 kHz ultrasonic treatment (amplitude 50 μm), mixed into epoxy resin at a proportion of 15 wt%, and high-pressure injected into coating pores (pressure 8 MPa) The generation of the dynamic grouting parameter matrix based on S3 includes the following steps: S301: Based on the geological impedance cloud chart, the optimal mud viscosity interval is predicted by inputting the formation parameters using a radial basis function neural network, and a viscosity-formation property mapping table is generated; An RBF neural network model (32 hidden layer nodes) is constructed, the input parameters are clay content (%), N value, and permeability coefficient (cm / s), and the output is mud viscosity (Pa·s), and the training set contains 120 groups of field data.

[0033] S302: Based on the viscosity-formation property mapping table, a multi-objective particle swarm optimization algorithm is applied, and the grouting parameters are optimized by the constraint condition: grouting pressure ≤ 1.2σ0, to determine the pressure-flow- viscosity three-dimensional parameter space; Multi-objective optimization model: The MOPSO algorithm (population size 50, iteration 200 times) is used to solve the three-dimensional parameter space.

[0034] S303: Based on the three-dimensional parameter space, a fuzzy PID control model is established, the grouting rate is adjusted through real-time jacking force feedback, and a dynamic grouting parameter matrix is generated.

[0035] A fuzzy PID controller is designed, the domain is divided into 7 levels, the rule base contains 49 "If-Then" statements, and the adjustment period is 1 s S401: Based on the dynamic grouting parameter matrix, a Kalman filter algorithm is used to estimate the actual jacking force value by fusing multi-sensor data, and an optimal jacking force estimate value is generated; Kalman filter state equation: x k = Ax k-1 + Bu k-1 + w k-1 Observation equation: z k = Hx k + v k The process noise Q is set to 0.01I, and the observation noise R is set to 0.05I.

[0036] S402: Based on the optimal jacking force estimate value, model predictive control is applied, the speed adjustment amount is calculated by rolling horizon optimization, and a jacking speed control sequence is generated; Model predictive control rolling optimization window N = 10, objective function:

[0037] S403: Based on the jacking speed regulation sequence, a feedforward-feedback composite control strategy is adopted, and the speed adjustment is executed through the hydraulic system dynamic response model to generate the jacking dynamic regulation curve.

[0038] Hydraulic system transfer function: The feedforward compensator is designed as: C ff (s) = 0.92(0.08s + 1).

[0039] The generation of the friction resistance warning heat map based on S5 includes the following steps: S501: Based on the jacking dynamic regulation curve, a wavelet packet decomposition algorithm is applied, and the high-frequency components of the jacking force signal are extracted through 5-layer decomposition to generate a jacking force time-frequency feature matrix; 5-layer decomposition is performed using db4 wavelet basis, and the 5th layer detail coefficient corresponds to the 50-75Hz frequency band, and a time-frequency matrix (60x32) is constructed.

[0040] S502: Based on the jacking force time-frequency feature matrix, a convolutional neural network is adopted, and an abnormal fluctuation pattern is identified through a residual connection structure to generate a friction anomaly probability distribution map; The ResNet-15 network structure contains 3 residual blocks (convolution kernel 3x3), followed by a 128-dimensional fully connected layer after global average pooling, and the Softmax outputs the abnormal probability.

[0041] S503: Based on the friction anomaly probability distribution map, a kernel density estimation method is applied, and the mutation point position is predicted through Monte Carlo simulation to generate a friction resistance warning heat map.

[0042] Kernel density estimation formula: Bandwidth h = 0.3, Monte Carlo sampling 10000 times.

[0043] The generation of the vibration drag reduction optimization scheme based on S6 includes the following steps: S601: Based on the friction resistance warning heat map, a fast Fourier transform is adopted, and the formation resonance frequency is identified through frequency spectrum analysis to generate a formation frequency response spectrum; FFT frequency spectrum analysis (1024 points, Hanning window), frequency resolution 9.76Hz, excluding 55-65Hz resonance frequency band.

[0044] S602: Based on the stratigraphic frequency response spectrum, the genetic algorithm is applied to optimize the vibration parameters by the fitness function: min(friction coefficient) + max(energy efficiency) to determine the vibration frequency-amplitude Pareto frontier. Genetic algorithm encoding scheme: frequency (50-150 Hz) 10 bits, amplitude (0.05-0.2 mm) 8 bits, fitness function:

[0045] S603: Based on the Pareto frontier, the entropy weight TOPSIS decision method is adopted to select the optimal parameter combination by weighted Euclidean distance to generate the vibration drag reduction optimization scheme.

[0046] Entropy weight method to calculate index dispersion: TOPSIS closeness calculation:

[0047] The S7-based generation of top-in process optimization instruction set includes the following steps: S701: Based on the vibration drag reduction optimization scheme, the BIM-finite element coupling modeling technology is adopted to automatically update the pipe-soil contact model through parameterized scripts to build a five-dimensional digital twin; The BIM model contains LOD400 geometric data (tolerance ±2mm) and material properties (elastic modulus 35GPa, Poisson's ratio 0.2).

[0048] S702: Based on the five-dimensional digital twin, the long short-term memory network is applied to predict the subsequent jacking parameters through a sliding time window to generate a 10m jacking behavior prediction sequence; The LSTM network input dimension is 12 (including 6 historical state parameters), the hidden layer has 64 units, and the output is the subsequent 10m jacking parameters (step 0.5m).

[0049] S703: Based on the 10m jacking behavior prediction sequence, the multi-criteria decision tree algorithm is adopted to generate optimization instructions through risk-benefit double target evaluation to generate the top-in process optimization instruction set.

[0050] Decision tree splitting criterion: Set the risk threshold: low risk (<3%), medium risk (3%-5%), high risk (5%).

[0051] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A method for constructing large-diameter pipe jacking systems to reduce frictional resistance, characterized in that, Includes the following steps: S1: Based on three-dimensional laser scanning and ground-penetrating radar data, an improved Kriging spatial interpolation algorithm is used to generate a formation impedance spatial distribution model and a geological impedance cloud map by introducing anisotropic variability function. S2: Based on the geoimpedance cloud map, a gradient composite manufacturing process is adopted, and a biomimetic lubrication structure layer is formed on the pipe wall by microcapsule phase change temperature control to prepare a biomimetic drag-reducing pipe section. S3: Based on the geological impedance cloud map and the parameters of the biomimetic drag-reducing pipe section, the multi-objective particle swarm optimization algorithm is adopted to determine the optimal grouting parameters and generate a dynamic grouting parameter matrix by using the constraint that the grouting pressure is ≤1.2 times the static earth pressure. S4: Based on the dynamic grouting parameter matrix, a fuzzy PID control algorithm is adopted to regulate the jacking speed and grouting volume through real-time jacking force sensor data, and generate a dynamic jacking control curve; S5: Based on the jacking dynamic control curve, a long short-term memory network is used to analyze the jacking force fluctuation characteristics, predict the friction resistance change point, and generate a friction resistance early warning heat map. S6: Based on the friction resistance early warning heat map, a genetic algorithm is used to optimize high-frequency micro-vibration parameters through vibration parameter space search, and generate a vibration drag reduction optimization scheme; S7: Based on the vibration reduction and drag reduction optimization scheme, digital twin technology is adopted to predict the jacking parameters for the next 10m by fusing BIM model with real-time data and generating a jacking process optimization instruction set.

2. The method for constructing large-diameter pipe jacking systems to reduce frictional resistance according to claim 1, characterized in that: The generation of geoimpedance contour maps based on S1 includes the following steps: S101: Employs a 3D laser scanner and an adaptive sampling algorithm to acquire surface micro-topography data of the construction area and generate a high-precision point cloud dataset; S102: Based on a high-precision point cloud dataset, a wavelet transform data fusion algorithm is used to perform multi-source data registration between ground-penetrating radar reflection signals and borehole data to generate a three-dimensional geological feature matrix. S103: Based on the three-dimensional geological feature matrix, an improved Kriging spatial interpolation algorithm is applied to construct the stratigraphic parameter field through Bayesian regularization and generate a geological impedance cloud map.

3. The method for constructing large-diameter pipe jacking systems to reduce frictional resistance according to claim 1, characterized in that: The fabrication of a biomimetic drag-reducing pipe section based on S2 includes the following steps: S201: Based on the geological impedance cloud map, the finite element topology optimization algorithm is used to design the distribution pattern of pipe wall reinforcing ribs through stress constraints, and generate the layout scheme of pipe wall reinforcing ribs. S202: Based on the pipe wall reinforcing rib layout scheme, a plasma spraying process is adopted to deposit a PTFE composite coating through gradient temperature control to form a micron-level friction-reducing coating. S203: Based on a micron-level friction-reducing coating, a biomimetic drag-reducing pipe section is prepared by using microcapsule embedding technology and ultrasonic-assisted dispersion of phase change lubricating materials.

4. The method for constructing large-diameter pipe jacking systems to reduce frictional resistance according to claim 1, characterized in that: The generation of dynamic grouting parameter matrix based on S3 includes the following steps: S301: Based on the geoimpedance cloud map, a radial basis function neural network is used to predict the optimal mud viscosity range by inputting formation parameters, and a viscosity-formation characteristic mapping table is generated. S302: Based on the viscosity-formation characteristic mapping table, the multi-objective particle swarm optimization algorithm is applied to optimize grouting parameters and determine the three-dimensional parameter space of pressure-flow-viscosity by using the constraint that grouting pressure ≤ 1.2σ0. S303: Based on the three-dimensional parameter space, a fuzzy PID control model is established, and the grouting rate is adjusted through real-time top force feedback to generate a dynamic grouting parameter matrix.

5. The method for constructing large-diameter pipe jacking systems to reduce frictional resistance according to claim 1, characterized in that: The generation of the jacking dynamic control curve based on S4 includes the following steps: S401: Based on the dynamic grouting parameter matrix, the Kalman filter algorithm is used to estimate the actual jacking force value by fusing multi-sensor data and generate the optimal jacking force estimate. S402: Based on the optimal estimate of the jacking force, model predictive control is applied, and the speed adjustment is calculated through rolling time domain optimization to generate the jacking speed control sequence; S403: Based on the jacking speed control sequence, a feedforward-feedback composite control strategy is adopted. The speed adjustment is performed through the dynamic response model of the hydraulic system to generate the jacking dynamic control curve.

6. The method for constructing large-diameter pipe jacking systems to reduce frictional resistance according to claim 1, characterized in that: The generation of a friction resistance early warning heat map based on S5 includes the following steps: S501: Based on the jacking dynamic control curve, the wavelet packet decomposition algorithm is applied to extract the high-frequency components of the jacking force signal through 5-level decomposition to generate the jacking force time-frequency feature matrix; S502: Based on the top force time-frequency feature matrix, a convolutional neural network is used to identify abnormal fluctuation patterns through the residual connection structure and generate a friction anomaly probability distribution map; S503: Based on the probability distribution map of friction anomalies, the kernel density estimation method is applied to predict the location of abrupt change points through Monte Carlo simulation, and a friction resistance early warning heat map is generated.

7. The method for constructing large-diameter pipe jacking to reduce frictional resistance according to claim 1, characterized in that: The vibration damping reduction optimization scheme based on S6 includes the following steps: S601: Based on the early warning thermal map of frictional resistance, a fast Fourier transform is used to identify the formation resonance frequency through spectrum analysis and generate the formation frequency response spectrum; S602: Based on the formation frequency response spectrum, a genetic algorithm is applied to optimize vibration parameters and determine the vibration frequency-amplitude Pareto front through the fitness function: min(friction coefficient) + max(energy efficiency); S603: Based on the Pareto front, the entropy-weighted TOPSIS decision method is adopted to select the optimal parameter combination through weighted Euclidean distance and generate an optimized vibration drag reduction scheme.

8. The method for constructing large-diameter pipe jacking to reduce frictional resistance according to claim 1, characterized in that: The steps for generating an optimized jacking process instruction set based on S7 are as follows: S701: Based on the vibration damping reduction optimization scheme, BIM-finite element coupled modeling technology is adopted to automatically update the pipe-soil contact model through parametric scripts and construct a five-dimensional digital twin; S702: Based on a five-dimensional digital twin, using a long short-term memory network, predicts subsequent jacking parameters through a sliding time window, generating a 10m jacking behavior prediction sequence; S703: Based on the 10m jacking behavior prediction sequence, a multi-criteria decision tree algorithm is adopted to generate optimization instructions through risk-benefit dual-objective evaluation, thus generating a jacking process optimization instruction set.

Images