Sandy soil remodeling gene compiling method and system driven by dynamic characteristics

Abstract

The invention relates to a sandy soil remodeling gene compiling method and system driven by dynamic characteristics, and the method comprises the steps: obtaining a small strain shear modulus of indoor and outdoor sandy soil through the density of the sandy soil, and obtaining a normalized small strain shear modulus after correction through an overlying stress correction formula; indoor and outdoor test results are integrated to obtain an indoor-in-situ fusion data set, so that an artificial neural network is trained, and optimal variable state parameters of indoor remolded sand are searched through an intelligent optimization algorithm; and taking an absolute value of a difference between a predicted value of the artificial neural network and a corresponding outdoor normalized small strain shear modulus as an optimization objective function, carrying out iterative optimization to establish an indoor-in-situ intelligent remodeling model, and generating an optimal variable state parameter and a corresponding sandy soil gene. The problem of dynamic characteristic deviation caused by the fact that a traditional sandy soil remodeling technology depends on static parameters is solved, and high-precision equivalence of indoor and outdoor soil dynamic response is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of soil dynamics sand remolding, and particularly relates to a sand remolding gene compiling method and system driven by dynamic characteristics. BACKGROUND

[0002] In the field of geotechnical engineering, the core mission of soil dynamics is to accurately reproduce the stiffness response characteristics of soil under dynamic load (such as seismic load and traffic load). With the continuous development of soil dynamics, the research problem has gradually shifted from macroscopic failure phenomenon to microscopic dynamic response control. The elastic stiffness of sand foundation in the small strain range, i.e. the small strain shear modulus, directly controls the wave propagation speed and anti-liquefaction capacity, and has become a key dynamic index for modern seismic design. At the same time, for different sands, if the normalized small strain shear modulus of two kinds of sands is the same after correction, their shear deformation resistance (stiffness) must be the same in the elastic deformation range (under small strain conditions).

[0003] However, for a long time, the development of soil dynamics test methods has seriously lagged behind the theoretical demand. The current soil dynamics test, such as centrifuge test, shaking table model test, cyclic triaxial test, etc., needs to remold the sand. However, the traditional sand remolding technology still takes the relative density as the core of the static characteristic parameter index as the control standard. When the remolded sand is tested, it will cause differences in liquefaction evaluation capacity and elastic shear deformation resistance between the remolded sand and the outdoor natural undisturbed sand. For example, the centrifuge test does not match the normalized small strain shear modulus and permeability characteristics of the outdoor natural undisturbed sand, and the pore structure of the indoor remolded sand is different from that of the outdoor natural undisturbed sand (such as particle contact stiffness and pore connectivity), which leads to a faster rate of rise of pore water pressure under dynamic load than that of the outdoor natural undisturbed sand, resulting in an earlier judgment of liquefaction occurrence and an overestimation of the liquefaction risk of the soil. The normalized small strain shear modulus of the indoor remolded sand in the shaking table model test is greatly different from that of the outdoor natural undisturbed sand, leading to a difference in the propagation speed, phase and energy dissipation of the seismic wave in the model from the actual site. If the normalized small strain shear modulus of the indoor remolded sand is different from that of the outdoor natural undisturbed sand, the particle contact stiffness and stress transmission path will be different, leading to a deviation of the shear dilation / contraction characteristics of the indoor remolded sand from those of the outdoor natural undisturbed sand under cyclic load, and misjudgment of the soil liquefaction time node. Therefore, the traditional sand remolding technology cannot meet the high-precision requirements of modern seismic analysis on soil dynamic response, and therefore it is necessary to consider a similar model of the outdoor natural undisturbed sand and the indoor remolded sand driven by the normalized small strain shear modulus to reproduce the in-situ dynamic characteristics in the indoor test.

[0004] With the development of science and technology, artificial neural networks have been gradually applied to the field of soil dynamics, but the traditional sand remolding technology still relies on manual trial and error and experience to adjust the static characteristic parameters such as relative density, which has a long experimental period, high cost, and is difficult to cope with the complex and variable conditions of the soil, and without the combination of artificial neural networks, the physical and mechanical properties of the outdoor undisturbed natural sand and the indoor remolded sand are different.

[0005] The sand remolding genetic compilation process driven by dynamic characteristics controls the normalization small strain shear modulus index and combines with artificial neural networks, and various physical and mechanical property parameters of the indoor remolded sand required for the soil dynamics test. SUMMARY

[0006] In view of the deficiencies of the prior art, the technical problems to be solved by the present application are to provide a sand remolding genetic compilation method and system driven by dynamic characteristics, and to construct an indoor-in-situ intelligent remolding model based on the normalization small strain shear modulus of sand, so as to remold the variable state parameters of the sand by using the indoor-in-situ intelligent remolding model, solve the dynamic characteristic deviation problem caused by the dependence of the traditional sand remolding technology on the static parameter of relative density, and realize the high-precision equivalence of the indoor and outdoor soil dynamic response.

[0007] The technical solution adopted by the present application to solve the technical problem is: In a first aspect, the present application provides a sand remolding genetic compilation method driven by dynamic characteristics, characterized in that the compilation method comprises the following contents: The small strain shear modulus of the indoor remolded sand under different vertical effective stresses, water contents and relative densities is obtained by indoor test when the particle size, particle size distribution, maximum void ratio and minimum void ratio of the standard sand are unchanged, the small strain shear modulus of the indoor remolded sand is corrected, and the indoor normalized small strain shear modulus is obtained; the indoor normalized small strain shear modulus is corresponding to the particle size, particle size distribution, maximum void ratio and minimum void ratio of the standard sand, and the corresponding vertical effective stress, water content and relative density, forming an indoor data set; The vertical effective stress and small strain shear modulus of the outdoor undisturbed natural sand under different burial depths are obtained by outdoor test, and the particle size, particle size distribution, maximum void ratio, minimum void ratio, water content and relative density of the outdoor undisturbed natural sand are obtained by sampling and laboratory test, the small strain shear modulus of the outdoor undisturbed natural sand is corrected, and the outdoor normalized small strain shear modulus is obtained; the outdoor normalized small strain shear modulus is corresponding to the particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content and relative density, forming an outdoor undisturbed data set; The indoor data set and the outdoor undisturbed data set are fused to form an indoor-in-situ fusion data set; An artificial neural network is constructed with particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content and relative density as input characteristics, and normalized small-strain shear modulus as target output, and the artificial neural network is trained using the indoor-in-situ fusion dataset to obtain a normalized small-strain shear modulus prediction model; An indoor-in-situ intelligent remodeling model driven by normalized small-strain shear modulus of sand is constructed using the normalized small-strain shear modulus prediction model and an intelligent optimization algorithm, and the indoor-in-situ intelligent remodeling model takes the normalized small-strain shear modulus outdoors as input to obtain the optimal variable state parameters of the indoor remodeled sand; wherein the optimization objective function of the indoor-in-situ intelligent remodeling model is the absolute value of the difference between the predicted value of the normalized small-strain shear modulus prediction model and the normalized small-strain shear modulus outdoors corresponding to the input; The variable state parameters are at least one of vertical effective stress, water content, and relative density.

[0008] In a second aspect, the present application provides a power characteristic driven sand remolding genetic compilation system, which comprises: An indoor bending element device acquires shear wave velocity data of standard sand under different working conditions; An electro-hydraulic servo universal testing machine is used to simulate vertical effective stress; An outdoor bending element device acquires shear wave velocity data of outdoor undisturbed natural sand under different working conditions; A function signal generator is used to apply a set voltage and frequency sine wave to the transmitting bending element sensor; A function signal amplifier is used to receive and amplify the electrical signal of the receiving bending element sensor; An oscilloscope is used to input waveform parameters and simultaneously display output and received waves; An intelligent remodeling data management platform comprises a data processing module, a database, a display module, and an indoor-in-situ intelligent remodeling model; The data processing module acquires different standard sands and their corresponding normalized small-strain shear modulus under different water contents, different relative densities, and different vertical effective stress controls, and simultaneously acquires outdoor undisturbed natural sand at different measurement positions and different depths, as well as corresponding particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content, relative density, and normalized small-strain shear modulus; and arranges the data to obtain an indoor-in-situ fusion dataset; The indoor-in-situ intelligent reshaping model takes the normalized small strain shear modulus of the outdoor soil as input and initializes and generates candidate combinations of variable state parameters within the constraints of variable state parameters. It calls an artificial neural network to predict the corresponding normalized small strain shear modulus of the candidate combinations of variable state parameters and uses an intelligent optimization algorithm to obtain the variable state parameters of the corresponding indoor reshaping sand. The database stores variable state parameters and inherent physical parameters of standard sand and outdoor undisturbed natural sand under different working conditions, as well as normalized small strain shear modulus, and supports engineering applications to call them. The database communicates with the data processing module, the indoor-in-situ intelligent reshaping model and the display module through standardized interfaces. The data processed by the data processing module is stored in the database. The indoor-in-situ intelligent reshaping model can call the data in the database to dynamically update and iteratively optimize the indoor-in-situ intelligent reshaping model. The display module supports the visualization of parameters.

[0009] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Existing indoor remolded sand technology uses relative density as the core static characteristic parameter as the control index, relying on manual trial and error adjustments, making it difficult to correlate with the dynamic mechanical response characteristics of the soil. This invention innovatively introduces the normalized small-strain shear modulus as an equivalent benchmark. By unifying the testing standards through indoor and outdoor bending element devices, it ensures the consistency of the normalized small-strain shear modulus between outdoor undisturbed natural sand and indoor remolded sand, thereby reflecting the equivalence of the soil's liquefaction resistance and elastic deformation capacity. This overcomes the problem of the disconnect between traditional static and dynamic characteristics, significantly improving the response fidelity of indoor remolded sand under complex dynamic loads.

[0010] 2. This invention unifies the measurement method of indoor and outdoor shear wave velocity—bending element test, and drives the establishment of an indoor-in-situ intelligent remodeling model through artificial neural network and intelligent optimization algorithm. It establishes a dynamic mapping between the normalized small strain shear modulus and variable state parameters in the outdoor environment, forming a closed loop from in-situ testing to laboratory remodeling.

[0011] 3. Establish a unified and standardized process for obtaining normalized small strain shear modulus both indoors and outdoors, covering the entire process from sample preparation and parameter measurement to data processing. Construct an indoor-in-situ fusion dataset, and combine artificial neural networks and intelligent optimization algorithms for iterative optimization to form a closed-loop system of "measurement-modeling-optimization-verification". This will significantly shorten the test cycle, reduce costs, and improve the accuracy and adaptability of remodeling variable state parameters under complex soil conditions.

[0012] 4. The present invention makes targeted improvements to the indoor and outdoor bending element device, which improves the stability, durability and measurement accuracy of the device, and provides hardware guarantee for the reliability of test data.

[0013] 5、Traditional sand remolding lacks systematic data management, and test parameters are stored in a scattered manner, which is difficult to iterate and optimize. The sand remolding gene compiling system driven by dynamic characteristics is based on the integration of outdoor detection, laboratory remolding, and intelligent remolding data management platform, realizes the full-process automation and intelligentization from field testing, data acquisition, model training to parameter optimization, has data full-life cycle management and strong analysis and processing capability, can quickly respond to different engineering requirements, and provides an efficient and intelligent solution for soil dynamics test and engineering application.

[0014] 6、In the application, the shear wave velocity of the indoor and outdoor sand is preferably measured by the bending element device, the small strain shear modulus of the indoor and outdoor sand is obtained by the density of the sand, and the normalized small strain shear modulus is obtained after correction by the overburden stress correction formula; the test results of the indoor and outdoor are integrated to obtain an indoor-in-situ fusion data set, an artificial neural network is trained based on the data set, and an intelligent optimization algorithm is used to find the optimal variable state parameters of the indoor remolding sand; the absolute value of the difference between the predicted value of the artificial neural network and the corresponding normalized small strain shear modulus of the outdoor is taken as an optimization objective function, and an indoor-in-situ intelligent remolding model is established through iterative optimization; the trained indoor-in-situ intelligent remolding model is written into the intelligent remolding data management platform, and the optimal variable state parameters generated by the indoor-in-situ intelligent remolding model and the corresponding sand gene are monitored and displayed in real time. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 It is a flowchart of an embodiment of the sand remolding gene compiling method driven by dynamic characteristics.

[0016] Figure 2 It is a structural diagram of the indoor bending element device of an embodiment of the application.

[0017] Figure 3 It is a side plate diagram of the indoor bending element device of an embodiment of the application.

[0018] Figure 4 It is a structural diagram of the protective sleeve body after placing the bending element sensor.

[0019] Figure 5 It is a structural diagram of the outdoor bending element device of an embodiment of the application.

[0020] Figure 6 It is a structural diagram of the sand remolding gene compiling system driven by dynamic characteristics.

[0021] Figure 7 It is a structural diagram of the normalized small strain shear modulus prediction model.

[0022] Figure 8This is a schematic diagram of the modeling process for the indoor-in-situ intelligent reshaping model in this invention. Detailed Implementation

[0023] The present invention will be further explained below with reference to the embodiments and accompanying drawings, but this is not intended to limit the scope of protection of this application.

[0024] The present invention provides a method for compiling sand remodeling genes driven by dynamic characteristics, comprising the following: The small-strain shear modulus of remolded sand was obtained through indoor experiments under different vertical effective stresses, water contents, and relative densities, with standard sand having constant particle size, particle size distribution, maximum void ratio, and minimum void ratio. The small-strain shear modulus of the indoor remolded sand was corrected using the overlying stress correction formula to obtain the normalized small-strain shear modulus. The normalized small-strain shear modulus was then correlated with the particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content, and relative density of the standard sand used to form an indoor dataset. The indoor tests employed indoor bending element tests, using soil dynamics standard sands such as Fujian standard sand, Toyoura sand from Japan, or Ottawa sand from the United States. With fixed particle size, particle size distribution, maximum void ratio, and minimum void ratio, the shear wave velocity was measured using an indoor bending element device by varying the effective vertical stress, water content, and relative density. Simultaneously, the density data of the standard sand is obtained by combining the sand cone method or the ring cutter method. Through standardized physical models Accurate calculation of small strain shear modulus of remolded sand in the laboratory .

[0025] Using the overburden stress correction formula The small-strain shear modulus of the remolded sand in the laboratory was corrected to obtain the normalized small-strain shear modulus. ,in This represents the effective vertical stress.

[0026] The vertical effective stress and small strain shear modulus of undisturbed natural sand at different burial depths were obtained through outdoor experiments. Samples of the undisturbed natural sand were taken, and the particle size, particle size distribution, maximum void ratio, minimum void ratio, water content, and relative density of the undisturbed natural sand were obtained in the laboratory. The small strain shear modulus of the undisturbed natural sand was corrected using the overlying stress correction formula to obtain the normalized small strain shear modulus of the outdoor sand. The normalized small strain shear modulus of the outdoor sand was then correlated with particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content, and relative density to form an outdoor undisturbed dataset. The outdoor test employed an outdoor bending element test, which used an outdoor bending element device to measure the shear wave velocity of sand at different burial depths h along the soil depth direction. Simultaneously, layered density tests were conducted using the sand cone method and the ring cutter method to obtain the density of sand at corresponding depths. ; through formula Obtain the vertical effective stress of outdoor undisturbed natural sand. ,in For effective severity; through a standardized physical model Accurate calculation of small strain shear modulus of outdoor undisturbed natural sand ; Using the overburden stress correction formula The small-strain shear modulus of the outdoor undisturbed natural sand is corrected to obtain the normalized small-strain shear modulus of the outdoor soil.

[0027] Meanwhile, samples of unspoiled natural sand were taken outdoors and tested in the laboratory for particle size, particle size distribution, maximum void ratio, minimum void ratio, moisture content, and relative density.

[0028] The indoor dataset and the outdoor untouched dataset are merged to form an indoor-in-situ fused dataset; Using particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, moisture content, and relative density as input features, and normalized small-strain shear modulus as the target output, an artificial neural network is constructed. This artificial neural network is then trained using an indoor-in-situ fusion dataset to obtain a normalized small-strain shear modulus prediction model (see [link to relevant documentation]). Figure 7 ); An indoor-in-situ intelligent remodeling model (hereinafter referred to as the indoor-in-situ intelligent remodeling model) based on the normalized small strain shear modulus prediction model and intelligent optimization algorithm is constructed. The indoor-in-situ intelligent remodeling model takes the outdoor normalized small strain shear modulus as input to obtain the optimal variable state parameters of the indoor remodeled sand. The optimization objective function of the indoor-in-situ intelligent remodeling model is the absolute value of the difference between the predicted value of the normalized small strain shear modulus prediction model and the corresponding input outdoor normalized small strain shear modulus. The variable state parameter is at least one of vertical effective stress, moisture content, and relative density.

[0029] The present invention provides an indoor-in-situ intelligent remodeling model, which uses the normalized small strain shear modulus of outdoor undisturbed natural sand as a benchmark. It uses an intelligent optimization algorithm to call the normalized small strain shear modulus prediction model to invert the laboratory remodeling parameters, which are the variable state parameters.

[0030] In actual use, the particle size, particle size distribution, maximum void ratio, and minimum void ratio of the outdoor original natural sand and the standard sand are all preset inherent physical parameters in the indoor-in-situ intelligent remodeling model. The candidate combination of variable state parameters generated by the intelligent optimization algorithm in each iteration is input into the normalized small-strain shear modulus prediction model together with the inherent physical parameters, and the corresponding normalized small-strain shear modulus prediction value is output by the normalized small-strain shear modulus prediction model. The intelligent optimization algorithm takes the input outdoor normalized small-strain shear modulus as the target value, iteratively optimizes by calculating the deviation (the optimization objective function is the absolute value of the deviation) between the prediction value and the target value, and outputs the optimal variable state parameters until the convergence condition is met.

[0031] In the indoor-in-situ intelligent remodeling model, the outdoor normalized small-strain shear modulus (target value) is input, the intelligent optimization algorithm is executed, the particle swarm is initialized, and a series of candidate combinations of variable state parameters are generated by initialization. Then, the normalized small-strain shear modulus prediction model is called to predict the corresponding normalized small-strain shear modulus (prediction value) using the candidate combination, and then the absolute value of the difference between the prediction value and the target value is calculated as the optimization objective function to obtain the optimal variable state parameters of the indoor remolded sand corresponding to the current outdoor original natural sand.

[0032] In the artificial neural network, the inherent physical parameters and the variable state parameters are used as input features together, and the normalized small-strain shear modulus is used as the target output. The inherent physical parameters include the particle size, particle size distribution, maximum void ratio, and minimum void ratio of the indoor standard sand and the outdoor original natural sand, and the variable state parameters include the vertical effective stress, water content, and relative density. The artificial neural network is trained using the indoor-in-situ fusion dataset, which contains the properties of the standard sand and the outdoor original natural sand, and combines the response of the standard sand with the complex response of the outdoor original natural sand. It can learn the mapping rule of the normalized small-strain shear modulus under different conditions of variable vertical effective stress, variable water content, and variable relative density, and new test data (such as sand in different regions) can be injected into the artificial neural network in real time. Through continuous optimization of network weights by learning, the adaptability is improved.

[0033] The application is trained by an artificial neural network, and then optimized by an intelligent optimization algorithm, wherein the normalized small strain shear modulus of the outdoor is taken as the input of the intelligent optimization algorithm, the variable state parameters are taken as the output of the intelligent optimization algorithm, and the required indoor remolded sand soil is obtained by combining the inherent physical parameters of the standard sand and the variable state parameters and the inherent physical parameters. When the indoor sand soil type is determined (such as Fujian standard sand), the particle size, particle size distribution, maximum / minimum void ratio can be fixed, and only one parameter of the vertical effective stress, water content and relative density is optimized, which is suitable for the scene of needing to quickly match a single index. For complex sites (such as multi-layered sand soil interbedding), the vertical effective stress, water content and relative density can be optimized at the same time, the global optimal combination is searched by the intelligent optimization algorithm, the equivalence of the normalized small strain shear modulus is ensured, and the sample prepared according to the corresponding parameters is subjected to the indoor bending element test. The normalized small strain shear modulus obtained by the indoor bending element test is fed back to the indoor-in-situ intelligent remolding model as a verification result, which is used to evaluate the performance of the indoor-in-situ intelligent remolding model and continuously update the indoor-in-situ intelligent remolding model. The optimized indoor-in-situ intelligent remolding model can optimize the parameters according to different scenes and different needs, and has very high applicability.

[0034] The artificial neural network can be at least one of a BP neural network, an ANN neural network or an RNN neural network, and the intelligent optimization algorithm can be at least one of a PSO algorithm, a GWO algorithm or a GA algorithm.

[0035] In the present application, the composition elements of the sand soil gene include inherent physical parameters and variable state parameters, wherein the particle size, particle size distribution, maximum void ratio and minimum void ratio are taken as the inherent physical parameters, and the vertical effective stress, water content and relative density are taken as the variable state parameters, and the gene compilation is mainly the compilation of the variable state parameters by the intelligent optimization algorithm.

[0036] The dynamic characteristic driving in the present application is driven by the calculation of the normalized small strain shear modulus. The optimal combination of the variable state parameters of the indoor remolded sand soil is found by the intelligent optimization algorithm (such as PSO), the dynamic characteristic directional expression of the sand soil parameter genotype deep compilation is realized, and the variable state parameter optimization process is directly described by using the term of “gene compilation”.

[0037] Specifically, the sand soil gene includes a particle size D50, a non-uniformity coefficient Cu, a maximum void ratio , a minimum void ratio , a water content w, a relative density and a vertical effective stress ; the gene compilation refers to the optimization of , w, The combination of the two makes the normalized small strain shear modulus predicted by the artificial neural network approach the in-situ target value.

[0038] The present application takes "predicted value approaching in-situ target value" as the optimization direction instead of directly taking the normalized small strain shear modulus output parameter in the outdoor as the target optimization. The intelligent optimization algorithm is used to deduce the optimal remodeling parameters reversely. The artificial neural network is used as a forward proxy model, which essentially learns the "constitutive behavior" of the sand. The optimization objective function of the intelligent optimization algorithm (such as PSO) is to "narrow the gap between the prediction and the in-situ target value", which is equivalent to letting the algorithm automatically explore the feasible solution set in the parameter space that can meet the dynamic stiffness equivalence. The biggest wisdom of this method is that it does not pre-set the parameter combination, but lets the physical law itself converge to the optimal solution through data-driven way. Direct parameter fitting cannot be done, and the error of each experimental verification will be fed back to the indoor-in-situ intelligent remodeling model iteration, which is equivalent to installing an "adaptive calibrator". "Gene compilation": this is not a simple parameter copy, but a simulation of the "dynamic performance type" of the sand. The training data of the artificial neural network comes from the indoor-in-situ fusion data set, which naturally contains the parameter feasible region. The intelligent optimization algorithm automatically avoids the violation area in the search, which is more reasonable than direct parameter optimization.

[0039] The function signal generator 001 can generate and accurately adjust multiple standard waveform signals (such as sine wave, square wave, triangle wave), and can generate a stable sine wave signal by controlling parameters such as frequency, amplitude and phase.

[0040] The function signal amplifier 002 is used to amplify weak electrical signals (such as voltage, current or power) to improve signal strength for subsequent processing or transmission.

[0041] The oscilloscope 003 is used to capture and visualize electrical signal waveforms, measure voltage, frequency, phase and other parameters.

[0042] Embodiment 1: The sand remodeling gene compilation method driven by dynamic characteristics in this embodiment includes the following contents: 1) Indoor standardized bending element detection Select a specific soil dynamics standard sand, strictly control and record its inherent physical parameters (particle size, particle size distribution, maximum void ratio, minimum void ratio, etc.), use a layered compaction instrument, a shaking table or an air / water sedimentation method based on the target void ratio to prepare remodeling samples with a series of relative densities. For unsaturated state, use controllable spray infiltration to accurately prepare samples with a series of initial water contents. Saturated samples use standard saturation processes such as CO2 replacement and reverse pressure saturation. Install the prepared samples in the indoor bending element device, and simulate the vertical effective stress through an electro-hydraulic servo universal testing machine.

[0043] The indoor bending element device is plated with nickel on the inner and outer surfaces, which ensures the rust and corrosion resistance of the iron plate. The indoor bending element device 005 includes a rectangular experimental box sealed by four corrosion-resistant side plates and a bottom plate, a heavy load foot 2 installed at the bottom of the experimental box, and a protective sleeve; a sealing groove is processed at the connection position of the four side plates, and an acid and alkali resistant high temperature fluorine rubber circular sealing strip 1 is arranged in the sealing groove, the adjacent side plates are connected by bolts, and when the bolts are tightened, the fluorine rubber circular sealing strip 1 is in excess of the volume of the sealing groove, thereby ensuring that the waterproof sand box does not leak. Four heavy load feet 2 are used to bear the weight of the experimental box and the load, and the heavy load foot 2 adopts M10 heavy load foot, each heavy load foot can bear 300KG pressure, and four heavy load feet can completely bear the weight of the experimental box and the load. A pipe threaded hole 7 is arranged at the bottom of one side plate of the experimental box, and a metal water permeable stone is arranged in the pipe threaded hole, so that only the water in the sand soil flows out when the sand soil in the experimental box body is subjected to pressure, and the sand soil does not overflow. The pipe threaded hole 7 is provided with an electromagnetic valve 3 outside, which controls the drainage condition during the test.

[0044] The four side plates and the bottom plate are all made of iron plates, which are connected by bolts to form a box body, and the inner and outer surfaces of the five iron plates are plated with nickel after processing, which ensures the rust and corrosion resistance of the iron plate. A pair of opposite side plates are provided with a reserved hole for mounting a protective sleeve, and the electromagnetic valve 3 adopts AC220V power supply voltage to control the drainage condition during the test. The protective sleeve includes a protective sleeve cover 5 and a protective sleeve body 6, and the protective sleeve cover 5 and the protective sleeve body 6 are provided with a reserved packaging groove at the front end, and the protective sleeve cover 5 and the protective sleeve body 6 are connected by bolts in alignment, and the bending element sensor 4 is packaged with epoxy resin to form a protective layer on the surface, which is insulated from the outside. The bending element sensor is fixed in the protective sleeve by bolts, and after the protective sleeve cover 5 and the protective sleeve body 6 and the bending element sensor 4 are fixed and installed, the packaging groove is filled with silica gel, and then it is left to wait for the silica gel to completely solidify. At this time, the bending element sensor is completely packaged in the protective sleeve except for the exposed part, which is insulated and can control the length of the bending element sensor extending from the front end by bolts, thereby adjusting the accuracy of the measurement data. After the protective sleeve and the silica gel are solidified, the whole formed by the protective sleeve and the bending element sensor is passed through the reserved hole on the side plate, and the protective sleeve is fixed on the side plate by bolts, thereby forming a complete loading experimental environment, which is completely sealed and does not leak.

[0045] The function signal generator 001 is connected with the transmitting bender element sensor through one end of the BNC connecting line and connected with the oscilloscope 003 through the other end. The function signal amplifier 002 is connected with the receiving bender element sensor through one end of the BNC connecting line and connected with the oscilloscope 003 through the other end. The function signal generator 001 is used to apply the sine wave with the set voltage and frequency to the transmitting bender element sensor. The piezoelectric effect forces the transmitting bender element sensor to elongate on one side and shorten on the other side, to generate the bending motion and the transverse vibration in the surrounding soil. The shear wave generated in the direction perpendicular to the vibration is transmitted to the receiving bender element sensor at the other end through the soil. After receiving the shear wave, the receiving bender element sensor is forced to generate a transverse swing and to convert the mechanical swing into the electrical signal which is displayed in the oscilloscope 003 through the function signal amplifier 002.

[0046] After the target vertical effective stress of the sample is stabilized in the electro-hydraulic servo universal testing machine, the integrated intelligent remolding data management platform 004 is started to collect the shear wave velocity data. After the test of the bender element sensor is completed, the density test of the sample is immediately performed. The small strain shear modulus of the remolding sand soil in the laboratory is calculated by using the standardized physical model. The small strain shear modulus of the remolding sand soil in the laboratory is obtained by correcting through the overburden stress correction formula. All the parameters (particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content, relative density, normalized small strain shear modulus) of the current sample in the laboratory test are associated and stored to form a highly structured indoor data set.

[0047] 2) Outdoor in-situ bender element detection The outdoor bender element device includes a transverse fixed rod 8 and two vertical fixed rods 9. The transverse fixed rod 8 can adjust the spacing of the two vertical fixed rods 9 and ensure that the vertical fixed rods 9 are in the same plane. Each vertical fixed rod 9 is composed of two half steel sheets and connected into a whole through bolts with the transverse fixed rod. A tapering hammer 10 is assembled at the bottom of the vertical fixed rod 9 to facilitate insertion into the soil and effectively reduce the disturbance to the surrounding soil during the tapering process. A screw hole is provided at the lower end of the vertical fixed rod 9 for installing the bender element sensor 4. Flat iron sheets 12 are installed above and below the bender element sensor 4 to protect the device when it is inserted into and pulled out of the soil. At the same time, a groove is provided in the vertical fixed rod 9 for the cable line of the bender element sensor to pass through. The cable line is connected with the function signal generator 001 and the function signal amplifier 002 respectively. A scale 13 is engraved on the outside of the vertical fixed rod to detect the depth of the device into the soil. The exposed part of the bender element sensor 4 is encapsulated with epoxy resin and sealed with silicone at the gap of the vertical fixed rod 9 to ensure strict waterproofing and accurate measurement data. The transmitting and receiving bender element sensors of the outdoor bender element device need to be inserted into the same depth of the soil to be tested.

[0048] After selecting the measurement site, the outdoor bending element device is used to test the outdoor original natural sand. The function signal generator 001 applies a sine wave with a set voltage and frequency to the transmitting bending element sensor. The piezoelectric effect forces the transmitting bending element sensor to elongate on one side and shorten on the other side, generating bending motion and producing transverse vibration in the surrounding soil. The shear wave generated in the direction perpendicular to the vibration is transmitted through the soil to the receiving bending element sensor at the other end. The receiving bending element sensor is forced to produce a transverse swing after receiving the shear wave, and the mechanical swing is converted into an electrical signal and displayed on the oscilloscope 003 through the function signal amplifier 002. The outdoor bending element device is precisely lowered to the predetermined test depth h, and the function signal generator 001 sends instructions to excite the bending element sensor to generate shear waves and record the waveform signal to obtain the shear wave velocity of the outdoor original natural sand. The formula is used to obtain the vertical effective stress of the outdoor original natural sand where is the effective density. After the test is completed, a small amount of outdoor original natural sand is obtained, and the density, particle size, particle size distribution, maximum void ratio, minimum void ratio, water content, and relative density are precisely measured in the laboratory. The standardized physical model is used to accurately calculate the small strain shear modulus of the outdoor in-situ natural sand at the depth, and the overburden stress correction formula is used for correction to obtain the normalized small strain shear modulus of the outdoor. All parameters (particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content, relative density, and normalized small strain shear modulus) of the current measurement point are associated and stored to form a highly structured outdoor original data set.

[0049] 3) Indoor-in-situ fusion data set The indoor data set and the outdoor original data set are strictly cleaned, formatted, and standardized. All parameter definitions are clear and measurement methods are clear. The "standard" response in the laboratory and the "true" response in-situ are placed in the same model framework to learn, so that the model has generalization ability and can predict whether the selected standard sand can achieve the target normalized small strain shear modulus under a specific parameter combination.

[0050] 4) An intelligent optimization algorithm is employed, setting the constraint range of the variable state parameters. The normalized small strain shear modulus from the outdoor environment is input, and the optimal combination of variable state parameters is searched. Laboratory personnel conduct indoor bending element tests on the reshaped specimens based on the optimal variable state parameters and corresponding inherent physical parameters provided by the intelligent optimization algorithm. The actual normalized small strain shear modulus of the reshaped specimens is calculated to verify the equivalence of the indoor-in-situ intelligent reshaping model. This verification result is also fed back to the system to evaluate the performance of the indoor-in-situ intelligent reshaping model or to expand the indoor-in-situ fusion dataset as new data points, continuously optimizing the indoor-in-situ intelligent reshaping model and forming a closed loop. The constraint range is determined by the physical boundaries and physical states of the variable state parameters. The constraint range refers to the reshaping parameters (vertical effective stress) calculated by the intelligent optimization algorithm. Moisture content (w), relative density During the search and inversion process, hard or soft constraints are applied. These constraints are derived directly from the physical properties of the soil, the limits of laboratory equipment capabilities, and engineering experience, aiming to ensure that the optimization result is not only mathematically optimal but also meets engineering practice requirements.

[0051] Example 2: The method for compiling sand remodeling genes driven by dynamic characteristics in this example (see...) Figure 1 The process is as follows: Step 1: Collect indoor dataset For indoor testing, standard sands such as Fujian standard sand, Toyoura sand from Japan, or Ottawa sand from the United States can be used. This example uses Fujian standard sand produced by Xiamen Aisio Standard Sand Co., Ltd. for illustration. 1.1 Sand Sample Pretreatment According to the "Standard for Geotechnical Testing Methods", the particle size (D50=0.55mm), particle size distribution (Cu=1.54), and maximum void ratio of Fujian standard sand were determined. =0.83) and minimum void ratio ( (=0.48) remained unchanged. The Fujian standard sand was dried in an oven at 105℃ for 24 hours to completely remove moisture and obtain dry sand.

[0052] 1.2 Moisture Content Control Preparation of wet Fujian standard sand: Assume that the mass of dried Fujian standard sand is weighed. For a weight of 1000g, the moisture content w is selected as 2%, 4%, 6%, 8%, 10%, 12%, 14%, 16%, 18%, and 20%, and calculated using the formula... Calculate target moisture content Spray water on the surface of the sand sample layer by layer with a precision sprayer, stir by hand for 3 minutes after each layer is sprayed, until the target water content is fully sprayed into Fujian standard sand, obtain wet Fujian standard sand with corresponding water content, and place the wet Fujian standard sand in a sealed bag, and stand in a constant temperature environment (20±1°C) for ≥24 hours, so that the moisture is uniformly penetrated.

[0053] Preparation of saturated Fujian standard sand: after the dried Fujian standard sand is placed in a mold, vacuumize for 30 minutes, inject degassed water, and apply a counter pressure ≥200 kPa until the B value ≥0.95. The B value refers to the coefficient of pore water pressure, which is the ratio of the increment of pore water pressure to the increment of confining pressure.

[0054] 1.3 Relative density control Selection of the relative density of Fujian standard sand 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, and 90%, respectively, the pore ratio of Fujian standard sand corresponding to different relative densities is calculated by the pore ratio formula , wherein , are the maximum and minimum pore ratios, respectively.

[0055] The experimental box volume V of the indoor flexural element device is obtained as follows: the bottom area of the experimental box is , the height is , and the volume is : The solid volume V of Fujian standard sand is calculated by the formula , The dry sand mass of Fujian standard sand is obtained by the formula , , wherein, is the specific gravity of soil particles, the of Fujian standard sand is 2.65 g / cm 3 ; is the density of water, and the value is 1 g / cm 3 .

[0056] Preparation of wet Fujian standard sand with different water contents corresponding to different relative densities according to the water content control method: first, obtain the dry sand mass according to the set relative density through the above process, then calculate the target water content according to the target water content and the dry sand mass, and spray water in the dry sand to obtain wet Fujian standard sand with the set relative density and water content. Then, the wet Fujian standard sand is divided into 3-5 layers and placed in the indoor flexural element device, each layer is compacted with a 2.5 kg rammer, and the rammer is dropped for 10-15 times (the energy is about 600 kJ / m 3 ) with a drop distance of 30 cm to achieve the target relative density.

[0057] 1.4 Vertical effective stress control Put the indoor bending element device filled with prepared wet Fujian standard sand under the electro-hydraulic servo universal testing machine. Before pressurization, place the pressurization block on the wet Fujian standard sand to enable the universal testing machine to uniformly apply pressure. Control the electro-hydraulic servo universal testing machine to apply stable pressures of 100 kPa, 200 kPa, 300 kPa, 400 kPa, 500 kPa, 600 kPa, 700 kPa, and 800 kPa, respectively, and keep them unchanged during the test stage. After each stress loading, monitor the deformation rate ≤0.005 mm / min to be considered as consolidation stability.

[0058] 1.5 Shear wave velocity and density measurement Connect the two ends of the function signal generator 001 to the transmitting bending element sensor and the oscilloscope 003, respectively. Connect the function signal amplifier 002 to the receiving bending element sensor. The data collected by the receiving bending element sensor is connected to the function signal amplifier 002 and the oscilloscope 003.

[0059] Turn on the function signal generator 001, function signal amplifier 002, and oscilloscope 003. Input the waveform parameters in the oscilloscope 003. The oscilloscope 003 simultaneously displays the output wave and the received wave. Read the time difference between the first peaks of the two waveforms .

[0060] Use the scale to measure the horizontal distance between the transmitting and receiving bending element sensors of the indoor bending element device . Calculate the shear wave velocity of the standard sand through the formula . ; After the indoor test, measure the density of the sample using the cutting ring method . Obtain the small strain shear modulus corresponding to the indoor test through the formula . .

[0061] Correct the small strain shear modulus of the remolded sand soil using the formula to obtain the normalized small strain shear modulus .

[0062] Collect data on different standard sands with different water contents, different relative densities, and different vertical effective stress controls through the intelligent remolding data management platform 004 and correspond one by one. Preliminarily arrange the collected data and check the rationality and consistency of the data. Store all the arranged data in a unified format to form an indoor data set.

[0063] Step two: Collecting the outdoor undisturbed dataset 2.1 Calculation of vertical effective stress Align the outdoor flexural element device with the measurement point, press it into the soil at a constant speed of 0.5 cm / s to the target depth by using a cone hammer, and record the depth of penetration h by the external scale of the outdoor flexural element device. Calculate the vertical effective stress layer by layer. .

[0064] For the soil layer above the groundwater level, the vertical effective stress is obtained by the formula , , where i is the number of the target soil layer below the soil; is the natural density of the jth layer of soil below the soil, which can be determined from regional geological data; is the thickness of the jth layer of soil below the soil, and the corresponding top surface of the soil layer is higher than the groundwater level.

[0065] For the soil layer below the groundwater level, the vertical effective stress is obtained by the formula , where k is the number of the soil layer where the groundwater level is located, is the natural density of the jth layer of soil below the soil above the groundwater level, is the natural effective density of the jth layer of soil below the groundwater level, ; 3 ; is the thickness of the jth layer of soil below the soil.

[0066] 2.2 Shear wave velocity measurement Connect the two ends of the function signal generator 001 to the transmitting flexural element sensor of the outdoor flexural element device and the oscilloscope 003 respectively, and connect the receiving flexural element sensor of the outdoor flexural element device to the oscilloscope 003 through the function signal amplifier 002. Turn on the function signal generator 001, the function signal amplifier 002 and the oscilloscope 003, input the waveform parameters, and the oscilloscope 003 simultaneously displays the output wave and the received wave, and reads the time difference between the first wave peaks of the two waveforms . Measure the horizontal distance between the two flexural element sensors of the outdoor flexural element device with a scale , and calculate the shear wave velocity of the outdoor undisturbed natural sand soil by the formula . .

[0067] 2.3 Sampling of outdoor undisturbed natural sand soil and determination of laboratory parameters At the same depth within a horizontal distance of ≤30 cm from the outdoor flexural element device, use a thin-walled soil sampler to press into the measurement point at a constant speed of 1 m / min, and immediately seal the two ends with wax after taking out. After sampling, the density of the sample is obtained in the laboratory by the cutting ring method Through formula The small strain shear modulus of the corresponding outdoor undisturbed natural sand was obtained. Using the formula The small-strain shear modulus of undisturbed natural sand is corrected to obtain the normalized small-strain shear modulus. .

[0068] The moisture content of the soil at the testing points was obtained by the standard drying method; particle size D50, particle size distribution C u and the maximum and minimum void ratio ( , The results were obtained by standard sieving, laser spectroscopy, and vibration table spectroscopy, respectively; and obtained through formulas. ( The specific gravity of soil particles. The density of water, The void ratio of the undisturbed natural sand (where the dry sand density can be measured in a laboratory) is obtained. ; through formula The relative density of the outdoor undisturbed natural sand was obtained. .

[0069] By inserting outdoor bending element devices at different positions and depths, multiple sets of outdoor undisturbed data can be measured. This data is collected through the Intelligent Reshaping Data Management Platform 004, and the particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, moisture content, relative density, and normalized small strain shear modulus are mapped one-to-one. The collected outdoor undisturbed data is then preliminarily processed to check its rationality and consistency. All processed data is stored in a unified format to form an outdoor undisturbed data set.

[0070] Step 3: Merge the indoor dataset with the outdoor untouched dataset to form an indoor-in-situ fused dataset. The data from the indoor and outdoor untouched datasets are preprocessed using the Intelligent Reshaping Data Management Platform 004. Rigorous data cleaning, format standardization, and unit standardization are performed on both datasets to ensure clear parameter definitions and explicit measurement methods. The indoor and outdoor untouched datasets are then merged to obtain the indoor-in-situ fused dataset.

[0071] Step 4: Establish an indoor-in-situ intelligent remodeling model driven by the normalized small-strain shear modulus of sand. The indoor-in-situ intelligent reshaping model consists of a normalized small-strain shear modulus prediction model and an intelligent optimization algorithm. The construction process is as follows: First, an artificial neural network is established. Then, the intelligent optimization algorithm performs back-optimization to obtain the optimal variable state parameters for indoor reshaping of sand. The indoor-in-situ intelligent reshaping model can take an outdoor normalized small-strain shear modulus as input. The intelligent optimization algorithm generates a series of candidate combinations of variable state parameters (water content, relative density, and vertical effective stress). Then, the normalized small-strain shear modulus prediction model predicts the corresponding normalized small-strain shear modulus based on this series of candidate combinations. The intelligent optimization algorithm is then called again to calculate the fitness (where fitness is the sum of the optimization objective function and the penalty constraint term), and the historical optimal is updated. It is then determined whether the termination criterion (i.e., convergence condition) is met. If the convergence condition is met, the optimal reshaping parameters are obtained. The optimization objective function is the absolute value of the difference between the predicted value of the normalized small-strain shear modulus prediction model and the corresponding outdoor normalized small-strain shear modulus.

[0072] In this embodiment, a BP neural network can be used as the artificial neural network, and the PSO algorithm can be used as the intelligent optimization algorithm. The optimization of three variable state parameters—moisture content, relative density, and vertical effective stress—is explained below. 1) Constructing artificial neural networks All feature values ​​(particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, moisture content, relative density) and target value (normalized small strain shear modulus) in the indoor-in-situ fusion dataset are normalized to obtain the normalized indoor-in-situ fusion dataset. This dataset is then mapped to the [0, 1] interval to eliminate the influence of dimensions. The normalization formula is as follows: ,in: Eigenvalues The normalized value To and The maximum value among similar eigenvalues To and The minimum value among similar characteristic values.

[0073] The normalized indoor-in-situ fusion dataset was divided into training and testing sets. The training set accounted for 75% and the testing set accounted for 25%.

[0074] The basic architecture of a BP neural network: The input layer of the BP neural network has seven nodes, representing particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content, and relative density. The output layer of the BP neural network has one node, namely the normalized small strain shear modulus. There is one hidden layer, and the number of nodes in the hidden layer is determined by an empirical formula. In the formula, Let m be the number of hidden layer nodes, m be the number of input layer nodes, and n be the number of output layer nodes. The constant is between 1 and 10. According to the formula, the number of hidden layer nodes is calculated to be between 5 and 14. In this embodiment, it is initially selected as 10, thus obtaining the basic architecture of the BP neural network.

[0075] Training the BP neural network: First, initial parameters are set by assigning random numbers to the weight matrices W and V, setting the sample pattern counter and training count counter to 1, setting the error E to 0, and setting the learning rate η to a random number between 0 and 1. The training set is then input into the BP neural network for training to obtain the normalized small strain shear modulus prediction value. The output error of each layer of neurons is calculated using the error function and backpropagated. Then, the weights and thresholds of each layer of neurons are adjusted according to the error function.

[0076] Once the error function is determined, the weights and thresholds of the output and hidden layers can be changed using the gradient descent method to set the error limit. ,when < If training is stopped at a certain point, it indicates that the training error has reached the required level and the training effect is good; if... ≥ Then continue training until... < So far, the trained BP neural network is obtained, which is the normalized small strain shear modulus prediction model.

[0077] 2) PSO algorithm The optimization objective function of the PSO algorithm

[0078]

[0079] in, The predicted value is from the BP neural network (unit: kPa). Let be the normalized small-strain shear modulus (unit: kPa) for outdoor conditions. The objective function represents the difference in dynamic stiffness between indoor remolded sand and outdoor undisturbed natural sand. The optimization direction is to find... Minimizing the dynamic stiffness allows the algorithm to automatically explore feasible solution sets in the parameter space that satisfy dynamic stiffness equivalence, achieving high-precision reproduction of soil dynamic response.

[0080] Choose the decision variable vector:

[0081] Wherein, vector x represents the three core control parameters of indoor remolded sand, and its component w represents the moisture content, with a constraint range of [0, 20%], which is set according to indoor test settings, covering unsaturated to saturated states; The relative density is constrained within the range of [10%, 90%], corresponding to a loose to a dense state; For vertical effective stress, the constraint range is [100, 800] kPa, covering the common foundation stress range, and the constraint basis is the physical limit of laboratory equipment and the feasible range of soil physical state; For vector transpose symbol, indicating column vector (actually row vector).

[0082] Particle swarm initialization is performed:

[0083] The formula represents the initial position of the i-th particle in the particle swarm optimization (PSO) algorithm , the first component represents the water content initialization, is the lower limit of water content (such as 0%), is the feasible range of water content (such as 20% - 0% = 20%), is a random number, making the initial water content uniformly distributed in [ ], is the upper limit of water content (such as 20%); the second component represents the relative density initialization, is the lower limit of relative density (such as 10%), is the range of relative density (such as 90% - 10% = 80%), is a random number, ensuring that the initial density is randomly distributed within the feasible range, is the upper limit of relative density (such as 90%); the third component represents the vertical effective stress initialization, is the lower limit of stress (such as 100 kPa), is the range of stress (such as 800 kPa - 100 kPa = 700 kPa), is a random number, generating a compliant initial stress value.

[0084] Through a random linear mapping, a set of feasible and diversified initial variable state parameter candidate combinations is generated for each particle, establishing a search starting point for global optimization.

[0085] Particle velocity initialization is performed according to the following formula:

[0086] where, is the initial velocity vector of the i-th particle in the PSO algorithm; is the velocity limiting coefficient (typical value 0.2), which constrains the velocity range to prevent oscillation and ensure the stability of particle motion; , , is a [0, 1] uniformly distributed random number, giving the particle a random exploration direction, and the random term makes the particle swarm uniformly cover the search direction, avoiding premature convergence.

[0087] The individual history optimal initialization is performed according to the following formula:

[0088] The above formula defines the initialization rule of the individual history optimal position of the i-th particle in the PSO algorithm, is the initial individual history optimal position of the particle i, is the initial position of the particle i. The formula defines the default value of the individual optimal solution in the PSO initialization, that is, the initial position of the particle itself is the best position it can find at the moment of generation. This is the starting point of the “memory of own history optimal” mechanism of the PSO algorithm.

[0089] Then, the normalized small-strain shear modulus prediction model is called, and the fitness calculation is performed according to the following formula:

[0090] is the fitness function value of the i-th particle, is the predicted normalized small-strain shear modulus of the i-th particle by the BP neural network. represents the penalty constraint term. is the penalty coefficient (typical value is 1000), which is used to amplify the violation cost and force the parameters within the safe range. is the k-th constraint function (a total of 6), which is respectively the six violation conditions of sample saturation liquefaction, undefined water content, compaction energy exceeding limit, loose collapse of sand sample, overload of universal testing machine, and inability to simulate stress state of shallow soil, is the variable state parameter vector of the i-th particle, is the constraint violation activation function.

[0091] The velocity update in the particle state update is performed:

[0092] The formula is the velocity update equation in the PSO algorithm, is the velocity vector of the particle i at the t-th iteration, is the particle motion inertia, is the velocity vector of the particle i at the t-1-th iteration, is the individual experience trust degree, is the group knowledge trust degree, and is a random disturbance factor, is the cognitive term (weight of moving to the individual history optimal position of the particle), is the social term (weight of moving to the global history optimal position of the group). Wherein is the position of particle i at the previous time step, is the individual historical optimal position (the best solution found so far by particle i), is the global historical optimal position (the best solution found so far by the whole swarm).

[0093] The position update in the particle state update is:

[0094] where, is the position of particle i at the tth iteration, is the position of particle i at the (t-1)th iteration, is the velocity vector of particle i at the tth iteration.

[0095] The individual historical optimal update and the global historical optimal update mechanisms in the PSO algorithm are respectively:

[0096]

[0097] where, in the first formula is the individual historical optimal position of particle i (the best solution found so far), is the position of particle i at the tth iteration; is the fitness value of particle i at the tth iteration (the smaller the better); is the fitness value of the historical optimal position, i.e., the first formula indicates that if the current new position has a lower fitness (for example, 5.2 kPa < 8.7 kPa), the current position covers the historical optimal ; In the second formula is the global historical optimal position (the best solution found so far by the whole swarm); is the parameter selected to minimize the function value; is the individual historical optimal position of particle i (the best solution found so far), i = 1, 2, …, N, and N is the total number of particles.

[0098] The first formula indicates the individual historical optimal update, i.e., the particle “self-learning” and accumulating local experience; the second formula indicates the global historical optimal update, i.e., the swarm “knowledge sharing” and avoiding local optimum.

[0099] After the historical optimal update, it is determined whether the termination criterion is met. If not, the particle swarm update is returned, and the normalized small-strain shear modulus prediction model is called. If yes, the optimal solution is output:

[0100] where, is the optimal variable state parameter; optimal water content, optimal relative density, optimal vertical effective stress; is the vector transpose symbol, indicating a column vector (actually a row vector).

[0101] The termination criterion is to terminate the optimization when any of the following conditions is met:

[0102] where t is the current iteration number, is the preset maximum iteration number, is the normalized small-strain shear modulus predicted by the global optimal parameter, i.e., the output of the BP neural network, is the normalized small-strain shear modulus of the in-situ target outdoor, is the modulus tolerance threshold, which is 100 kPa in this example. is the Euclidean distance between the particle position and the global optimum, i.e., representing the dispersion degree of the parameter space, represents the average distance of the particle swarm, is the convergence radius threshold, representing the percentage of the parameter range, which is 1% of the parameter range in this example.

[0103] The final model obtained based on artificial neural networks (an example is given by BP neural network) and intelligent optimization algorithms (an example is given by PSO algorithm) is called indoor-in-situ intelligent remodeling model. The indoor-in-situ intelligent remodeling model can input a normalized small-strain shear modulus, and the intelligent optimization algorithm starts to work, generating a series of candidate combinations of variable state parameters (water content, relative density, vertical effective stress). Then, the normalized small-strain shear modulus prediction model is called to predict the corresponding normalized small-strain shear modulus for the series of candidate combinations. Then, the intelligent optimization algorithm is called again to calculate the fitness, and the fitness value is fed back to the intelligent optimization algorithm to evaluate the pros and cons of the current candidate combination and update the particle position. When the penalty constraint satisfies the convergence condition, the optimal variable state parameter combination is obtained (see Figure 8 ).

[0104] Step five: access the intelligent remodeling data management platform The trained indoor-in-situ intelligent remolding model is burned into the intelligent remolding data management platform 004, which includes a data processing module, a database, a display module, and the indoor-in-situ intelligent remolding model; the database is physically connected to the data processing module and interacts with data through a standardized data interface; the normalized small-strain shear modulus data of the outdoor measured by the outdoor bending element device is imported into the database after being processed by the data preprocessing module, triggering the platform to call the indoor-in-situ intelligent remolding model, and obtaining the optimal combination of the variable state parameters of the indoor remolded sand soil.

[0105] Laboratory personnel prepare samples according to the corresponding sand soil gene of the optimal combination and carry out indoor bending element tests, and the obtained normalized small-strain shear modulus is fed back to the platform as a verification result and stored in the database for evaluating the performance of the model, and at the same time, the result can be used as a new data point to expand the indoor data set; in addition, the database contains the soil parameters of the outdoor measurement point area, and the current normalized small-strain shear modulus is associated with the soil parameters to form a new outdoor data point, expand the outdoor original data set, realize the dynamic update of the outdoor original data set, drive the model to continuously optimize, and form a closed-loop process of “data acquisition-model calculation-test verification-data feedback-model optimization”. The intelligent remolding data management platform 004 is developed with an OpenSees interface to realize seamless connection of “remolding parameters-numerical simulation”.

[0106] Embodiment 3: The sand soil remolding gene compiling system driven by dynamic characteristics (see Figure 6 ) in this embodiment includes: The indoor bending element device 005 acquires shear wave velocity data of standard sand under different working conditions; An electro-hydraulic servo universal testing machine is used to simulate vertical effective stress; The outdoor bending element device 006 acquires shear wave velocity data of outdoor original natural sand soil under different working conditions; The function signal generator 001 is used to apply a sine waveform with a set voltage and frequency to the transmitting bending element sensor; The function signal amplifier 002 is used to receive and amplify the electrical signal of the receiving bending element sensor; The oscilloscope 003 is used to input waveform parameters and simultaneously display output and received waves; The intelligent remolding data management platform 004 includes a data processing module, a database, a display module, and an indoor-in-situ intelligent remolding model; The indoor bending element device is a one-dimensional consolidation compression bending element device (see Figures 2-4 ), which includes a protective sleeve, and a rectangular experimental box sealed by four corrosion-resistant side plates and a bottom plate; Reserve holes for installing protective sleeve are arranged on the opposite pair of side plates, and pipe threaded holes are arranged on the remaining at least one side plate, and metal water permeable stones are filled in the pipe threaded holes; electromagnetic valves are arranged on the pipe threaded holes to control the drainage condition during the test; The protective sleeve comprises a protective sleeve upper cover and a protective sleeve main body, the protective sleeve upper cover and the protective sleeve main body are fixed together by bolts to form a protective sleeve with a front-end reserved packaging groove and a rear-end closed structure, the packaging groove is used for filling the curved element sensor, and the front end of the curved element sensor is exposed outside the packaging groove; the protective sleeve as a whole has a T shape, comprising a horizontal part and a vertical part, the vertical projection area of the horizontal part of the T shape is greater than and fully covers the vertical projection area of the vertical part, the vertical part of the protective sleeve is passed out of the experimental box outside the experimental box from the reserved hole of the experimental box, at this time, the horizontal part is attached to the inner side of the reserved hole, the horizontal part is fixed to the inner wall of the experimental box by bolts, and the vertical part is exposed outside the experimental box; Silica gel is filled between the packaging groove and the curved element sensor for sealing; The vertical part is provided with a plurality of height adjustment holes in the height direction for adjusting the area of the curved element sensor exposed at the front end of the protective sleeve; The two curved element sensors installed on the two side plates are one transmitting curved element sensor and the other receiving curved element sensor.

[0107] The outdoor curved element device (see Figure 5 ) comprises a horizontal fixed rod, two vertical fixed rods with tapered hammers symmetrically movably mounted on the horizontal fixed rod, a depth scale is engraved on the vertical fixed rod, and the lower part of each vertical fixed rod is provided with a curved element sensor, one of which is a receiving curved element sensor and the other is a transmitting curved element sensor; a protective iron sheet is arranged on the vertical fixed rod above and below the curved element sensor, and the curved element sensor is sealed from the vertical fixed rod by silica gel after being encapsulated by epoxy resin.

[0108] The function signal generator 001 is connected with the transmitting curved element sensor through one end of a BNC connecting line and connected with the oscilloscope 003 through the other end, the function signal amplifier 002 is connected with the receiving curved element sensor through one end of a BNC connecting line and connected with the oscilloscope 003 through the other end, and the transmitting sine wave and the receiving sine wave are displayed through the oscilloscope 003.

[0109] The intelligent remolding data management platform 004 is the core hub of the digital ecosystem, which mainly includes a data processing module, a database, a display module and an indoor-in-situ intelligent remolding model. The data processing module can obtain different kinds of standard sands under different water contents, different relative densities and different vertical effective stress controls and their corresponding normalized small strain shear moduli, and at the same time, it can obtain outdoor undisturbed natural sands at different measuring positions and different depths and the corresponding particle sizes, particle gradations, maximum pore ratios, minimum pore ratios, vertical effective stresses, water contents, relative densities and normalized small strain shear moduli; and it can arrange the data to obtain an indoor-in-situ fusion data set; The indoor-in-situ intelligent remolding model takes the normalized small strain shear modulus of the outdoor as the input, calls the artificial neural network to predict the corresponding normalized small strain shear modulus, and uses an intelligent optimization algorithm to drive to obtain the variable state parameters of the corresponding indoor remolding sand; The database is used to store the variable state parameters and inherent physical parameters of the standard sands and the outdoor undisturbed natural sands under different working conditions, and the normalized small strain shear moduli, and supports the calling of engineering applications; the database is connected with the data processing module, the indoor-in-situ intelligent remolding model and the display module through a standardized interface; the data processed by the data processing module is stored in the database, the indoor-in-situ intelligent remolding model can call the data in the database to dynamically update and iteratively optimize the model, and the display module supports the visual display of the parameters.

[0110] The unmentioned parts of the present application are applicable to the prior art.

Claims

1. A method for driving a sand soil remolding gene compilation method with power characteristics, characterized by, The compiling method Comprise the following contents: Through the indoor test, the normalized small strain shear modulus of the indoor remolded sand soil is obtained by correcting the small strain shear modulus of the indoor remolded sand soil under different vertical effective stresses, water contents and relative densities while the particle size, particle size distribution, maximum void ratio and minimum void ratio of the standard sand are unchanged; the indoor normalized small strain shear modulus is corresponded to the particle size, particle size distribution, maximum void ratio, minimum void ratio of the standard sand and the corresponding vertical effective stress, water content and relative density to form an indoor data set; Through the outdoor test, the vertical effective stress and the small strain shear modulus of the outdoor undisturbed natural sand soil under different buried depths are obtained, and the particle size, particle size distribution, maximum void ratio, minimum void ratio, water content and relative density of the outdoor undisturbed natural sand soil are obtained by sampling and testing in the laboratory; the normalized small strain shear modulus of the outdoor undisturbed natural sand soil is obtained by correcting the small strain shear modulus of the outdoor undisturbed natural sand soil; the normalized small strain shear modulus is corresponded to the particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content and relative density to form an outdoor undisturbed data set; The indoor data set and the outdoor undisturbed data set are fused to form an indoor-undisturbed fusion data set; The particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content and relative density are taken as input characteristics, and the normalized small strain shear modulus is taken as target output; an artificial neural network is constructed, and the indoor-undisturbed fusion data set is used to train the artificial neural network to obtain a normalized small strain shear modulus prediction model; An indoor-undisturbed intelligent remolding model based on the normalized small strain shear modulus of the sand soil is constructed by using the normalized small strain shear modulus prediction model and an intelligent optimization algorithm; the indoor-undisturbed intelligent remolding model takes the normalized small strain shear modulus of the outdoor as input to obtain the optimal variable state parameters of the indoor remolded sand soil; wherein the optimization objective function of the indoor-undisturbed intelligent remolding model is the absolute value of the difference between the predicted value of the normalized small strain shear modulus prediction model and the normalized small strain shear modulus of the outdoor corresponding to the input; The variable state parameters are at least one of the vertical effective stress, water content and relative density.

2. The method of claim 1, wherein the power characteristics are driven by a sand soil remolding genetic code. The artificial neural network is at least one of a BP neural network, an ANN neural network or an RNN neural network, and the intelligent optimization algorithm is at least one of a PSO algorithm, a GWO algorithm or a GA algorithm.

3. The method of claim 1, wherein the power characteristics are driven by a sand soil reshaping gene. The construction process of the indoor-undisturbed intelligent remolding model is: The normalized small strain shear modulus of the outdoor undisturbed natural sand soil is input, which is recorded as a target value; and the constraint range of each variable state parameter is set, and the variable state parameter candidate combination is generated by random linear mapping within the constraint range; Then, the normalized small strain shear modulus prediction model is called, and the corresponding normalized small strain shear modulus is predicted by using the variable state parameter candidate combination, which is recorded as a predicted value; The fitness is calculated by summing the absolute value of the difference between the predicted value and the target value and the penalty constraint term; An iterative process is performed to obtain an optimal variable state parameter with a minimum absolute value of the difference between the predicted value and the target value; The penalty constraint term includes six constraint violation activation functions, and the six constraint violation activation functions correspond to six violation conditions, i.e., sample saturation liquefaction, undefined water content, compaction energy overrun, loose collapse of sand sample, universal testing machine overload, and inability to simulate the stress state of shallow soil.

4. A sand remolded genetic code compiling system driven by power characteristics, characterized by, The system comprises: An indoor flexural element device is arranged to obtain shear wave velocity data of standard sand under different working conditions; An electro-hydraulic servo universal testing machine is arranged to simulate vertical effective stress; An outdoor flexural element device is arranged to obtain shear wave velocity data of outdoor undisturbed natural sand under different working conditions; A function signal generator is arranged to apply a set voltage and frequency sine wave to the transmitting flexural element sensor; A function signal amplifier is arranged to receive and amplify the electrical signal of the receiving flexural element sensor; An oscilloscope is arranged to input waveform parameters and simultaneously display output and received waves; An intelligent remolding data management platform comprises a data processing module, a database, a display module and an indoor-in-situ intelligent remolding model; The data processing module is arranged to obtain different kinds of standard sand and corresponding normalized small-strain shear modulus under different water contents, different relative densities and different vertical effective stress controls, and simultaneously obtain outdoor undisturbed natural sand at different measurement positions and different depths as well as corresponding particle size, particle size distribution, maximum void ratio, minimum void ratio, vertical effective stress, water content, relative density and normalized small-strain shear modulus; and to arrange data to obtain indoor-in-situ fusion data sets; The indoor-in-situ intelligent remolding model is arranged to take the normalized small-strain shear modulus of the outdoor as input, initialize to generate candidate combinations of variable state parameters within the constraint range of the variable state parameters, call an artificial neural network to predict corresponding normalized small-strain shear modulus of the candidate combinations of variable state parameters, and drive to obtain the variable state parameters of the corresponding indoor remolding sand soil by using an intelligent optimization algorithm; The database is arranged to store the variable state parameters and inherent physical parameters of the standard sand and the outdoor undisturbed natural sand under different working conditions, and the normalized small-strain shear modulus, and support engineering application calling; the database communicates with the data processing module, the indoor-in-situ intelligent remolding model and the display module through a standardized interface; the data processed by the data processing module is stored in the database, the indoor-in-situ intelligent remolding model can call the data in the database to dynamically update and iteratively optimize the indoor-in-situ intelligent remolding model, and the display module supports visual display of the parameters.

5. The power profile driven sand soil reshaping genetic compilation system of claim 4, wherein, The indoor flexural element device comprises a protective sleeve, and a cuboid experimental box sealed by four anticorrosion-treated side plates and a bottom plate; A reserved hole for mounting the protective sleeve is arranged on the opposite pair of side plates, and a pipe threaded hole is arranged on the remaining at least one side plate, the pipe threaded hole is filled with metal water-permeable stones, and an electromagnetic valve for controlling drainage conditions during the test is arranged on the pipe threaded hole; The protective sleeve comprises a protective sleeve upper cover and a protective sleeve main body, the protective sleeve upper cover and the protective sleeve main body are fixed together through bolts to form a protective sleeve with a front-end reserved packaging groove and a rear-end closure, the packaging groove is used for loading the bending element sensor, and the front end of the bending element sensor is exposed outside the packaging groove; the protective sleeve is in a T shape as a whole, comprising a horizontal part and a vertical part, the vertical part of the protective sleeve is passed out of the experimental box through a reserved hole of the experimental box, at this time, the horizontal part is attached to the inner side of the reserved hole, the horizontal part is fixed to the inner wall of the experimental box through bolts, and the vertical part is exposed outside the experimental box; Silica gel is filled between the packaging groove and the bending element sensor to seal; The vertical part is provided with a plurality of number adjusting holes in the height direction, which are used for adjusting the area of the bending element sensor exposed to the front end of the protective sleeve; Two bending element sensors installed on the two side plates are a transmitting bending element sensor and a receiving bending element sensor; The outdoor bending element device comprises a horizontal fixing rod, two vertical fixing rods with tapering hammers symmetrically movably installed on the horizontal fixing rod, a depth scale is engraved on the vertical fixing rod, the lower parts of the two vertical fixing rods are provided with bending element sensors, one of the bending element sensors is a receiving bending element sensor, and the other bending element sensor is a transmitting bending element sensor; protective iron sheets are arranged on the vertical fixing rods above and below the bending element sensors, and the bending element sensors are sealed from the vertical fixing rods through silica gel after being packaged by epoxy resin.

6. The power profile driven sand soil reshaping genetic compilation system of claim 5, wherein, Heavy-duty footings are arranged at the bottom of the experimental box.

Images