Expressway reconstruction and expansion composite foundation optimization method based on intelligent monitoring feedback

Abstract

A highway reconstruction and expansion composite foundation optimization method based on intelligent monitoring feedback, comprising: laying a monitoring network in the rigid boundary area to obtain foundation response data; identifying negative friction zones and extracting characteristic indicators based on the data to form a stress discontinuity characteristic matrix, while establishing a boundary parameter set; inputting the characteristic matrix and parameter set into the model, introducing a stiffness correction layer and a friction transition layer, dividing the rigid zone, transition zone and flexible zone through partition processing, and correcting the model using the measured data; generating control parameters such as pile spacing and pile length based on the corrected model; applying the parameters to the construction and continuously monitoring to form a closed-loop optimization until the settlement meets the standard. The method effectively solves the problem of negative friction zones in the rigid boundary area, realizes accurate prediction and control of the settlement mode, and ensures the safety and reliability of the highway reconstruction and expansion project.

Description

Technical Field

[0001] This invention belongs to the field of highway reconstruction and expansion, and more specifically, relates to a method for optimizing composite foundations for highway reconstruction and expansion based on intelligent monitoring feedback. Background Technology

[0002] In highway reconstruction and expansion projects, at rigid boundary locations such as bridge abutments or culvert openings, the stress transfer path of the composite foundation can form discontinuous areas due to different boundary constraints. The settlement characteristics of these areas differ significantly from those of free-field sections, especially behind bridge abutments where "negative skin friction zones" are prone to form, leading to discrepancies between the actual settlement pattern and conventional model predictions.

[0003] Currently, the main technical challenges in the practice of composite foundation engineering for highways are as follows:

[0004] Insufficient monitoring targeting: Existing monitoring systems mostly adopt a uniform distribution method, failing to optimize the layout for the stress concentration characteristics of rigid boundary areas. The monitoring point types are limited and the density is insufficient, making it difficult to accurately capture the stress abrupt change characteristics at the pile-soil interface, resulting in limited accuracy in identifying negative skin friction zones.

[0005] Poor model adaptability: Conventional foundation mechanics models are based on the assumption of a homogeneous and continuous medium, which cannot accurately describe the discontinuous stress behavior near rigid boundaries. The model parameters are mainly determined by empirical formulas, failing to fully consider the influence of boundary conditions such as abutment stiffness and fill compressibility, resulting in systematic deviations between the predicted results and actual monitoring data.

[0006] The control measures lack specificity: the control parameters generated by existing design methods are mostly static empirical values, which cannot be dynamically optimized based on measured data. There is a lack of differentiated treatment strategies for regions with significant negative friction and regions where friction changes in the opposite direction, making precise control difficult.

[0007] Lack of closed-loop control mechanism: Existing technologies lack a complete closed-loop system from monitoring and identification to regulation and correction, making it impossible to achieve dynamic optimization based on real-time monitoring data. When stress mutations or new negative friction trends persist after construction, it is difficult to adjust engineering measures in a timely manner, affecting project quality and long-term safety.

[0008] Therefore, there is an urgent need in this field for a new technical solution that can effectively solve the problem of stress discontinuity in rigid boundary regions, so as to overcome the above-mentioned defects of existing technologies and improve the safety and reliability of highway reconstruction and expansion projects. Summary of the Invention

[0009] To address the shortcomings of existing technologies, the present invention aims to overcome the aforementioned deficiencies and propose a method for optimizing composite foundations for highway reconstruction and expansion based on intelligent monitoring feedback.

[0010] The present invention adopts the following technical solution.

[0011] The first aspect of this invention discloses a method for optimizing composite foundations for highway reconstruction and expansion based on intelligent monitoring feedback, the method comprising:

[0012] Step 1: Deploy monitoring units in the bridge approach slab and culvert opening areas to form a multi-source monitoring network, acquire and output basic response datasets;

[0013] Step 2: Based on the basic response dataset, identify the negative skin friction zone region and extract feature indices to form a stress discontinuity feature matrix; at the same time, establish the correspondence between abutment stiffness, fill compressibility and pile spacing, and output the boundary parameter set.

[0014] Step 3: Input the stress discontinuity feature matrix and boundary parameter set into the foundation mechanics response model, introduce the boundary stiffness correction layer and the friction reverse transition layer, divide the pile-soil system into rigid zone, transition zone and flexible zone through zoning, use measured data for correction, and output the corrected zoning response model and final parameter set.

[0015] Step 4: Using the corrected partitioned response model and the final parameter set as input, generate control parameters, and provide feedback on the generation of control parameters and differentiated optimization design to output the total set of control parameters;

[0016] Step 5: Use the set of control parameters as input, continue monitoring after structural improvement, and end the closed loop when the pile-soil friction curve becomes continuous and the settlement difference is less than the threshold; otherwise, re-input the corrected data into Step 3 for iteration.

[0017] Step 1 includes:

[0018] Based on the basic design data of the bridge approach slab and culvert opening area, the influence range of the rigid boundary is identified, thereby determining the monitoring control area. This area mainly covers the reasonable distance from the back of the bridge abutment to the roadbed and the specific depth range above the pile tip. The final output is a set of coordinates of the monitoring area.

[0019] Within the coordinate set of the monitoring area, the location of the monitoring unit is determined along the interface between the pile and the soil, and the output is a set of monitoring unit placement points.

[0020] Additional monitoring points are added to the pile tip area to form a supplementary monitoring point set. The monitoring unit monitoring point set and the supplementary monitoring point set are then merged to obtain a complete monitoring point set.

[0021] All monitoring units in the complete set of monitoring points are connected to a unified data acquisition system. Dynamic data acquisition is carried out by setting a basic sampling frequency and combining it with the working condition adjustment coefficient to form a preliminary data set.

[0022] The raw monitoring data in the preliminary dataset are calculated and organized to form the basic response dataset.

[0023] Step 2 includes:

[0024] The parameters in the basic response dataset were reorganized to form a deep sequence dataset. To eliminate the influence of noise, the pile side friction in the basic response dataset was smoothed to obtain the smoothed friction value.

[0025] Based on the depth sequence dataset, the frictional change in the depth direction is calculated. According to the smoothed frictional value and the reasonable threshold range of the frictional change, the negative frictional trend and stress mutation point are determined, thereby screening out the depth intervals where negative frictional occurs continuously and forming a set of negative frictional regions.

[0026] By utilizing the depth segments corresponding to the set of negative friction regions, the depth points that meet the conditions and their corresponding features are combined to form a set of feature vectors.

[0027] The set of feature vectors is integrated into a matrix form to construct a stress discontinuous feature matrix, where each row of the matrix represents a depth feature point.

[0028] Based on design data and construction records, the corresponding relationship between abutment stiffness, fill compressibility and pile spacing is established and output, and the set of boundary parameters is output.

[0029] Step 3 includes:

[0030] The boundary stiffness correction coefficients at each depth point are calculated based on the stress discontinuity characteristic matrix and the parameters in the boundary parameter set, so as to obtain the boundary stiffness correction coefficient sequence.

[0031] The transition points where the friction direction changes significantly are identified by using the boundary stiffness correction coefficient sequence and the basic response dataset, so as to obtain the friction reverse transition parameter sequence.

[0032] Based on the boundary stiffness correction coefficient sequence and the friction reverse transition parameter sequence, the pile-soil system is divided into rigid zone, transition zone and flexible zone, forming a set of partition results. Initial pile-soil stress transfer curve parameters are defined for each partition, forming an initial stress transfer curve parameter set.

[0033] Step 3 also includes:

[0034] Based on the partitioned result set and the initial stress transfer curve parameter set, a foundation mechanical response model is constructed. A friction prediction function is defined through the partitioning function, and the prediction results are compared with the measured friction in the foundation response dataset. The residual function and objective function are defined to form the foundation mechanical response model.

[0035] Based on the foundation mechanical response model, an iterative algorithm is used to dynamically optimize the stress transfer curve parameters of each zone. By continuously updating the parameters and recalculating the objective function, the preset convergence conditions are met, and finally the corrected zone response model and the corresponding final parameter set are output.

[0036] Step 4 includes:

[0037] Based on the corrected partition response model and the final parameter set, the difference between the measured friction and the friction prediction function in each partition is calculated, the stress imbalance index at each depth point is defined, and the average stress imbalance of each partition is further calculated to form a partition stress imbalance index set.

[0038] By combining the set of zoning results and the set of zoning stress imbalance indexes, a negative skin friction significant area is identified by setting a negative skin friction criterion. The criterion includes that the pile side skin friction is negative and the stress imbalance index exceeds a specific threshold. The continuous depth intervals that meet the conditions are sorted and output as a set of negative skin friction significant areas.

[0039] For the set of areas with significant negative skin friction, suggestions for adjusting pile spacing and pile length are generated based on the magnitude of the stress imbalance index, in order to generate a set of adjustment parameters for pile spacing and pile length.

[0040] Identify sections where frictional resistance changes slowly in the opposite direction. These sections are located in the flexible zone. For these sections, define the grouting density increase ratio and the filler modulus increase value to generate a set of optimized grouting and filler parameters.

[0041] The set of parameters for adjusting pile spacing and pile length, as well as the set of parameters for optimizing grouting and filling, are integrated to form the final set of control parameters. This set includes the amount of pile spacing reduction, the amount of pile end lengthening, the proportion of grouting density increase, and the value of filling modulus increase.

[0042] Step 5 includes:

[0043] The on-site construction process is adjusted based on the set of control parameters. The adjusted construction parameters are calculated based on the amount of pile spacing reduction, pile end lengthening, grouting density increase ratio, and filler modulus increase value. These parameters include new pile spacing, new pile length, new grouting density, and new filler modulus, forming a set of construction parameter execution.

[0044] After completing the construction adjustment based on the construction parameters, the monitoring network continues to collect data to obtain the pile side friction, lateral displacement difference and pore pressure after the construction adjustment, forming an updated monitoring dataset;

[0045] The effect is evaluated based on the updated monitoring dataset. The friction change range at each depth point is calculated as the friction continuity index, and the settlement difference between the settlement value on the back of the bridge abutment and the settlement value of the main roadbed is calculated to form a set of evaluation indicators.

[0046] The closed-loop determination is based on the set of evaluation indicators. The determination conditions are set, requiring that the maximum value of the friction continuity index is less than the friction continuity threshold and the settlement difference is less than the settlement difference limit. When both of the above conditions are met at the same time, the closed-loop status flag is set to 1 and the closed loop ends. Otherwise, the closed-loop status flag is set to 0 and the corrected data is re-entered into step 3 for iteration.

[0047] The second aspect of this invention discloses a composite foundation optimization system for highway reconstruction and expansion based on intelligent monitoring feedback, used to implement the method described above, the system comprising:

[0048] The monitoring network deployment module is used to deploy monitoring units in the bridge approach slab and culvert opening areas to form a multi-source monitoring network, acquire and output basic response datasets;

[0049] The stress anomaly identification module is used to identify negative skin friction zone regions and extract feature indicators based on the basic response dataset, forming a stress discontinuity feature matrix; at the same time, it establishes the correspondence between abutment stiffness, fill compressibility and pile spacing, and outputs a set of boundary parameters.

[0050] The rigid-flexible transition model module is used to input the stress discontinuity feature matrix and boundary parameter set into the foundation mechanics response model, introduce the boundary stiffness correction layer and the friction reverse transition layer, divide the pile-soil system into rigid zone, transition zone and flexible zone through zoning, use measured data for correction, and output the corrected zoning response model and final parameter set.

[0051] The regulation parameter generation module is used to take the corrected partition response model and the final parameter set as input to generate regulation parameters, and to provide feedback on the generation of regulation parameters and differentiated optimization design, so as to output the total set of regulation parameters.

[0052] The closed-loop control module takes the set of control parameters as input, and continues to monitor after the structural improvement is completed. When the pile-soil friction curve tends to be continuous and the settlement difference is less than the threshold, the closed loop ends; otherwise, the corrected data is re-inputted into step 3 for iteration.

[0053] A third aspect of the present invention discloses a terminal, including a processor and a storage medium; the storage medium is used to store instructions; the processor is used to perform operations according to the instructions to execute the steps of the method.

[0054] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method.

[0055] The beneficial effects of the present invention are as follows: Compared with the prior art, the present invention has the following advantages:

[0056] This invention utilizes targeted multi-source monitoring units deployed in rigid boundary areas such as bridge abutment slabs and culvert openings. Instead of the traditional uniform distribution method, it employs a high-density deployment along the pile-soil interface and at key depths of the pile tip, forming a complete monitoring network capable of capturing stress abrupt changes, reverse frictional changes, and the development of negative frictional zones. Combining the stress discontinuity feature matrix, boundary stiffness correction coefficient sequence, and reverse frictional transition parameter sequence constructed in steps 2 and 3, this invention achieves, for the first time in engineering, the quantitative extraction and structured modeling of stress discontinuity behavior near rigid boundaries. Unlike existing foundation models based on the assumption of a continuous medium, this method uses a three-segment partitioning approach—"rigid zone – transition zone – flexible zone"—allowing boundary conditions such as abutment stiffness, filler compressibility, and pile spacing variations to be clearly, adjustably, and correctably integrated into the mechanical model. The resulting corrected partitioned response model resolves the systematic bias of traditional models in accurately predicting negative frictional resistance, reverse frictional resistance, and settlement gradients, achieving a true depiction of the stress transmission path. This ensures high consistency between prediction results and monitoring data, significantly improving prediction capability and model adaptability. Attached Figure Description

[0057] Figure 1 This is a flowchart of a method for optimizing composite foundations for highway reconstruction and expansion based on intelligent monitoring feedback, according to an embodiment of the present invention. Detailed Implementation

[0058] The present application will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, and should not be construed as limiting the scope of protection of the present application.

[0059] like Figure 1 As shown, in one embodiment, the present invention discloses a method for optimizing composite foundations for highway reconstruction and expansion based on intelligent monitoring feedback, the method comprising:

[0060] Step 1: Deploy monitoring units in the bridge approach slab and culvert opening areas to form a multi-source monitoring network, acquire and output basic response datasets;

[0061] Step 2: Based on the basic response dataset, identify the negative skin friction zone region and extract feature indices to form a stress discontinuity feature matrix; at the same time, establish the correspondence between abutment stiffness, fill compressibility and pile spacing, and output the boundary parameter set.

[0062] Step 3: Input the stress discontinuity feature matrix and boundary parameter set into the foundation mechanics response model, introduce the boundary stiffness correction layer and the friction reverse transition layer, divide the pile-soil system into rigid zone, transition zone and flexible zone through zoning, use measured data for correction, and output the corrected zoning response model and final parameter set.

[0063] Step 4: Using the corrected partitioned response model and the final parameter set as input, generate control parameters, and provide feedback on the generation of control parameters and differentiated optimization design to output the total set of control parameters;

[0064] Step 5: Use the set of control parameters as input, continue monitoring after structural improvement, and end the closed loop when the pile-soil friction curve becomes continuous and the settlement difference is less than the threshold; otherwise, re-input the corrected data into Step 3 for iteration.

[0065] In a specific embodiment, the present invention provides a method for optimizing the composite foundation of highway reconstruction and expansion based on intelligent monitoring feedback, comprising the following steps:

[0066] Step 1: Deployment of monitoring units in rigid boundary areas and acquisition of foundation response data. For example, multiple types of monitoring units, including strain gauges, lateral displacement gauges, and pore pressure gauges, are deployed in the bridge approach slab and culvert opening areas. These units are distributed along the interface between the pile and the soil, the bottom of the pile tip, and the back of the abutment, forming a multi-source monitoring network that reflects the collaborative stress state of the pile-soil-abutment system.

[0067] Sub-step 1.1: Identification of the influence range of rigid boundaries and determination of the monitoring deployment area.

[0068] Based on the design data, the boundary of the bridge approach slab and culvert opening area was identified. The area from the back of the abutment to the roadbed in the direction of 2 to 6 meters and from the pile tip to 0.5 to 2 meters was determined as the monitoring control area. This clarified the monitoring deployment range, ensuring that the subsequent deployment points corresponded to the stress anomaly area and improving the monitoring effectiveness.

[0069] The coordinate set C of the monitoring area is represented as:

[0070] C = {(x1,y1),(x2,y2),…,(xn,yn)};

[0071] Where xi and yi represent the horizontal and vertical coordinates of the layout point on the plane, respectively, both derived from the construction drawing measurement data.

[0072] Sub-step 1.2: Select the monitoring unit type and determine the location of the pile side installation.

[0073] Strain gauges, lateral displacement gauges, and pore pressure gauges are deployed along the interface between the pile and the soil within the coordinate set C of the monitoring area, thus forming a monitoring deployment system along the depth direction, which can capture the changing trends of pile side friction and lateral deformation. The set P of monitoring unit deployment points is represented as follows:

[0074] P = {p1,p2,...,pm};

[0075] Where pi represents the i-th monitoring point, which includes the depth value hi and the monitoring type ti. The value of hi ranges from 0.5 meters to 15 meters (determined according to the pile length), and ti represents the value of the strain gauge when i is 1, the value of the lateral displacement gauge when i is 2, and the value of the pore pressure gauge when i is 3. The spacing between monitoring points adopts a fixed depth interval formula:

[0076] h(k) = h0 + k*Δh;

[0077] Where h(k) is the depth of the kth placement point, h0 is the reference value of the pile top depth (generally taken as 0), and Δh is the placement point spacing (0.5 meters to 1 meter).

[0078] Sub-step 1.3: Supplement the pile end and the back of the abutment with additional piles.

[0079] Additional points are added to the pile tip area to form set E:

[0080] E = {e1,e2,...,eq};

[0081] Where ei includes the depth value hi (ranging from 0.5 meters to 15 meters, determined according to the pile length) and the monitoring type ti (strain gauge or pore pressure gauge). The monitoring unit deployment point set P and E are merged to obtain the complete monitoring point set S:

[0082] S = P∪E.

[0083] Meanwhile, pore pressure gauges are added along the height direction in the filling area on the back of the bridge abutment, and their placement is consistent with the layer thickness of the filling.

[0084] Sub-step 1.4: Monitoring network connectivity and data acquisition system setup.

[0085] All monitoring points in the complete monitoring point set S are connected to the data acquisition system, and data is read through a unified acquisition module. The sampling frequency is calculated using the formula:

[0086] f = f0*k;

[0087] Where f is the sampling frequency, f0 is the basic sampling frequency (once every 10 minutes), and k is the working condition adjustment coefficient (1 to 6). When k ≥ 3 is taken during the construction loading stage, the collected data includes strain ε, lateral displacement u, and pore pressure p, forming a preliminary data set D:

[0088] D = {(εi,ui,pi)};

[0089] Where εi, ui, and pi correspond to the measured values ​​of the i-th monitoring point, respectively.

[0090] Sub-step 1.5: Calculation and organization of basic response data.

[0091] The basic response dataset R is obtained from the initial dataset D. First, the pile side friction τ is inverted based on the strain data using the following formula:

[0092] τ = E*ε;

[0093] Where τ is the pile side friction, E is the material elastic modulus (range 20GPa to 35GPa, derived from pile material testing), and ε is the measured strain. Lateral displacement gradient calculation formula:

[0094] du = u(i) - u(i-1).

[0095] The pore pressure changes were processed using time series analysis to form the basic response dataset R:

[0096] R = {τ,du,p}.

[0097] Step 2: Boundary stress anomaly identification and discontinuity feature extraction. Based on the monitoring sequence from Step 1, the distribution trend of pile side friction along the depth direction is calculated, and the "negative friction zone" region appearing on the back of the abutment and both sides of the culvert is identified. Feature indicators including stress mutation depth, pile bending moment change rate, and soil lateral displacement mutation points are extracted to form a "stress discontinuity feature matrix". Simultaneously, the correspondence between abutment stiffness, fill compressibility, and pile spacing is established for subsequent model correction input.

[0098] Sub-step 2.1: Deep sequence construction and basic data preparation.

[0099] The pile side friction τ, lateral displacement difference du, and pore pressure p in the basic response dataset R are recombined according to the monitoring depth h to form a depth sequence. The depth sequence dataset S2 is represented in the following form:

[0100] S2 = {(h(i),τ(i),du(i),p(i))};

[0101] Where h(i) is the depth of the i-th monitoring point, derived from the deployment depth record; τ(i) is the pile side friction at the corresponding depth; du(i) is the lateral displacement gradient; and p(i) is the pore pressure value. To eliminate the influence of noise, τ(i) is smoothed using a moving average formula:

[0102] τs(i)=( τ(i-1) + τ(i) + τ(i+1)) / 3;

[0103] Where τs(i) is the friction value after smoothing, and finally forms the deep sequence dataset S2.

[0104] Sub-step 2.2: Calculation of pile side friction distribution trend and identification of negative friction.

[0105] The rate of change of friction in the depth direction is calculated based on τs(i) in the depth sequence dataset S2, as shown in the following formula:

[0106] dτ(i) = τs(i) - τs(i-1);

[0107] Where dτ(i) represents the change in frictional resistance. When τs(i) is in the range of 0.0 kPa to -20 kPa, it is considered a negative frictional resistance trend; when the decrease in dτ(i) is in the range of -5 kPa to -15 kPa, it is considered a stress abrupt change point. The change in frictional resistance is used to characterize the negative frictional resistance zone generated by the boundary constraints on the back of the abutment and both sides of the culvert. The depth range is selected based on the judgment criteria to form the set of negative frictional resistance regions Z:

[0108] Z = {(h(a),h(b))};

[0109] Where (h(a),h(b)) represents the depth range in which negative friction occurs continuously.

[0110] Sub-step 2.3: Extraction of abrupt changes in bending moment rate and lateral displacement characteristics.

[0111] Using the depth sequence dataset S2 and the negative skin friction region set Z, a feature vector set F is obtained. First, using the depth segment corresponding to the negative skin friction region Z, the locations of abrupt changes in bending moment rate and lateral displacement are extracted. The bending moment rate is estimated using the following formula:

[0112] M(i) = τs(i)*r;

[0113] Where M(i) is the bending moment value, and r is the pile radius (ranging from 0.2 meters to 0.6 meters). The expression for the rate of change of bending moment is as follows:

[0114] dM(i) = M(i) - M(i-1);

[0115] The threshold for lateral displacement abrupt change is determined using:

[0116] du(i) > Tdu;

[0117] Where Tdu is the threshold for abrupt lateral displacement change, typically taken as 2 to 5 millimeters. The depth points that meet the conditions are combined with their corresponding rates of change of bending moment to form a set of feature vectors F:

[0118] F = {(h(i),dM(i),du(i))}.

[0119] Sub-step 2.3 is used to realize the location of abrupt changes in the stress and lateral displacement of the pile under the influence of the quantified boundary rigidity, and is the core data source for constructing the stress discontinuity feature matrix.

[0120] Sub-step 2.4: Construction of the stress discontinuity characteristic matrix.

[0121] The eigenvector set F is integrated into a stress discontinuous eigenma matrix M2:

[0122] M2 = ;

[0123] Each row of the matrix represents a depth feature point, where h(n) is the depth, dM(n) is the rate of change of bending moment, and du(n) is the abrupt change in lateral displacement. All elements in the matrix are derived from F and do not contain any inferred parameters.

[0124] Sub-step 2.5: Establish the correspondence between stiffness, compressibility, and pile spacing.

[0125] The boundary parameter set B is obtained by using the characteristic matrix M2, as well as the abutment stiffness K, fill compressibility C, and pile spacing L, which can be directly obtained from the design data.

[0126] The relationship between abutment stiffness, fill compressibility, and pile spacing is established as follows:

[0127] B = {(K,C,L)};

[0128] Where K ranges from 100MN / m to 500MN / m; C ranges from 0.02 to 0.15; and L ranges from 1.0 m to 3.0 m, the data are derived from design documents and construction records. This set is used to describe the structural characteristics of the boundary zone and provides input conditions for the zoning model.

[0129] Step 3: Establishment and Dynamic Correction of the Rigid-Flexible Transition Response Model. The discontinuous characteristic matrix and boundary parameter set output from Step 2 are input into the foundation mechanics response model. A "boundary stiffness correction layer" and a "friction-reverse transition layer" are introduced into the model. Through zonal equivalence processing, the pile-soil system under the abutment is divided into rigid, transition, and flexible zones, and independent pile-soil stress transfer curves are defined for each zone. The zonal stiffness and friction rate of change of the model are corrected using measured data until the pile side friction distribution output by the model matches the measured trend.

[0130] Sub-step 3.1: Calculate the boundary stiffness correction coefficient based on the discontinuous characteristic matrix and the boundary parameter set.

[0131] The boundary stiffness correction coefficient sequence Kc is obtained by using the stress discontinuity characteristic matrix M2 and the boundary parameter set B.

[0132] The stress discontinuity characteristic matrix M2 is:

[0133] M2 = ;

[0134] Where h(i) is the depth of the i-th feature point in meters; dM(i) is the rate of change of bending moment at the corresponding depth in kilonewton-meters; du(i) is the abrupt change in lateral displacement at the corresponding depth in millimeters; and n is the number of feature points.

[0135] The boundary parameter set B is represented as:

[0136] B = (K,C,L);

[0137] Wherein, K is the overall stiffness of the abutment, in meganewtons per meter, ranging from 100 to 500; C is the compressibility coefficient of the fill material, a dimensionless coefficient ranging from 0.02 to 0.15, obtained through indoor compression tests; L is the pile spacing, in meters, ranging from 1.0 to 3.0, obtained from the design layout drawings.

[0138] First, calculate the discontinuous composite index q(i) using the following formula:

[0139] q(i) = |dM(i)| / dMmax + du(i)| / dumax;

[0140] Where q(i) is the discontinuous intensity index at point i, which is dimensionless; dMmax is the maximum value of |dM(i)| among all i, in kilonewton-meters; and dumax is the maximum value of |du(i)| among all i, in millimeters. dMmax and dumax are obtained by directly calculating the maximum value from the data in M2.

[0141] Then, based on q(i), abutment stiffness K, fill compressibility C, depth h(i), and pile spacing L, a boundary stiffness correction coefficient Kc(i) is constructed. When the amount of data is small, tree models such as random forest and gradient boosting tree (e.g., XGBoost) can be used. When the amount of data is large, a fully connected neural network can be used, with q(i), K, C, h(i), and L as inputs and Kc(i) as output. Kc(i) is the equivalent corrected stiffness at depth h(i), in meganewtons per meter.

[0142] The sequence of boundary stiffness correction coefficients Kc for all depth points is:

[0143] Kc = {Kc(1),Kc(2),...,Kc(n)}.

[0144] By using the Kc sequence, the degree of stress discontinuity is organically coupled with the abutment stiffness and the compressibility of the filler, transforming the originally uniform boundary stiffness into a "boundary stiffness correction layer" parameter that varies with depth, providing a quantitative stiffness basis for the subsequent construction and partitioning of the friction-reverse transition layer.

[0145] It should be noted that the values ​​of M2 and B are input into the formulas q(i) and Kc(i). The output Kc will be used in sub-step 3.2 for constructing the friction reverse transition layer parameters, and in sub-step 3.3 for dividing the rigid region and the transition region.

[0146] Sub-step 3.2: Construct the parameters of the friction reverse transition layer and generate the deep friction response sequence.

[0147] The reverse transition parameter sequence Tf of the friction is obtained using the boundary stiffness correction coefficient sequence Kc and the foundation response dataset R. The foundation response dataset R contains the pile side friction τ(i) along the depth direction and the depth h(i), which can be summarized as follows:

[0148] Rτ ={(h(1),τ(1)), (h(2),τ(2)) ,..., (h(m),τ(m))};

[0149] Where τ(i) is the measured pile side friction at the i-th depth point, in kilopascals; m is the number of monitoring points.

[0150] First, interpolation matching is performed between the Kc sequence and Rτ at depth to unify Kc(i) and τ(i) to the same depth discrete point, resulting in a joint sequence:

[0151] J = {(h(1),τ(1), Kc(1)) ,..., (h(m),τ(m),Kc(m))}.

[0152] Then, the transition point from positive to negative or from negative to positive friction is identified through the joint sequence J, and the friction direction index s(i) is defined:

[0153] s(i) = sign(τ(i));

[0154] Where, sign is the sign function, which takes the value 1 when τ(i) is greater than 0, takes the value -1 when τ(i) is less than 0, and takes the value 0 when the absolute value of τ(i) is less than 1 kPa.

[0155] Frictional reverse transition strength t(i) is defined as:

[0156] t(i) = | τ(i) - τ(i-1) | * Kc(i);

[0157] Where t(i) is the reverse transition strength of friction at point i, in units of kilopascals multiplied by meganewtons per meter, which reflects both the magnitude of friction change and the local stiffness amplification effect.

[0158] Using logistic regression, support vector machine, or tree model, input the current point and its neighboring points τ(i), Kc(i), h(i) to output which depth points are friction-reverse transition points.

[0159] Finally, the t(i) and s(i) of all depth points are organized into a frictional reverse transition parameter sequence Tf:

[0160] Tf = {(h(1),t(1),s(1)),...,(h(m),t(m),s(m))}.

[0161] By combining frictional variation with corrected stiffness, a frictional reverse transition parameter sequence Tf is constructed to identify which depth segments are regions of significant frictional direction change, thus providing a quantitative basis for the subsequent division of rigid, transition, and flexible regions into frictional transition layers.

[0162] Sub-step 3.3 involves dividing the pile-soil system into zones based on the stiffness correction coefficient and friction transition parameters.

[0163] The partitioning result set Zone and the initial stress transfer curve parameter set T0 of each zone are obtained by using the boundary stiffness correction coefficient sequence Kc, the friction reverse transition parameter sequence Tf, and the boundary parameter set B. The depth ranges of the rigid zone, transition zone, and flexible zone are determined based on Kc and Tf. First, the stiffness normalization index k(i) is defined:

[0164] k(i) = Kc(i) / K;

[0165] Where k(i) is the relative stiffness of the i-th point, which is dimensionless; K is the overall stiffness of the abutment in the boundary parameter set B.

[0166] Using random forests, gradient boosting trees, or neural networks, input k(i), Tf, t(i), h(i), τ(i) to output whether each depth point belongs to a rigid region, a transition region, or a flexible region.

[0167] The depth ranges for each region are summarized as follows:

[0168] Zone = {Zr, Zt, Zf};

[0169] Where Zr is the set of depths in the rigid region, Zt is the set of depths in the transition region, and Zf is the set of depths in the flexible region. Each set consists of multiple depth intervals, such as (h(a1), h(b1)).

[0170] Based on this, initial pile-soil stress transfer curve parameters are defined for each zone. Taking the linear approximation of pile side friction with depth as an example:

[0171] Rigid region:

[0172] τr(h) = ar*h + br

[0173] Transition zone:

[0174] τt(h) = at*h + bt

[0175] Flexible area:

[0176] τf(h) = af*h + bf;

[0177] Wherein, τr(h), τt(h), and τf(h) are the initial friction transmission curves in the rigid, transition, and flexible regions, respectively, in kilopascals; h is the depth in meters; ar, at, and af are the friction gradients within the regions, in kilopascals per meter, obtained by least-squares fitting of the measured friction τ(h) within the regions; br, bt, and bf are the friction intercepts within the regions, in kilopascals, also obtained by fitting.

[0178] The initial stress transfer curve parameter set for each region is as follows:

[0179] T0 = ​​{(ar,br),(at,bt), (af,bf)}.

[0180] Sub-step 3.4: Construct a mechanical response model for a rigid-flexible transition foundation and generate initial values ​​for the model parameters.

[0181] The foundation mechanical response model G0 is obtained by using the partition result set Zone, the initial stress transfer curve parameter set T0, and the basic response dataset R.

[0182] Based on the Zone zoning results, the pile-soil system is treated as a segmented one-dimensional continuous medium, with corresponding stress transfer curves used for the rigid zone, transition zone, and flexible zone. The friction prediction function τm(h) is defined as:

[0183] When h belongs to Zr:

[0184] τm(h) = τr(h)

[0185] When h belongs to Zt:

[0186] τm(h) = τt(h)

[0187] When h belongs to Zf:

[0188] τm(h) = τf(h);

[0189] Among them, τr(h), τt(h), and τf(h) are defined by (ar,br), (at,bt), and (af,bf) in T0, respectively.

[0190] By comparing the above piecewise function with the measured friction τ(h) in the basic response dataset R, the residual function e(h) is defined as follows:

[0191] e(h) = τ(h) - τm(h)

[0192] To facilitate subsequent iterative corrections, an objective function J is constructed to measure the overall fitting error:

[0193] J = Σ ;

[0194] Where J is the objective function of error, and is a non-negative real number; the summation is performed at all depth points i.

[0195] The initial foundation mechanical response model G0 is constructed by combining the zone structure, piecewise function form, and parameters (ar, br, at, bt, af, bf), and the parameter combination is defined as follows:

[0196] G0 = (ar,br,at,bt,af,bf)

[0197] Based on the rigid-flexible zoning, the distribution of pile side friction is uniformly described by a piecewise linear model, and the objective function J provides a quantitative evaluation standard for subsequent parameter correction, forming an iterative rigid-flexible transition foundation mechanical response model framework.

[0198] Sub-step 3.5: Dynamic model correction and convergence of rigid-flexible transition response model based on measured data.

[0199] Using the foundation mechanics response model G0 and the foundation response dataset R, the corrected zonal response model G3 and the final parameter set O* are obtained. Based on G0, an iterative correction method is used to optimize the stiffness and friction rate of change in each zone. Taking gradient descent as an example, the parameter update formula for k in each iteration is:

[0200] ar(k+1) = ar(k) - η* ;

[0201] br(k+1) = br(k) - η* ;

[0202] at(k+1) = at(k) - η* ;

[0203] bt(k+1) = bt(k) - η* ;

[0204] af(k+1) = af(k) - η* ;

[0205] bf(k+1) = bf(k) - η* ;

[0206] Where ar(k), br(k), at(k), bt(k), af(k), and bf(k) are the parameter values ​​at the k-th iteration; η is the learning step size coefficient, ranging from 0.01 to 0.1, which is determined based on a combination of convergence speed and stability. The equal partial derivatives are obtained by numerical difference calculation of the objective function J.

[0207] In each iteration, τm(h) is reconstructed using the updated parameters, and the new residual e(h) and objective function J are calculated. The iteration terminates when the change in J satisfies one of the following convergence conditions:

[0208] First, the objective function J is less than a set threshold Jmin, where Jmin ranges from 1* Up to 1* ;

[0209] Second, the difference between two adjacent iterations of the objective function is less than the set tolerance ε, which ranges from 1 to 10.

[0210] After convergence, the final parameter set is obtained:

[0211] O* = ( ar*,br*,at*,bt*,af*,bf* )

[0212] The corresponding friction prediction function τm(h) and the zone structure Zone together constitute the corrected zoned response model G3. At this time, the pile side friction distribution output by the model basically matches the measured friction trend in the foundation response dataset R within each zone.

[0213] Through iterative optimization, the pile-soil stress transfer curves in the rigid zone, transition zone, and flexible zone fully absorb the measured information, forming a foundation mechanics response model that can truly reflect the rigid-flexible transition behavior, providing a reliable basis for the generation of control parameters in the subsequent step 4, which includes the zonal stiffness and friction rate of change.

[0214] Step 4: Feedback Control Parameter Generation and Differentiated Optimization Design. Using the revised zonal response model from Step 3 as feedback input, control parameters for the composite foundation in the abutment area are generated in real time, including adjustment suggestions for pile spacing, pile length, pile diameter, and composite ratio. For areas with significant negative skin friction, stress transfer is balanced by reducing pile spacing, adding transition layer pile groups, or increasing the depth of the bearing layer at the pile tip. For areas where the reverse change in skin friction is slow, the pile tip grouting density and the modulus of the transition filler are optimized to smoothly connect with the settlement curve of the flexible subgrade. This feedback result is transmitted back to the construction and monitoring system in real time, forming an executable parameter adjustment plan.

[0215] Sub-step 4.1: Calculation of stress imbalance in different zones and identification of control requirements.

[0216] By correcting the partition response model G3 and the final parameter set, the partition stress imbalance index set U is obtained. Based on the friction prediction function τm(h) of each depth segment in G3, and combined with the final fitting parameters (ar*,br*,at*,bt*,af*,bf*) in O*, the difference between the measured friction and the predicted friction in each partition is calculated. The stress imbalance index u(i) is defined as follows:

[0217] u(i) = | τ(h(i)) - τm(h(i)) |;

[0218] Where u(i) is the stress imbalance index at the i-th depth point, in dry Pascals; τ(h(i)) comes from the basic response data R; and τm(h(i)) is calculated by G3.

[0219] Meanwhile, to assess the overall degree of zonal imbalance, the average stress imbalance degree Uj is defined:

[0220] Uj = Σ u(i) / Nj;

[0221] Where Uj is the average stress imbalance of the j-th zone (rigid zone, transition zone, flexible zone), in kilopascals; Nj is the number of depth points in that zone.

[0222] The imbalance indicators corresponding to the three zones are summarized as follows:

[0223] U = {Ur,Ut,Uf};

[0224] Ur is the imbalance index of the rigid region, Ut is the imbalance index of the transition region, and Uf is the imbalance index of the flexible region, all of which are derived from the above average calculation results.

[0225] Sub-step 4.2: Quantitative identification of regions with significant negative friction.

[0226] By using the set of stress imbalance indices U and the set of partitioning results Zone, a set of regions with significant negative skin friction ZN is obtained. Combining the Zone partitioning information with the imbalance index U, regions with significant negative skin friction are identified. The negative skin friction criterion is defined as follows:

[0227] First condition: τ(h(i))<0

[0228] Second condition: u(i) > Tu;

[0229] Tu represents the negative friction threshold, ranging from 20 kPa to 50 kPa, determined through engineering experience. Depth points meeting both conditions are assigned to the significant negative friction region.

[0230] The depth points that meet the conditions are divided into sets according to continuous depth ranges:

[0231] ZN = {(h(a1),h(b1)),...,(h(an),h(bn))};

[0232] Where (h(a1),h(b1)) represents the first segment of continuous negative friction interval.

[0233] Sub-step 4.3: Generate the pile spacing and pile length control parameters for areas with significant negative skin friction.

[0234] By using the set of negative skin friction significant zones ZN and the final parameter set O*, the set of adjustment parameters P1 for pile spacing and pile length is obtained. For the depth section within ZN, adjustment suggestions for pile spacing and pile length are generated based on the magnitude of the skin friction imbalance index. The pile spacing reduction amount ΔL is defined as follows:

[0235] ΔL = β*Ur;

[0236] Where ΔL is the pile spacing reduction in meters; Ur is the rigid zone imbalance index, derived from set U; β is the adjustment coefficient, ranging from 0.1 to 0.3, determined based on construction feasibility and stress transfer requirements. If ΔL is less than 0.2 meters, then 0.2 meters is taken as the minimum adjustment.

[0237] Define the pile tip extension ΔH:

[0238] ΔH = γ*Ut;

[0239] Where ΔH is the pile end lengthening amount in meters; Ut is the imbalance index of the transition zone; and γ is the pile end adjustment coefficient, ranging from 0.5 to 1.5.

[0240] The set of pile spacing and pile length adjustment parameters P1 is organized as follows:

[0241] P1 = {ΔL,ΔH}

[0242] Sub-step 4.4: Optimization of grouting density and filler modulus in the region where frictional resistance changes slowly in the opposite direction.

[0243] By using the set of stress imbalance indices U and the set of zoning results Zone, the set of optimized parameters P2 for grouting and filling is obtained.

[0244] Identify sections where frictional resistance changes slowly in the opposite direction, i.e., sections that meet the following conditions:

[0245] First, Ut is less than 15 kPa.

[0246] Second, the section is located in the flexible region Zf

[0247] For the above-mentioned sections, it is recommended to increase the grouting density at the pile ends and the modulus of the transition filler. Define the grouting density adjustment coefficient ΔG:

[0248] ΔG = δ*Uf;

[0249] Where ΔG is the grouting density increase ratio, ranging from 0.1 to 0.4; δ is the adjustment ratio coefficient, ranging from 0.5 to 1.0, determined by construction experience; and Uf is the imbalance index of the flexible zone.

[0250] Define the increase in packing modulus ΔE:

[0251] ΔE = ε*Uf;

[0252] Where ΔE is the increase in packing modulus, in megapascals; ε is the packing adjustment coefficient, ranging from 1 to 5.

[0253] The optimized parameter set P2 for grouting and filling is summarized as follows:

[0254] P2 = {ΔG,ΔE}

[0255] Sub-step 4.5: Parameter feedback integration and generation of executable optimization scheme.

[0256] The set of parameters for adjusting pile spacing and pile length, P1, is integrated with the set of parameters for optimizing grouting and filling, P2, to form the final set of control parameters:

[0257] P = {ΔL,ΔH,ΔG,ΔE};

[0258] ΔL is used for adjusting the pile layout during construction; ΔH is used for extending the pile end during construction; ΔG is used for grouting process control; and ΔE is used for selecting and adjusting the filler material type and ratio.

[0259] An executable solution is developed, including: reducing the pile spacing in the negative skin friction zone; increasing the pile tip bearing depth; increasing the grouting density; increasing the modulus of the transition layer; and transmitting the data back to the monitoring system in real time to achieve dynamic construction control. This generates a parameter scheme that can be directly used for on-site construction adjustments, enabling dynamic optimization design based on measured feedback.

[0260] Step 5: Construction Verification and Closed-Loop Dynamic Correction. The control commands output from Step 4 are applied to on-site construction. By adjusting the pile foundation construction parameters and transition layer material configuration, structural improvements are completed, and settlement and pile stress changes are continuously monitored. When the monitoring system detects that the pile-soil friction curve tends to be continuous and the difference between the settlement on the back of the abutment and the main roadbed settlement is less than the threshold, the closed-loop control ends. If there are still stress abrupt changes or a trend of new negative friction, the corrected data is re-inputted into Step 3 for model iteration until stable convergence.

[0261] Sub-step 5.1: Application of control instructions and adjustment of construction parameters.

[0262] The on-site construction process is adjusted based on the various control parameters in the control parameter set P. The control parameter set P is represented as follows:

[0263] P = {ΔL,ΔH,ΔG,ΔE};

[0264] Where ΔL is the reduction in pile spacing in meters; ΔH is the increase in pile tip length in meters; ΔG is the increase in grouting density, dimensionless, ranging from 0.1 to 0.4; and ΔE is the increase in the modulus of the transition layer filler in megapascals. The adjusted new pile spacing Lnew is expressed as:

[0265] Lnew = L - ΔL;

[0266] Where L is the original design pile spacing, in meters. The adjusted new pile length Hnew is represented as:

[0267] Hnew = H + ΔH;

[0268] Where H is the original pile length in meters. The adjusted new grouting density is expressed as:

[0269] Gnew = G * (1 + ΔG);

[0270] Where G is the original grouting density, in kilograms per cubic meter. The adjusted modulus of the new filler is expressed as:

[0271] Enew = E + ΔE;

[0272] Where E represents the original modulus of the transition layer filler, in megapascals (MPa). The above adjustment parameters are then summarized to form the set of execution parameters for construction.

[0273] C1 = { Lnew, Hnew, Gnew, Enew}

[0274] C1 is used to transform model feedback into actionable construction adjustments, enabling targeted structural optimization to improve stress transmission paths and settlement coordination.

[0275] Sub-step 5.2: Improve the acquisition of monitoring data after the structure is completed.

[0276] Following the construction adjustments based on C1, the monitoring network established in step 1 continues to be used to collect strain ε(i), lateral displacement u(i), and pore pressure p(i). This forms an updated monitoring dataset.

[0277] R2 = { τ2(i),du2(i),p2(i)};

[0278] Where τ2(i) is the inverse value of pile side friction after construction adjustment, in kilopascals, obtained through the formula:

[0279] τ2(i) = E*ε(i);

[0280] Where E is the elastic modulus of the pile, in kilopascals, ranging from 20 to 35 gigapascals; du2(i) is the difference in lateral displacement, in millimeters; and p2(i) is the pore pressure, in kilopascals.

[0281] Sub-step 5.3 evaluates frictional continuity and settlement difference to achieve quantitative construction adjustment of stress continuity and settlement coordination.

[0282] The friction continuity index Cτ is calculated based on the updated monitoring dataset R2:

[0283] Cτ(i) = | τ2(i) - τ2(i-1) |;

[0284] Where Cτ(i) represents the change in friction at the i-th depth point, in kilopascals. The settlement difference ΔS is defined as follows:

[0285] ΔS = |Sb - Sr|;

[0286] Where ΔS represents the settlement difference between the abutment back and the main roadbed, in millimeters; Sb represents the settlement value of the abutment back; and Sr represents the settlement value of the main roadbed, all obtained from settlement data at monitoring points. This forms a set of evaluation indicators:

[0287] E1 = { Cτ, ΔS}

[0288] E1 represents the set of evaluation indicators.

[0289] Sub-step 5.4: Loop closure determination and model iteration triggering.

[0290] The closed-loop state identifier F is obtained through the evaluation index set E1. First, the closed-loop determination conditions are set:

[0291] First condition: max(Cτ(i)) <Tτ

[0292] Second condition: ΔS <TS

[0293] Where Tτ is the frictional continuity threshold, ranging from 5 kPa to 15 kPa; TS is the settlement difference limit, ranging from 5 mm to 15 mm. Define the closed-loop state identifier F:

[0294] When both of the above conditions are met:

[0295] F = 1;

[0296] If any condition is not met:

[0297] F = 0.

[0298] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0299] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0300] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0301] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0302] Various aspects of this disclosure are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0303] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0304] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0305] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0306] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for optimizing composite foundations for highway reconstruction and expansion based on intelligent monitoring feedback, characterized in that, The method includes: Step 1: Deploy monitoring units in the bridge approach slab and culvert opening areas to form a multi-source monitoring network, acquire and output basic response datasets; Step 2: Based on the basic response dataset, identify the negative skin friction zone region and extract feature indices to form a stress discontinuity feature matrix; at the same time, establish the correspondence between abutment stiffness, fill compressibility and pile spacing, and output the boundary parameter set; among which, the feature indices include: stress mutation depth, pile bending moment change rate and soil lateral displacement mutation point; the boundary parameter set includes: abutment stiffness, fill compressibility and pile spacing. Step 3: Input the stress discontinuity characteristic matrix and boundary parameter set into the foundation mechanics response model, introduce the boundary stiffness correction coefficient and the friction reverse transition coefficient, divide the pile-soil system into rigid zone, transition zone and flexible zone through zoning, use measured data for correction, and output the corrected zoning response model and final parameter set. Step 4: Using the corrected partitioned response model and the final parameter set as input, generate control parameters, and provide feedback on the generation of control parameters and differentiated optimization design to output the total set of control parameters; Step 5: Use the set of control parameters as input, continue monitoring after structural improvement, and end the closed loop when the pile-soil friction curve becomes continuous and the settlement difference is less than the threshold; otherwise, re-input the corrected data into Step 3 for iteration. Step 2 includes: The parameters in the basic response dataset were reorganized to form a deep sequence dataset. To eliminate the influence of noise, the pile side friction in the basic response dataset was smoothed to obtain the smoothed friction value. Based on the depth sequence dataset, the frictional change in the depth direction is calculated. According to the smoothed frictional value and the reasonable threshold range of the frictional change, the negative frictional trend and stress mutation point are determined, thereby screening out the depth intervals where negative frictional occurs continuously and forming a set of negative frictional regions. By utilizing the depth segments corresponding to the set of negative friction regions, the depth points that meet the conditions and their corresponding features are combined to form a set of feature vectors. The set of feature vectors is integrated into a matrix form to construct a stress discontinuous feature matrix, where each row of the matrix represents a depth feature point. Based on design data and construction records, establish and output the correspondence between abutment stiffness, fill compressibility and pile spacing, and output the set of boundary parameters. Step 3 includes: The boundary stiffness correction coefficients at each depth point are calculated based on the stress discontinuity characteristic matrix and the parameters in the boundary parameter set, so as to obtain the boundary stiffness correction coefficient sequence. The transition points where the friction direction changes significantly are identified by using the boundary stiffness correction coefficient sequence and the basic response dataset, so as to obtain the friction reverse transition parameter sequence. Based on the boundary stiffness correction coefficient sequence and the friction reverse transition parameter sequence, the pile-soil system is divided into rigid zone, transition zone and flexible zone, forming a set of partition results. Initial pile-soil stress transfer curve parameters are defined for each partition, forming an initial stress transfer curve parameter set. Based on the partitioned result set and the initial stress transfer curve parameter set, a foundation mechanical response model is constructed. A friction prediction function is defined through the partitioning function, and the prediction results are compared with the measured friction in the foundation response dataset. The residual function and objective function are defined to form the foundation mechanical response model. Based on the foundation mechanical response model, an iterative algorithm is used to dynamically optimize the stress transfer curve parameters of each zone. By continuously updating the parameters and recalculating the objective function, the preset convergence conditions are met, and finally the corrected zone response model and the corresponding final parameter set are output.

2. The method for optimizing composite foundations for highway reconstruction and expansion based on intelligent monitoring feedback as described in claim 1, characterized in that, Step 1 includes: Based on the basic design data of the bridge abutment slab and culvert opening area, the influence range of the rigid boundary is identified, thereby determining the monitoring control area. This area covers the distance from the back of the abutment to the roadbed and the depth range above the pile tip. The final output is a set of coordinates of the monitoring area. Among them, the distance from the back of the abutment to the roadbed is 2 meters to 6 meters, and the depth range above the pile tip is 0.5 meters to 2 meters. Within the coordinate set of the monitoring area, the location of the monitoring unit is determined along the interface between the pile and the soil, and the output is a set of monitoring unit placement points. Additional monitoring points are added to the pile tip area to form a supplementary monitoring point set. The monitoring unit monitoring point set and the supplementary monitoring point set are then merged to obtain a complete monitoring point set. All monitoring units in the complete set of monitoring points are connected to a unified data acquisition system. Dynamic data acquisition is carried out by setting a basic sampling frequency and combining it with the working condition adjustment coefficient to form a preliminary data set. The raw monitoring data in the preliminary dataset are calculated and organized to form the basic response dataset.

3. The method for optimizing composite foundations for highway reconstruction and expansion based on intelligent monitoring feedback as described in claim 1, characterized in that, Step 4 includes: Based on the corrected partition response model and the final parameter set, the difference between the measured friction and the friction prediction function in each partition is calculated, the stress imbalance index at each depth point is defined, and the average stress imbalance of each partition is further calculated to form a partition stress imbalance index set. By combining the set of zoning results and the set of zoning stress imbalance indexes, a negative skin friction significant area is identified by setting a negative skin friction criterion. The criterion includes that the pile side skin friction is negative and the stress imbalance index exceeds the negative skin friction identification threshold. The continuous depth intervals that meet the conditions are sorted and output as a set of negative skin friction significant areas. For the set of areas with significant negative skin friction, suggestions for adjusting pile spacing and pile length are generated based on the magnitude of the stress imbalance index, in order to generate a set of adjustment parameters for pile spacing and pile length. Identify sections where frictional resistance changes slowly in the opposite direction. These sections are located in the flexible zone. For these sections, define the grouting density increase ratio and the filler modulus increase value to generate a set of optimized grouting and filler parameters. The set of parameters for adjusting pile spacing and pile length, as well as the set of parameters for optimizing grouting and filling, are integrated to form the final set of control parameters. This set includes the amount of pile spacing reduction, the amount of pile end lengthening, the proportion of grouting density increase, and the value of filling modulus increase.

4. The method for optimizing composite foundations for highway reconstruction and expansion based on intelligent monitoring feedback as described in claim 1, characterized in that, Step 5 includes: The on-site construction process is adjusted based on the set of control parameters. The adjusted construction parameters are calculated based on the amount of pile spacing reduction, pile end lengthening, grouting density increase ratio, and filler modulus increase value. These parameters include new pile spacing, new pile length, new grouting density, and new filler modulus, forming a set of construction parameter execution. After completing the construction adjustment based on the construction parameters, the monitoring network continues to collect data to obtain the pile side friction, lateral displacement difference and pore pressure after the construction adjustment, forming an updated monitoring dataset; The effect is evaluated based on the updated monitoring dataset. The friction change range at each depth point is calculated as the friction continuity index, and the settlement difference between the settlement value on the back of the bridge abutment and the settlement value of the main roadbed is calculated to form a set of evaluation indicators. The closed-loop determination is based on the set of evaluation indicators. The determination conditions are set, requiring that the maximum value of the friction continuity index is less than the friction continuity threshold and the settlement difference is less than the settlement difference limit. When both of the above conditions are met at the same time, the closed-loop status flag is set to 1 and the closed loop ends. Otherwise, the closed-loop status flag is set to 0 and the corrected data is re-entered into step 3 for iteration.

5. A composite foundation optimization system for highway reconstruction and expansion based on intelligent monitoring feedback, characterized in that, The system for implementing the method according to any one of claims 1 to 4, the system comprising: The monitoring network deployment module is used to deploy monitoring units in the bridge approach slab and culvert opening areas to form a multi-source monitoring network, acquire and output basic response datasets; The stress anomaly identification module is used to identify negative skin friction zone regions and extract feature indicators based on the basic response dataset, forming a stress discontinuity feature matrix; at the same time, it establishes the correspondence between abutment stiffness, fill compressibility and pile spacing, and outputs a set of boundary parameters. The rigid-flexible transition model module is used to input the stress discontinuity characteristic matrix and boundary parameter set into the foundation mechanics response model, introduce the boundary stiffness correction coefficient and the friction reverse transition parameter, divide the pile-soil system into rigid zone, transition zone and flexible zone through zoning, use measured data for correction, and output the corrected zoning response model and final parameter set. The regulation parameter generation module is used to take the corrected partition response model and the final parameter set as input to generate regulation parameters, and to provide feedback on the generation of regulation parameters and differentiated optimization design, so as to output the total set of regulation parameters. The closed-loop control module takes the set of control parameters as input, and continues to monitor after the structural improvement is completed. When the pile-soil friction curve tends to be continuous and the settlement difference is less than the threshold, the closed loop ends; otherwise, the corrected data is re-inputted into step 3 for iteration.

6. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-4.

Images