Arch bridge construction counterweight optimization method and system based on stress monitoring

Abstract

The invention relates to the technical field of bridge construction, in particular to an arch bridge construction counterweight optimization method and system based on stress monitoring. The method comprises the following steps: arranging a plurality of stress sensors at different parts of an arch bridge, and collecting stress monitoring data; performing data preprocessing operation and feature extraction operation on the stress monitoring data; inputting the features, the construction state parameters and the structure parameters into a machine learning model to obtain an arch bridge multi-section stress trend; generating an arch bridge construction counterweight optimization scheme by adopting a non-dominated sorting genetic algorithm; dynamically adjusting the counterweight optimization scheme according to stress monitoring data and the multi-section stress trend of the arch bridge; therefore, accurate and scientific optimization of the counterweight scheme in the arch bridge construction process is realized.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of bridge construction, in particular to an arch bridge construction counterweight optimization method and system based on stress monitoring. BACKGROUND

[0002] In the arch bridge construction process (such as cable hoisting, support cast-in-place or swivel body construction), counterweights (such as sandbags, water tanks and the like) are often applied at specific positions to balance the internal force of the structure, control deformation or meet the stress requirements of a specific construction stage. There are schemes for adjusting the counterweights in the bridge construction process in the prior art. For example, a dynamic balance counterweight adjustment method for a swivel bridge is disclosed in Chinese Invention Patent (CN120250493A), which comprises a real-time weighing monitoring system, a counterweight adjustment system, a hydraulic pulling device and a counterweight moving device. The counterweight moving device comprises a plurality of movable counterweight blocks, a plurality of flat track cars and a track car longitudinal moving track. The movable counterweight blocks are mounted on the flat track cars, and two adjacent movable counterweight blocks are connected by a connecting rope. The hydraulic pulling device comprises two fixed counterweight blocks and two cross-hole jacks. The two fixed counterweight blocks are arranged at the two ends of the longitudinal track, and the two cross-hole jacks are installed on the two fixed counterweight blocks. The cross-hole jacks are connected with the adjacent end movable counterweight blocks. The counterweight adjustment system calculates the distribution points and the loading weight of the counterweight blocks in real time according to the monitoring data of the real-time weighing monitoring system, and controls the corresponding cross-hole jacks to pull the movable counterweight blocks.

[0003] However, when the above scheme adjusts the counterweights, the counterweights are adjusted by experience, which results in that the counterweight adjustment is relatively rough, and the real-time changes of the structure profit state are not considered, so that accurate and scientific adjustment of the counterweights in the arch bridge construction process cannot be realized. Meanwhile, when the stress monitoring data is processed, the weak stress signal processing is not accurate. Meanwhile, when the intelligent model is used to optimize and predict the configuration of the arch bridge construction, all time steps of historical information are treated equally, and the key period (such as the pouring peak period) and the regular period (such as the night without construction) cannot be automatically distinguished. Meanwhile, when the existing scheme determines the multi-objective function of the counterweight optimization, the accurate matching of the engineering actual demand in different construction stages cannot be met. SUMMARY

[0004] In order to solve the above technical problems, the application provides an arch bridge construction counterweight optimization method and system based on stress monitoring, which is used to solve the problems in the prior art.

[0005] The application provides an arch bridge construction counterweight optimization method based on stress monitoring, which comprises the following steps: S1: arranging a plurality of stress sensors at different positions on the arch bridge, and collecting stress monitoring data; S2: performing data preprocessing operation and feature extraction operation on the stress monitoring data; S3: input the extracted features, construction state parameters and structure parameters into a machine learning model to obtain an arch bridge multi-section stress trend; The machine learning model is a long short-term memory network model with a time attention mechanism layer; the time attention mechanism layer adopts key time step weighting to highlight the influence of the construction stage on the stress trend; S4: generating an arch bridge construction counterweight optimization scheme by using a non-dominated sorting genetic algorithm; S5: dynamically adjusting the counterweight optimization scheme according to the stress monitoring data and the arch bridge multi-section stress trend.

[0006] Preferably, the time attention mechanism layer adopts key time step weighting to highlight the influence of important construction stages on the stress trend, specifically: Sa: calculating an energy score of each time step t; Sb: normalizing the energy score of each time step t to obtain an attention weight of each time step t; Sc: generating a context vector according to the attention weight of each time step t, the context vector being input into a fully connected layer for final prediction of future stress values.

[0007] Preferably, the S4 is specifically: S4.1: establishing a multi-constraint condition for the arch bridge construction counterweight optimization; S4.2: establishing an objective function for the arch bridge construction counterweight optimization; The expression of the objective function is: ; wherein f1 is the total amount of counterweight, f2 is the deformation amount, f3 is the construction period, λ 1 is a weight coefficient of the total amount of counterweight, λ 2 is a weight coefficient of the deformation amount, λ 3 is a weight coefficient of the construction period; in the early construction period, λ 1=0.7, λ 2=0.2, λ 3=0.1; in the middle construction period, λ 1=0.4, λ 2=0.3, λ 3=0.3; in the late construction period, λ 1=0.5, λ 2=0.1, λ 3=0.4; S4.3: generating a counterweight optimization scheme by using a non-dominated sorting genetic algorithm according to the multi-constraint condition and the objective function.

[0008] Preferably, the constraints include stress safety constraints, deformation coordination constraints, construction period constraints, and safety margin constraints.

[0009] Preferably, the stress safety constraints are specifically: for each key section i ( i =1, 2,..., n ), the maximum tensile stress t,i and the maximum compressive stress σ c,i satisfy: ; wherein n is the total number of key sections, σ t and σ c are the design allowable tensile stress limit and compressive stress limit, respectively; The deformation coordination constraints are specifically: the maximum inclined displacement Δ of the temporary support top satisfies: ; ; wherein H is the support height, and Δ 允许 is the maximum allowable inclined displacement of the temporary support; The construction period constraints are specifically: the total number of counterweight adjustments during construction N 调整 satisfies: ; wherein N 允许 is the maximum allowable adjustment number of the construction period; The safety margin constraints are specifically: the ratio of the actual stress to the limit value satisfies: ; wherein k is a coefficient.

[0010] Preferably, k is 0.8 or 0.9.

[0011] Preferably, in S4.2, the main objective function is to minimize the total amount of counterweight under the premise of satisfying the above constraints; and the secondary objective function is to minimize deformation and to shorten the construction period.

[0012] Preferably, in S1, the stress sensors are arranged at the key sections of the main arch ring of the arch bridge, the temporary supports, and the counterweight application points, respectively.

[0013] Preferably, in step S2, the data preprocessing involves using an adaptive filtering algorithm based on wavelet transform to perform data filtering on the stress monitoring data; specifically: The stress monitoring data is subjected to zero-mean normalization to obtain zero-mean normalized data; The zero-mean processed data was subjected to a 7-level discrete wavelet transform using Morlet wavelets to obtain the approximation coefficients A7 and detail coefficients D1~D7. The approximation coefficients and detail coefficients are subjected to combined threshold denoising; specifically: No thresholding is applied to the approximation coefficient layer; For the detail coefficients D 1- D 4. Use a universal threshold T for threshold processing; For the detail coefficients D5-D7, an adaptive soft threshold T is applied. j It retains some weak high-frequency components that are related to the effective signal.

[0014] According to another aspect of the present invention, a counterweight optimization system for arch bridge construction based on stress monitoring is provided. The system employs the aforementioned method for optimizing counterweights for arch bridge construction based on stress monitoring. The system includes: The data acquisition module is used to deploy multiple stress sensors at different locations on the arch bridge and collect stress monitoring data; The data processing module is used to perform data preprocessing and feature extraction operations on the stress monitoring data; The stress trend prediction module is used to input the features, construction state parameters and structural parameters into the machine learning model to obtain the stress trend of the multi-section of the arch bridge. The machine learning model is a long short-term memory network model with an added time attention mechanism layer. The time attention mechanism layer uses key time steps to weight the impact of important construction stages on stress trends. The counterweight optimization module is used to generate an optimized counterweight scheme for arch bridge construction using a non-dominated sorting genetic algorithm. The dynamic adjustment module is used to dynamically adjust the counterweight optimization scheme based on the stress monitoring data and the stress trend of the multi-section of the arch bridge.

[0015] The embodiments of the present invention have the following technical effects: Firstly, a plurality of stress sensors are arranged at different positions of the arch bridge, and stress monitoring data is collected; data preprocessing operation and feature extraction operation are performed on the stress monitoring data; the features, construction state parameters and structure parameters are input into a machine learning model to obtain arch bridge multi-section stress trends; a non-dominated sorting genetic algorithm is used to generate an arch bridge construction counterweight optimization scheme; the counterweight optimization scheme is dynamically adjusted according to the stress monitoring data and the arch bridge multi-section stress trends; thereby, accurate and scientific optimization of the counterweight scheme in the arch bridge construction process is realized.

[0016] The approximate coefficient and the detail coefficient are combined threshold denoising processing, through the adaptive threshold strategy, the low frequency effective component (<1Hz) and the high frequency noise (>10Hz) in the construction stress signal are successfully separated. The scheme not only improves the signal-to-noise ratio of the strain monitoring data, but also provides a reliable input data basis for subsequent dynamic response prediction and counterweight optimization, which is a key link in the intelligent monitoring system of the arch bridge construction.

[0017] The importance weight of each time step is dynamically calculated, the key information is automatically focused, the time step corresponding to the key construction stage such as concrete pouring, counterweight adjustment and hanging basket movement can be automatically identified, and higher weight is given; the interference of the conventional period such as night without construction and material transportation is reduced, and the influence of irrelevant data on the prediction result is avoided.

[0018] Meanwhile, the dynamic weight guiding algorithm is used to prioritize the key target in different stages, and the priority of each target in the multi-objective optimization (such as stress safety, deformation control and construction period efficiency) is dynamically adjusted according to the construction progress, so that the counterweight scheme accurately matches the actual engineering requirements in different construction stages. This design breaks through the limitations of static weight and global balance in traditional optimization methods, significantly improves the engineering applicability and rationality of the optimization result. The dynamic weight coefficient adjustment realizes the technical path of 'construction stage perception-weight dynamic allocation-multi-objective collaborative optimization', so that the counterweight scheme can focus on the most critical requirements in each construction stage: safety in the early stage, deformation and construction period in the medium term, and cost reduction and efficiency improvement in the later stage. This design not only improves the engineering rationality of the optimization result, but also provides a 'accurate, flexible and economic' counterweight strategy for arch bridge construction through dynamic adaptation. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the drawings needed in the specific embodiments or prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0020] Figure 1 is a flowchart of an arch bridge construction counterweight optimization method based on stress monitoring provided by an embodiment of the present application; Figure 2 is a flowchart of a key time step weighting adopted by the time attention mechanism layer provided by an embodiment of the present application; Figure 3 is a flowchart of generating an arch bridge construction counterweight optimization scheme using a non-dominated sorting genetic algorithm provided by an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to make the objectives, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described clearly and completely below. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application.

[0022] Embodiment 1, Appendix Figure 1 shows a flowchart of an arch bridge construction counterweight optimization method based on stress monitoring, as shown in Appendix Figure 1 The arch bridge construction counterweight optimization method based on stress monitoring comprises the following steps: S1: arranging a plurality of stress sensors at different positions on the arch bridge and collecting stress monitoring data; Among them, the stress sensors are arranged at the key sections (such as arch foot, L / 4 section, arch top) of the main arch ring, temporary supports (such as supports, hanging baskets) and counterweight application points of the arch bridge, and a full-dimensional stress monitoring network is constructed.

[0023] In the arch bridge construction counterweight optimization method, the strain sensor is a key device for monitoring the stress state of the structure, mainly used for measuring the linear strain of materials such as concrete and steel. By monitoring the linear strain and combining the strain-stress conversion relationship, while considering temperature compensation, the true stress of the structure can be obtained. This embodiment adopts two types of strain sensors, fiber Bragg grating strain gauges and resistance strain gauges, to combine and arrange them to fully exert their respective advantages and achieve more accurate and comprehensive monitoring of the stress of the structure.

[0024] S2: performing data preprocessing operation and feature extraction operation on the stress monitoring data; Among them, the data preprocessing is to perform data filtering operation on the stress monitoring data using an adaptive filtering algorithm based on wavelet transform; Specifically, according to the frequency characteristics of the stress signal, a wavelet basis function and a decomposition layer number are designed to separate the effective stress signal and the noise component, the wavelet basis function is, for example, Morlet wavelet, and the decomposition layer number is 7 layers.

[0025] The adaptive filtering algorithm based on wavelet transform is used to perform data filtering operation on the stress monitoring data, and the operation is specifically: The stress monitoring data is subjected to zero-mean processing to obtain zero-mean processed data; The zero-mean processed data is subjected to 7-layer discrete wavelet transform by using Morlet wavelet to obtain approximation coefficients A 7and detail coefficients D 1~ D 7; Table 1 shows the frequency range of the wavelet transform coefficients and the corresponding signal type.

[0026] Table 1 Frequency range of wavelet transform coefficients and corresponding signal type In this step, the approximation coefficients contain the key information of the structural stress during the construction process, such as the slow stress change due to the gradual increase of the material self-weight during concrete pouring, the quasi-static load effect on the structure caused by the movement of the hanging basket, etc. These information is an important basis for analyzing the stress state of the structure and evaluating the construction safety. Therefore, by not processing the approximation coefficients, the key effective signals can be ensured not to be damaged by threshold processing, thereby providing accurate data basis for subsequent dynamic response prediction and weight optimization. At the same time, the stress fluctuation caused by the construction load is mainly low-frequency signal, and the frequency is usually less than 1 Hz. These signals reflect the quasi-static load effect of the operation such as concrete pouring and hanging basket movement during the construction process, as well as the material time-varying characteristics such as concrete creep, which is crucial for the evaluation of the stress state of the structure. The interference such as vibration noise is high-frequency signal, and the frequency is greater than 10 Hz. These noises are derived from the mechanical vibration of the concrete vibrator, the start-stop impact of the construction machinery, etc., and have the characteristics of suddenness and wide frequency band, with energy concentrated in the high-frequency band, which can mask the effective low-frequency signal. When there is a mixture of weak signal and strong noise in the stress monitoring data, a method combining soft and hard threshold values can be used. For the high-frequency part dominated by noise, the hard threshold method is used to remove the obvious strong noise first; then for the remaining coefficients containing weak effective signals, the soft threshold method is used for further processing to reduce the signal deviation. For example, in some construction stages, the vibration noise is very strong, and the hard threshold method is used to remove most of the high-frequency peak noise in the high-frequency detail coefficient layer first, and then the soft threshold method is used to retain the weak effective stress signal more accurately in the subsequent processing. Therefore, the approximation coefficients and the detail coefficients are subjected to combined threshold denoising processing, and the processing is specifically: The approximation coefficient layer is not subjected to threshold processing; The detail coefficients D 1- D4, threshold processing is performed using a general threshold T to directly eliminate coefficients with absolute values less than T ; wherein the general threshold T is calculated according to the formula: ; wherein, σ is the standard deviation of the detail coefficients, N is the length of the stress monitoring data.

[0027] For the detail coefficients D5-D7, an adaptive soft threshold T j is used to retain part of the weak high-frequency components related to the effective signal, such as early high-frequency precursor signals of concrete micro-crack expansion.

[0028] wherein the frequency of the detail coefficients D5-D7 is in the mixed frequency band of 1.56-6.25Hz, and therefore an adaptive soft threshold is used for processing; wherein the adaptive soft threshold T j is calculated according to the formula: ; wherein j represents the number of wavelet decomposition layers, and a is an adjustment coefficient, which is taken as 1.2-1.5; The approximate coefficients and the detail coefficients after threshold processing are reconstructed through inverse wavelet transform to obtain the preprocessed stress monitoring data.

[0029] By precisely matching the Morlet wavelet base function with 7 layers of decomposition layers and combining the adaptive threshold strategy, efficient separation of low-frequency effective components (<1Hz) and high-frequency noise (>10Hz) in the construction stress signal is successfully achieved. This scheme not only improves the signal-to-noise ratio of the strain monitoring data, but also provides a reliable input data basis for subsequent dynamic response prediction and weight optimization, which is a key link in the intelligent monitoring system of arch bridges.

[0030] The preprocessed stress monitoring data is subjected to a feature extraction operation; wherein the extracted features include: statistical features: mean of the preprocessed stress monitoring data, variance of the preprocessed stress monitoring data, extreme value of the preprocessed stress monitoring data; time-frequency features: wavelet energy entropy of the preprocessed stress monitoring data; trend features: stress change rate calculated by a sliding window.

[0031] S3: inputting the features, construction state parameters and structure parameters into a machine learning model to obtain the stress trend of the multi-section arch bridge; The construction state parameters are concrete pouring progress (current pouring section number / accumulative volume), applied counterweight value (left and right cantilever end counterweight difference), and hanging basket position; and the structure parameters include current concrete strength (measured rebound value conversion), temporary support stiffness (bracket deformation monitoring value back calculation), and environmental temperature gradient.

[0032] The arch bridge multi-section stress trend is the maximum value and change trend of the key section of the main arch ring within 2-6 hours in the future; and the key section includes arch foot compressive stress, L / 4 section shear stress, and arch top tensile stress.

[0033] The machine learning model is a long short-term memory network model with a time attention mechanism layer; The long short-term memory network model is stacked by three LSTM layers (Stacked LSTM), and the number of hidden units of each LSTM layer is 128, 64 and 32 respectively, which is reduced layer by layer to reduce the calculation complexity and extract high-level abstract features.

[0034] The input dimension of the long short-term memory network model is 12; The time step of the long short-term memory network model is t=6 (corresponding to the monitoring data of the first 3 hours, every 30 minutes as a time step); The activation function of the long short-term memory network model is that the hidden layer uses a tanh function (limiting the output to the interval [-1, 1] to enhance the nonlinear expression ability), and the gating mechanism uses a sigmoid function (outputting a probability value of 0-1 to control the information flow).

[0035] In the arch bridge construction counterweight optimization scenario, the dynamic response (such as stress and deformation) of the main arch ring is significantly affected by the dynamic changes of multi-stage loads in the construction process. For example, the material accumulation during concrete pouring, the cantilever end load offset caused by the movement of the hanging basket, the sudden increase and decrease of the counterweight block and other key operations will produce “high-influence load” at a certain time step. The monitoring data of these time periods is crucial for predicting future stress evolution. However, the traditional attention mechanism long short-term memory network model treats all time steps of historical information equally and cannot automatically distinguish between key periods (such as pouring peak) and regular periods (such as night without construction). Therefore, the embodiment sets a time attention mechanism layer and automatically focuses on key information by dynamically calculating the importance weight of each time step.

[0036] The input of the time attention mechanism layer is the hidden state sequence of all time steps output by the LSTM layer of the long short-term memory network model H =[ h 1, h 2,..., h t ]t is the time step, e.g. t = 6 corresponds to the first 3 hours of monitoring data, h t is the hidden state sequence of the t-th time step, these h t In the training data, the semantic information of the stages such as "whether it is concrete pouring" and "whether it is counterweight adjustment" will be contained, because the sensor readings, operation records, text descriptions, etc. of these stages in the training data will have statistical rules, such as the sudden rise of sensor load during concrete pouring, and the specific change pattern of displacement / stress during counterweight adjustment, wherein each h t is a vector with a length of d , e.g. d = 32, determined by the number of hidden units of the last layer of LSTM.

[0037] Specifically, the time attention mechanism layer adopts key time step weighting to highlight the influence of important construction stages on stress trends; as shown in the accompanying Figure 2 , the specific process is as follows: Sa: calculate the energy score of each time step t; For each time step t, the energy score e t of the time step is calculated through a learnable linear transformation; the calculation formula is as follows: ; wherein v a T is a weight vector in the attention mechanism, and the dimension of the weight vector is usually consistent with the dimension of the tanh( ) output. The weight vector and the linear transformation result are dot- multiplied to weight or map the scores of different positions, so as to finally obtain the attention score of the position.

[0038] W a is a weight matrix in the attention mechanism, used for linear transformation of each hidden state h t ; its role is to map the input hidden state to a new feature space, and the dimension is determined by the number of columns of W a , so as to perform inner product operation with v a later; h t represents the hidden state of the LSTM layer at the time step t , which encodes the information of the previous t time points in the sequence, and is one of the objects to be focused on by the attention mechanism; b a is a bias term, used to increase the fitting ability of the model by adding a learnable constant offset after the linear transformation of W a h t

[0039] Sb: normalize the energy score of each time step t to obtain the attention weight of each time step t; In this step, the energy score e t of each time step is converted into an attention weight a t by a softmax function, ensuring that the sum of all weights is 1; Sc: generate a context vector according to the attention weight of each time step t; In this step, the hidden state h h t of each time step t is weighted and summed according to the attention weight a α t to obtain a context vector c; The specific formula is: ; In this embodiment, c is a weighted summary of all historical time step information, where the h t of high-weight time steps (such as pouring periods) has a greater contribution to c, and the h t of low-weight time steps (such as stable periods) has little effect on the result; this vector will be used as the input of a fully connected layer for the final prediction of future stress values. Through the scheme of this embodiment, the time steps corresponding to key construction stages such as concrete pouring, counterweight adjustment, and hanging basket movement can be automatically identified and given higher weights; the interference of regular periods such as nighttime non-construction and material transportation is reduced, and irrelevant data is prevented from affecting the prediction results.

[0040] S4: generate an arch bridge construction counterweight optimization scheme using a non-dominated sorting genetic algorithm; In the context of arch bridge construction counterweight optimization, the formulation of the counterweight optimization scheme is not simply to minimize the total amount of counterweight, but rather to balance the conflicts between multiple objectives through intelligent algorithms (non-dominated sorting genetic algorithm) under the premise of meeting structural safety, deformation coordination, construction period control, and material cost constraints, to generate the optimal counterweight adjustment strategy. This embodiment uses an improved non-dominated sorting genetic algorithm combined with dynamic weight coefficients to achieve multi-objective collaborative optimization of arch bridge construction counterweight.

[0041] Specifically, as shown in FIG. 6, the steps of the method are as follows: Figure 3 ​As shown, the S4 is specifically: S4.1: Establishing the multi-constraint conditions for the arch bridge construction counterweight optimization; In this step, the constraint conditions include stress safety constraints, deformation coordination constraints, construction period constraints, and safety margin constraints.

[0042] Among them, the stress safety constraint is to ensure that the stress of the key section of the main arch ring (such as the arch foot, L / 4 section, and arch top) under the action of the counterweight does not exceed the allowable limit of the material, avoiding concrete cracking or steel yielding; Specifically: For each key section i ( i =1,2,..., n ), the maximum tensile stress σ t,i and the maximum compressive stress σ c,i need to meet: ; In the formula, n is the total number of key sections, [ σ t ] and [ σ c ] are the design allowable tensile stress limit and compressive stress limit respectively.

[0043] Among them, the deformation coordination constraint is to limit the inclination displacement of the temporary support (such as the support and the hanging basket) under the action of the counterweight, avoiding the chain reaction caused by the instability of the support; Specifically: The maximum inclination displacement Δ of the top of the temporary support needs to meet: ; ; Δ 允许 is the maximum allowable inclination displacement of the temporary support.

[0044] Among them, the construction period constraint is to control the operation frequency and time cost of the counterweight adjustment, avoiding the impact of frequent hoisting on the overall construction progress.

[0045] Specifically: The total number of counterweight adjustments during construction N 调整 needs to meet: ; In the formula, N 允许 is the maximum allowable adjustment frequency of the construction period.

[0046] Among them, the safety margin constraint is to cope with construction errors (such as material strength fluctuations and environmental temperature mutations), leaving a certain stress safety margin.

[0047] Specifically: The ratio of actual stress to the limit value must meet the following requirements: ; In the formula, k is a coefficient. In this embodiment, k is 0.8 or 0.9; that is, the actual stress does not exceed 80% to 90% of the limit value, and a redundancy space of 10% to 20% is reserved.

[0048] S4.2: Establish the objective function for optimizing the construction counterweight of the arch bridge; The objective function includes a primary objective function and a secondary objective function. The primary objective function is to minimize the total counterweight while satisfying the above constraints, so as to reduce material costs and hoisting workload. The secondary objective function is to minimize deformation and shorten the construction period.

[0049] The expression for the main objective function is as follows: ; In the formula, f1 is the total counterweight; ω a For the first a The weight of each counterweight m This represents the total number of counterweights.

[0050] The secondary objective function is specifically as follows: Minimize deformation: ; In the formula, f2 is the deformation amount, Δ 允许 The maximum allowable tilt displacement for temporary supports; Shortest construction period: ; In the formula, f3 represents the construction period.

[0051] The expression for the objective function is: ; In the formula, f1 is the total counterweight, f2 is the deformation, and f3 is the construction period. λ 1 is the weighting coefficient of the total amount of counterweight. λ 2 is the weighting coefficient for the deformation amount. λ 3 is the weighting coefficient for the construction period.

[0052] In this step, this embodiment sets dynamic weight coefficients to adapt to changes in different construction stages based on the engineering characteristics of arch bridge counterweight optimization; that is, in the early stage of construction (such as the concrete pouring stage), stress safety is the primary concern (requiring high weight); in the later stage of construction (such as the formwork moving stage), the construction period and deformation control are more important (requiring weight adjustment).

[0053] That is, in the early stages of construction,λ 1 = 0.7, λ 2 = 0.2, λ 3 = 0.1; in the middle of construction, λ 1 = 0.4, λ 2 = 0.3, λ 3 = 0.3; in the later stage of construction, λ 1 = 0.5, λ 2 = 0.1, λ 3 = 0.4.

[0054] This embodiment prioritizes key objectives at different stages through a dynamic weight guiding algorithm, dynamically adjusts the priority of each objective in multi-objective optimization (such as stress safety, deformation control, and construction efficiency) according to the construction progress, and makes the counterweight scheme accurately match the actual needs of the project at different construction stages. This design breaks through the limitations of traditional optimization methods with static weights and global balance, significantly improving the engineering applicability and rationality of the optimization results. Dynamic weight coefficient adjustment through the technical path of "construction stage perception → dynamic weight distribution → multi-objective collaborative optimization" enables the counterweight scheme to focus on the most critical needs at each construction stage: safety in the early stage, deformation and construction period in the middle stage, and cost reduction and efficiency improvement in the later stage. This design not only improves the engineering rationality of the optimization results, but also provides a "precise, flexible, and economical" counterweight strategy for arch bridge construction through dynamic adaptation capability.

[0055] S4.3: According to the multi-constraint conditions and the objective function, a non-dominated sorting genetic algorithm is used to generate a counterweight optimization scheme; In this step, S4.3 is specifically: Initialize the population; Randomly generate N counterweight schemes, each containing m the weight of counterweight blocks to ensure that the initial scheme meets the basic engineering constraints, such as the total weight of the counterweight blocks not exceeding the lifting equipment capacity.

[0056] Multi-objective evaluation and sorting; Calculate the values of the four objective functions (f1, f2, f3, and the target F) for each individual (counterweight scheme), and check the constraint conditions. If the constraint conditions (stress safety constraint, deformation coordination constraint, construction period constraint, and safety margin constraint) are violated, mark it as an infeasible solution and eliminate it.

[0057] According to the Pareto frontier level division, the advantages and disadvantages are sorted, and the crowding distance is calculated.

[0058] The tournament selection method is adopted, and individuals with high non-dominated level and large crowding distance are preferentially reserved to ensure the uniformity of population diversity and front distribution. Simulated binary crossover (SBX) is performed on the selected parent individuals to generate offspring counterweight schemes; Gaussian mutation (small perturbation) is performed on the offspring counterweight schemes to introduce local optimization possibilities.

[0059] Dynamic updating and iteration; The dynamic weight coefficients λ1, λ2, and λ3 are updated according to the current construction progress; the multi-objective evaluation and sorting steps and the selection, crossover, and mutation steps are repeated. The termination condition is that the number of iterations is greater than or equal to 100 generations, or the Pareto front converges, or the calculation time exceeds the limit.

[0060] Optimal strategy extraction; From the final Pareto front, a scheme that meets all the constraint conditions, has the minimum total counterweight, and balances the secondary objectives is selected, and specific counterweight instructions are output through a construction management platform.

[0061] Exemplarily, the specific counterweight instruction is: "add 3 tons of counterweight to the left cantilever end and reduce 1 ton of counterweight to the right side".

[0062] S5: dynamically adjusting the counterweight optimization scheme according to the stress monitoring data and the multi-section stress trend of the arch bridge.

[0063] When any of the following conditions is met, the counterweight adjustment process is started: The maximum stress monitored in real time approaches the warning threshold; The dynamic response prediction model shows that if the current counterweight is maintained, stress will exceed the limit within the next 2 hours; The warning threshold is 80% of the stress design limit.

[0064] The adjustment process is: based on S3, calculate the stress trend under different candidate counterweight schemes; Using S4, the optimal counterweight value that meets all the constraint conditions is selected from the candidate schemes; Adjustment instructions are issued to the site through the construction management platform, and the adjustment process is monitored through unmanned aerial vehicle inspection or camera monitoring to confirm that the counterweight adjustment is in place.

[0065] In embodiment 2, the application also provides an arch bridge construction counterweight optimization system based on stress monitoring, which adopts the arch bridge construction counterweight optimization method based on stress monitoring in embodiment 1, and the system comprises: A data acquisition module is arranged for arranging multiple stress sensors at different positions on the arch bridge and acquiring stress monitoring data; A data processing module is arranged for performing data preprocessing and feature extraction operations on the stress monitoring data; a stress trend prediction module configured to input the features, the construction state parameters, and the structure parameters into a machine learning model to obtain the stress trend of the multi-section of the arch bridge; the machine learning model is a long short-term memory network model with a time attention mechanism layer, and the time attention mechanism layer adopts key time step weighting to highlight the influence of important construction stages on the stress trend; a counterweight optimization module configured to generate an arch bridge construction counterweight optimization scheme by using a non-dominated sorting genetic algorithm; a dynamic adjustment module configured to dynamically adjust the counterweight optimization scheme according to the stress monitoring data and the stress trend of the multi-section of the arch bridge.

[0066] In some embodiments, the electronic device further includes one or more processors and a memory.

[0067] The processor can be a central processing unit (CPU) or other form of processing unit having data processing and / or instruction execution capabilities, and can control other components in the electronic device to perform desired functions.

[0068] The memory can include one or more computer program products, which can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM), cache memory, and / or the like. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, and / or the like. One or more computer program instructions can be stored on the computer-readable storage media, and the processor can execute the program instructions to implement the stress monitoring based arch bridge construction counterweight optimization method of any embodiment of the present application described above and / or other desired functions. Various contents such as initial extrinsic parameters, threshold values, and the like can also be stored in the computer-readable storage media.

[0069] In one example, the electronic device can further include an input device and an output device, which are interconnected by a bus system and / or other forms of connection mechanism (not shown). The input device can include, for example, a keyboard, a mouse, and the like. The output device can output various information to the outside, including pre-warning prompt information, braking force, and the like. The output device can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.

[0070] Of course, components such as buses, input / output interfaces, and the like are omitted for simplicity. In addition, the electronic device can include any other appropriate components according to specific application cases.

[0071] In addition to the above method and device, the embodiments of the present application can also be a computer program product, which comprises computer program instructions, and the computer program instructions enable the processor to realize the functions of the stress monitoring based arch bridge construction counterweight optimization method provided by any embodiment of the present application when the processor runs.

[0072] In addition, the embodiments of the present application can also be a computer readable storage medium, which stores computer program instructions, and the computer program instructions enable the processor to realize the stress monitoring based arch bridge construction counterweight optimization method provided by any embodiment of the present application when the processor runs.

[0073] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present application.

Claims

1. A method for optimizing the counterweight during arch bridge construction based on stress monitoring, characterized in that, Includes the following steps: S1: Multiple stress sensors are arranged at different locations on the arch bridge to collect stress monitoring data; S2: Perform data preprocessing and feature extraction operations on the stress monitoring data; S3: Input the extracted features, construction state parameters, and structural parameters into the machine learning model to obtain the stress trend of the arch bridge's multi-section. The machine learning model is a long short-term memory network model with an added time attention mechanism layer; the time attention mechanism layer uses key time step weighting to highlight the impact of the construction stage on the stress trend. S4: An optimized scheme for the counterweight of arch bridge construction is generated using a non-dominated sorting genetic algorithm. Specifically, S4 is: S4.1: Establish multiple constraints for optimizing the construction counterweight of the arch bridge; The multiple constraints include stress safety constraints, deformation compatibility constraints, construction period constraints, and safety margin constraints. The stress safety constraint is specifically as follows: For each critical section i ( i =1,2,..., n Its maximum tensile stress σ t,i and maximum compressive stress σ c,i Must meet: ； In the formula, n is the total number of critical sections, [ σ t ]and[ σ c These are the design-permissible tensile stress limits and compressive stress limits, respectively. The deformation compatibility constraints are specifically as follows: The maximum tilt displacement Δ at the top of the temporary support must meet the following requirements: ； ； In the formula, H is the support height, Δ 允许 The maximum allowable tilt displacement for temporary supports; The specific time constraints are as follows: Total number of counterweight adjustments during construction N 调整 Must meet: ； In the formula, N 允许 This represents the maximum number of adjustments allowed for the construction period. The safety margin constraint is specifically as follows: The ratio of actual stress to the limit value satisfies: ； In the formula, k is a coefficient; S4.2: Establish the objective function for optimizing the construction counterweight of the arch bridge; The expression for the objective function F is: ； In the formula, f1 is the total counterweight, f2 is the deformation, and f3 is the construction period. λ 1 is the weighting coefficient of the total amount of counterweight. λ 2 is the weighting coefficient for the deformation amount. λ 3 represents the weighting coefficient for the construction period; In the early stages of construction λ 1 = 0.7, λ 2=0.2, λ 3=0.1; During the middle of construction, λ 1 = 0.4, λ 2 = 0.3, λ 3=0.3; In the later stages of construction, λ 1 = 0.5, λ 2=0.1, λ 3 = 0.4; S4.3: Based on the multiple constraints and the objective function, a non-dominated sorting genetic algorithm is used to generate a weight optimization scheme; S5: The counterweight optimization scheme is dynamically adjusted based on the stress monitoring data and the stress trend of the multi-section of the arch bridge.

2. The method for optimizing the counterweight in arch bridge construction based on stress monitoring according to claim 1, characterized in that: The aforementioned time attention mechanism layer employs key time step weighting to highlight the impact of the construction phase on stress trends, specifically as follows: Sa: Calculate the energy fraction at each time step t; Sb: Normalize the energy fraction at each time step t to obtain the attention weight at each time step t; Sc: Generate a context vector based on the attention weights at each time step t. This context vector serves as the input to the fully connected layer and is used to ultimately predict future stress values.

3. The method for optimizing the counterweight in arch bridge construction based on stress monitoring according to claim 2, characterized in that: k is 0.8 or 0.

9.

4. The method for optimizing the counterweight in arch bridge construction based on stress monitoring according to claim 1, characterized in that: In S4.2, the objective function includes a primary objective function and a secondary objective function. The primary objective function is to minimize the total amount of counterweight while satisfying the above constraints. The secondary objective function is to minimize deformation and minimize the construction period.

5. The method for optimizing the counterweight in arch bridge construction based on stress monitoring according to claim 1, characterized in that: In step S1, stress sensors are arranged at key sections of the main arch ring, temporary supports, and counterweight application points of the arch bridge.

6. The method for optimizing the counterweight in arch bridge construction based on stress monitoring according to claim 1, characterized in that: In step S2, the data preprocessing involves using an adaptive filtering algorithm based on wavelet transform to perform data filtering on the stress monitoring data; specifically: The stress monitoring data is subjected to zero-mean normalization to obtain zero-mean normalized data; The zero-mean processed data was subjected to a 7-level discrete wavelet transform using Morlet wavelets to obtain the approximation coefficients A7 and detail coefficients D1~D7. The approximation coefficients and detail coefficients are subjected to combined threshold denoising; specifically: No thresholding is applied to the approximation coefficient layer; For the detail coefficients D 1- D 4. Use a universal threshold T for threshold processing; For the detail coefficients D5-D7, an adaptive soft threshold T is applied. j It retains some weak high-frequency components that are related to the effective signal.

7. A counterweight optimization system for arch bridge construction based on stress monitoring, characterized in that, The system employs the stress monitoring-based counterweight optimization method for arch bridge construction as described in any one of claims 1-6, and the system comprises: The data acquisition module is used to deploy multiple stress sensors at different locations on the arch bridge and collect stress monitoring data; The data processing module is used to perform data preprocessing and feature extraction operations on the stress monitoring data; The stress trend prediction module is used to input the features, construction state parameters and structural parameters into the machine learning model to obtain the stress trend of the multi-section of the arch bridge. The machine learning model is a long short-term memory network model with an added time attention mechanism layer. The time attention mechanism layer uses key time steps to weight the impact of the construction stage on the stress trend. The counterweight optimization module is used to generate an optimized counterweight scheme for arch bridge construction using a non-dominated sorting genetic algorithm. The dynamic adjustment module is used to dynamically adjust the counterweight optimization scheme based on the stress monitoring data and the stress trend of the multi-section of the arch bridge.

Images