A method for evaluating the safety of an initial state of installation of a highway bridge erecting machine

Abstract

This invention relates to the field of safety evaluation technology for highway bridge erecting machines, and discloses a method for safety evaluation of the initial installation state of a highway bridge erecting machine. First, a multimodal safety evaluation model is established, including a structural layer, a dynamic parameter layer, and a risk factor layer. Then, a fuzzy clustering algorithm is used to divide the feature intervals of the dynamic parameter layer, and a grey relational analysis method is used to calculate the correlation coefficient of risk factors. The objective weights of the parameters in the dynamic parameter layer are determined using the entropy weight method, and the two are combined to obtain a comprehensive weight. Next, multi-source sensor data of the bridge erecting machine installation are collected in real time and discretized. Finally, the risk index is calculated layer by layer to determine the safety level of the initial installation state of the bridge erecting machine. This method comprehensively considers the structure, dynamic parameters, and risk factors of the bridge erecting machine, uses multiple algorithms to improve the accuracy of the evaluation, and achieves automatic determination of the safety level. It can effectively reduce the installation risk of the bridge erecting machine and ensure construction safety and the smooth progress of the project.

Description

Technical Field

[0001] This invention relates to the field of safety evaluation technology for highway bridge erecting machines, specifically a method for safety evaluation of the initial installation state of a highway bridge erecting machine. Background Technology

[0002] In modern highway construction, bridge erecting machines are widely used as key equipment in bridge erection projects. With the continuous expansion of highway bridge construction scale and the increasing demands for construction efficiency and quality, the frequency and complexity of bridge erecting machine usage are also constantly rising. However, in the initial installation phase, due to their complex structure, numerous installation stages, and various dynamic parameter changes, bridge erecting machines pose high safety risks. Once a safety accident occurs, it will not only cause serious casualties and property losses but also have a significant impact on the progress and cost of the entire highway construction project.

[0003] Currently, traditional safety assessment methods for highway bridge erecting machines are relatively simplistic, relying mostly on manual experience and basic equipment inspections. This approach has significant limitations: firstly, manual experience is highly subjective, with varying judgment standards among different personnel, making it difficult to accurately assess complex safety conditions; secondly, simple equipment inspections can only identify some superficial and obvious problems, failing to effectively monitor and assess potential risks within the equipment and safety hazards during dynamic changes.

[0004] From the perspective of the bridge erecting machine's structure, it consists of multiple complex components, such as the main beam, outriggers, and supports. These components have close physical topological relationships and are interconnected. During installation, changes in the state of any component can trigger a chain reaction, affecting overall safety. For example, inaccurate positioning of the main beam may lead to deviations during subsequent beam erection; problems with the synchronization of the outriggers can cause uneven stress on the entire machine; and insufficient load-bearing capacity of the supports may lead to localized structural damage. However, existing safety assessment methods often fail to fully consider the complex relationships between these components and the resulting safety risks.

[0005] In terms of dynamic parameter monitoring, various dynamic parameters are generated during the installation of bridge erecting machines, such as displacement deviation, angular offset, and stress distribution. These parameters change continuously over time, reflecting the real-time status of the bridge erecting machine during installation. However, existing monitoring technologies struggle to accurately and comprehensively collect and analyze these multi-dimensional dynamic parameters, failing to promptly detect safety risks indicated by abnormal parameter changes. Furthermore, there is a lack of scientifically effective methods for defining the characteristic intervals of these dynamic parameters, resulting in a lack of accurate basis for assessing safety status.

[0006] Risk factor assessment also presents challenges. Due to the lack of a comprehensive historical fault database, it is difficult to accurately determine the critical threshold ranges for each module, hindering the effective identification of potential risk factors. Traditional methods also suffer from low accuracy in calculating the correlation coefficients of risk factors and determining the weights of each parameter, resulting in the final safety evaluation failing to accurately reflect the actual safety status of the bridge erecting machine in its initial installation state. Summary of the Invention

[0007] The purpose of this invention is to provide a method for safety evaluation of the initial installation state of a highway bridge erecting machine to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a method for safety evaluation of the initial installation state of a highway bridge erecting machine, the method comprising:

[0009] S1. Establish a multimodal safety evaluation model for highway bridge erecting machines, including structural layer, dynamic parameter layer and risk factor layer;

[0010] S2. Divide the multi-dimensional feature intervals of the dynamic parameter layer based on the fuzzy clustering algorithm;

[0011] S3. Use grey relational analysis to calculate the correlation coefficient of each parameter in the risk factor layer;

[0012] S4. Determine the objective weights of each parameter in the dynamic parameter layer using the entropy weight method to obtain the entropy weights;

[0013] S5. Combine the correlation coefficient and entropy weight to fuse parameters and generate a comprehensive weight for the dynamic parameter layer.

[0014] S6. Real-time acquisition of multi-source sensor data during the installation process of the bridge erecting machine and discretization processing;

[0015] S7. Calculate the risk index layer by layer to determine the safety level of the initial installation state of the highway bridge erecting machine.

[0016] Preferably, in step S1:

[0017] The structural layer is divided into a main beam positioning module, a leg synchronization module, a support bearing module, and a whole machine balance module based on the physical topology of the bridge erecting machine's installation components.

[0018] The dynamic parameter layer collects displacement deviation, angle offset, and stress distribution parameters of each module through a sensor network.

[0019] The risk factor layer extracts the critical threshold range of each module based on the historical failure database.

[0020] Preferably, the step S2, which involves dividing the feature intervals based on a fuzzy clustering algorithm, includes:

[0021] The fuzzy membership function is set according to the parameter distribution characteristics of the dynamic parameter layer;

[0022] The similarity matrix between different parameters is calculated using an iterative optimization algorithm;

[0023] The dynamic parameter layer is clustered and grouped based on the similarity matrix.

[0024] Preferably, the implementation steps of the grey relational analysis method in step S3 include:

[0025] Select a benchmark reference sequence for each parameter in the risk factor layer;

[0026] Calculate the correlation coefficients between each parameter sequence and the benchmark reference sequence;

[0027] Generate a correlation coefficient matrix based on the mean of the correlation coefficients.

[0028] Preferably, the objective weight determination step of the entropy weight method in step S4 includes:

[0029] The historical dataset of the dynamic parameter layer is transformed by probability density function, and the information entropy value and difference coefficient of each parameter are calculated. The objective weights of the dynamic parameter layer are generated by normalizing the difference coefficients.

[0030] Preferably, the parameter fusion step in step S5 includes:

[0031] The correlation coefficient is used as a subjective weighting factor, and the entropy weight is used as an objective weighting factor.

[0032] A Bayesian network model is used to probabilistically fuse subjective and objective weights, and a comprehensive weight for the dynamic parameter layer is generated through the posterior probability distribution.

[0033] Preferably, the discretization process in step S6 includes:

[0034] Anomaly fluctuation detection and noise filtering are performed on multi-source sensor data. Data with different dimensions are converted into a unified discrete interval through the range standardization method, and a discrete distribution map of dynamic parameters is generated according to the time series.

[0035] Preferably, the calculation of the risk index in step S7 includes:

[0036] The risk index for each structural layer is generated by multiplying the comprehensive weight of the dynamic parameter layer with the discretized data.

[0037] A total risk score is generated by superimposing the preset structural layer weight coefficients and risk index, and safety levels are divided based on the score range.

[0038] Preferably, the dynamic parameter layer includes at least one of the following parameters:

[0039] The longitudinal displacement deviation, lateral torsion angle, and track levelness of the main beam positioning module;

[0040] Hydraulic pressure difference, extension / retraction speed synchronization rate, and vertical tilt angle of the outrigger synchronization module;

[0041] The difference in support force between the front and rear outriggers of the overall balance module and the symmetry of the transverse track.

[0042] Preferably, the step of generating the discretized distribution map includes:

[0043] The standardized sensor data is divided into equal-length data segments according to time windows. The wavelet packet decomposition algorithm is used to extract the frequency domain energy features of each data segment. The feature difference matrix of the data segments in different time windows is calculated by Mahalanobis distance.

[0044] Compared with the prior art, the beneficial effects of the present invention are:

[0045] At the model construction level, a multimodal safety evaluation model for highway bridge erecting machines was established, comprising a structural layer, a dynamic parameter layer, and a risk factor layer. The structural layer is meticulously divided based on the physical topology of the bridge erecting machine's installation components, covering the main beam positioning module, outrigger synchronization module, support bearing module, and overall machine balance module, comprehensively and accurately reflecting the structural state of each key part of the bridge erecting machine. The dynamic parameter layer collects parameters such as displacement deviation, angular offset, and stress distribution of each module through a sensor network, enabling real-time monitoring of the bridge erecting machine's dynamic operating characteristics. The risk factor layer extracts the critical threshold ranges of each module based on a historical fault database, providing a scientific and accurate reference standard for risk assessment. This multimodal model integrates the bridge erecting machine's structure, dynamic operating parameters, and historical fault information, overcoming the shortcomings of traditional evaluation methods that only focus on a single factor, and greatly improving the comprehensiveness and accuracy of the safety evaluation.

[0046] In terms of algorithm application, a fuzzy clustering algorithm is used to divide the multi-dimensional feature intervals of the dynamic parameter layer. A fuzzy membership function is set according to the distribution characteristics of the dynamic parameters, and then an iterative optimization algorithm is used to calculate the similarity matrix and cluster the data. This effectively handles the complexity and uncertainty of the dynamic parameters, reasonably distinguishes the parameter characteristics under different operating states, and lays a solid foundation for subsequent accurate assessment of safety risks. Grey relational analysis is used to calculate the correlation coefficients of each parameter in the risk factor layer. By selecting a benchmark reference sequence, calculating the correlation coefficients, and generating a correlation coefficient matrix, the degree of correlation between each risk factor can be accurately analyzed, key risk factors can be identified, and the evaluation results can be made more targeted. The entropy weight method is used to determine the objective weights of each parameter in the dynamic parameter layer. By performing probability density function transformation on the historical dataset, calculating the information entropy value and difference coefficient, and normalizing them, the information inherent in the data is fully utilized, avoiding the arbitrariness of subjective weighting and ensuring the scientific and objective nature of the weight determination.

[0047] In the parameter fusion stage, the correlation coefficient and entropy weight are combined for parameter fusion. The correlation coefficient is used as a subjective weighting factor, and the entropy weight as an objective weighting factor. A Bayesian network model is used for probabilistic fusion to generate a comprehensive weight. This fusion method takes into account both expert experience and data-driven approaches, effectively integrating subjective and objective information, making the evaluation results more consistent with reality.

[0048] In the data processing and evaluation process, multi-source sensor data is collected in real time during the bridge erecting machine installation and then discretized. First, abnormal fluctuations are detected and noise is filtered. Then, the data dimensions are unified using the range standardization method to generate a discretized distribution map, ensuring the data input to the evaluation model is authentic, reliable, and easy to analyze. The risk index is calculated layer by layer to determine the safety level. The comprehensive weight of the dynamic parameter layer is multiplied by the discretized data to obtain the risk index for each structural layer. Combined with the structural layer weight coefficients, a total risk score is generated and safety levels are assigned. This achieves full automation and standardization from data acquisition and processing to safety evaluation, enabling timely and accurate reflection of the safety status of the bridge erecting machine at the initial installation stage. It provides construction personnel with intuitive and effective safety warnings, helping them to take timely measures to eliminate safety hazards, greatly reducing safety risks during bridge erecting machine installation, and ensuring the safety of construction personnel and the smooth progress of the project. Simultaneously, the application of this method also helps improve the installation quality and construction efficiency of the bridge erecting machine, reduces project construction costs, and is of great significance for promoting technological progress and sustainable development in the highway bridge construction industry. Attached Figure Description

[0049] Figure 1 This is a schematic diagram illustrating the working principle of the initial state safety evaluation method for highway bridge erecting machine installation described in this invention.

[0050] Figure 2A flowchart illustrating the steps involved in generating the correlation coefficient matrix using grey relational analysis.

[0051] Figure 3 A diagram illustrating the working principle of generating a comprehensive weight by combining the correlation coefficient and entropy weight;

[0052] Figure 4 This is a flowchart for discretizing multi-source sensor data. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] Please see Figures 1-4 This invention provides a method for safety evaluation of the initial installation state of a highway bridge erecting machine, the specific implementation steps of which are as follows:

[0055] A multimodal safety evaluation model was constructed, comprising a structural layer, a dynamic parameter layer, and a risk factor layer. The structural layer is divided according to the physical topology of the bridge erecting machine's installation components; the dynamic parameter layer collects relevant parameters using a sensor network; and the risk factor layer extracts critical threshold ranges based on a historical fault database.

[0056] Based on the parameter distribution characteristics of the dynamic parameter layer, a fuzzy membership function is set, and an iterative optimization algorithm is used to calculate the similarity matrix between different parameters. Then, the dynamic parameter layer is clustered and grouped according to the similarity matrix.

[0057] Select a benchmark reference sequence for each parameter in the risk factor layer, calculate the correlation coefficient between each parameter sequence and the benchmark reference sequence, and generate a correlation coefficient matrix based on the mean of the correlation coefficients.

[0058] The historical dataset of the dynamic parameter layer is transformed by probability density function, and the information entropy value and difference coefficient of each parameter are calculated. The objective weights of the dynamic parameter layer are generated by normalizing the difference coefficients.

[0059] The correlation coefficient is used as a subjective weight factor, and the entropy weight is used as an objective weight factor. A Bayesian network model is used to probabilistically fuse the subjective and objective weights, and the comprehensive weight of the dynamic parameter layer is generated through the posterior probability distribution.

[0060] Multi-source sensor data is collected in real time during the bridge erecting machine installation process. Abnormal fluctuations and noise filtering are performed on the data. Data with different dimensions are converted into a unified discrete interval through the range standardization method. Discrete distribution map of dynamic parameters is generated according to the time series.

[0061] The dynamic parameter layer comprehensive weight is multiplied by the discretized data to generate the risk index of each structural layer; the total risk score is generated by superimposing the preset structural layer weight coefficients and risk indices, and the safety level is divided based on the score interval.

[0062] The present invention will be further described below with reference to Examples 1 to 5:

[0063] Example 1:

[0064] When establishing a multimodal safety evaluation model for highway bridge erecting machines, the structural layer is divided based on the physical topological relationships of the machine's installation components. These components have complex physical connections and collaborative working relationships. The main beam positioning module is crucial for ensuring the precise positioning of the main beam during installation. Its stability and accuracy directly affect the quality and safety of subsequent bridge erection. In actual installation, deviations in the main beam's position can lead to uneven stress on the bridge, creating potential safety hazards. The outrigger synchronization module is primarily responsible for the synchronization of multiple outriggers during operation. If the outriggers are not synchronized, the bridge erecting machine will experience stress imbalance during support and movement, potentially leading to overturning accidents. The support bearing module bears the weight of the entire bridge erecting machine and the bridge to be erected; its load-bearing capacity and stability are paramount. Problems with the support bearing can cause the bridge erecting machine to collapse. The overall machine balancing module ensures the balance of the bridge erecting machine during installation, comprehensively considering the stress and positional relationships of each component to ensure stable operation.

[0065] The dynamic parameter layer collects displacement deviation, angular offset, and stress distribution parameters of each module through a sensor network. In the main beam positioning module, parameters such as longitudinal displacement deviation, lateral torsion angle, and track levelness need to be accurately collected. These parameters reflect the real-time position and attitude changes of the main beam. In the outrigger synchronization module, parameters such as hydraulic pressure difference, telescopic speed synchronization rate, and vertical tilt angle are collected to monitor whether the outriggers are working in sync. For the overall machine balance module, parameters such as the difference in support force between the front and rear outriggers and the symmetry of the lateral track are collected to provide a basis for judging the overall machine balance. The risk factor layer extracts the critical threshold ranges of each module based on a historical fault database. Through the analysis and organization of a large amount of historical fault data, the safe parameter range of each module under different working conditions is determined. When the collected real-time parameters exceed this range, it means that there is a certain safety risk. When building the model, it is necessary to ensure smooth data interaction between each layer to accurately reflect the actual situation during the bridge erecting machine installation process.

[0066] In the installation scenario of highway bridge erecting machine, taking the installation project of a certain type of highway bridge erecting machine as an example, the bridge erecting machine is used for the construction of a cross-river highway bridge. The bridge is 500 meters long and consists of 20 precast box girders, each weighing about 80 tons.

[0067] Before actual construction, the structural layers were divided according to the physical topology of the bridge erecting machine's installation components. The main girder of the bridge erecting machine is a key component that supports the precast box girder and enables its hoisting and installation; its installation position accuracy directly affects the overall structural stability of the bridge. In this project, the main girder positioning module includes devices for accurately measuring the longitudinal, lateral, and height positions of the main girder. Longitudinal positioning utilizes a high-precision laser rangefinder, installed on one side of the bridge erecting machine's track, to monitor the distance deviation between the front end of the main girder and the predetermined installation position in real time. Lateral positioning uses horizontal displacement sensors installed at both ends of the main girder to monitor lateral offset. For height adjustment, multiple high-precision pressure sensors are installed at the main girder support points, and pressure changes indicate whether the main girder height is within the design requirements.

[0068] The outrigger synchronization module is crucial for the stable operation of the bridge erecting machine. This machine uses four hydraulic outriggers, and the module primarily monitors hydraulic system parameters and outrigger movement status. Each outrigger's hydraulic cylinder is equipped with a pressure sensor to collect real-time hydraulic pressure differences and determine if the force on each outrigger is balanced. Simultaneously, displacement sensors are installed on the outrigger extension and retraction mechanisms to calculate the synchronization rate of extension and retraction speeds, ensuring coordinated and consistent movement of each outrigger. Furthermore, tilt sensors monitor the vertical tilt angle of the outriggers to prevent excessive tilting that could lead to instability of the bridge erecting machine.

[0069] The support structure's load-bearing module is crucial for supporting the bridge erecting machine's own weight and the load of the hoisted box girders. Stress sensors are installed at key stress-bearing locations, such as the connections between columns and crossbeams, and the bottom support points, to monitor the magnitude and distribution of stress on the support in real time. These sensors transmit the collected data to the control system in real time. If the stress approaches or exceeds the design tolerance, the system immediately issues an alarm, prompting operators to take appropriate measures to prevent damage to the support due to overload. The overall balance module ensures the stability of the bridge erecting machine. By installing weighing sensors on the front and rear outriggers, the difference in support force between the front and rear outriggers is measured to determine whether the weight distribution of the bridge erecting machine is balanced. Simultaneously, laser alignment devices are installed on both sides of the transverse track to monitor the symmetry of the transverse track, ensuring the bridge erecting machine remains balanced during transverse movement and preventing skewing that could lead to safety accidents.

[0070] During the installation of the bridge erecting machine, a sensor network is used to collect displacement deviations, angular offsets, and stress distribution parameters of each module. For the main beam positioning module, the longitudinal displacement deviation is collected every 10 seconds by a laser rangefinder, recording the deviation of the main beam in the longitudinal direction relative to the ideal installation position; the lateral torsion angle is obtained using dual-axis tilt sensors installed at both ends of the main beam, which can measure the change in the torsion angle of the main beam on the horizontal plane in real time; the track levelness is monitored by multiple levels installed on the track, with one level installed every 5 meters, collecting the height difference data of the track, and then calculating the track levelness deviation.

[0071] Regarding the outrigger synchronization module, the hydraulic pressure difference is collected in real time by pressure sensors installed on the hydraulic cylinders of each outrigger, with data collected once per second, so as to promptly detect pressure differences between outriggers; the telescopic speed synchronization rate is calculated by collecting displacement data during the telescopic process of the outriggers through displacement sensors, combined with time intervals, and recorded once every 20 seconds; the vertical tilt angle is continuously monitored by tilt sensors installed on the outriggers, with data collection frequency of once per second, to ensure timely monitoring of the outrigger tilt.

[0072] In the overall balance module, the difference in support force between the front and rear outriggers is measured in real time by weighing sensors installed at the bottom of the front and rear outriggers, with data collected every 15 seconds; the symmetry of the transverse track is monitored by a laser alignment device, which continuously emits laser signals, and the receiving end receives and analyzes the signals in real time to calculate the track symmetry deviation, recording data every 30 seconds.

[0073] Based on the historical fault database of this type of bridge erecting machine, as well as design standards and construction specifications, the critical threshold ranges for each module were extracted. For the main beam positioning module, the critical threshold for longitudinal displacement deviation is set at ±50 mm. If this range is exceeded, it may lead to excessive deviation in the installation position of the box girder, affecting the stress on the bridge structure. The critical threshold for lateral torsion angle is ±0.5°. Exceeding this angle will cause uneven stress on the main beam, posing a safety hazard. The critical threshold for track levelness is ±3 mm / m. Exceeding this range will affect the stability of the bridge erecting machine's movement.

[0074] In the outrigger synchronization module, the critical threshold for hydraulic pressure difference is set at ±0.5MPa. Excessive pressure difference will cause uneven force on the outriggers, leading to tilting of the bridge erecting machine. The critical threshold for telescopic speed synchronization rate is ±5%. Exceeding this range will cause uncoordinated outrigger movements. The critical threshold for vertical tilt angle is ±1°. Excessive tilt angle will easily cause instability of the bridge erecting machine.

[0075] In the overall balance module, the critical threshold for the difference in support force between the front and rear outriggers is set at ±10 tons. Exceeding this range will affect the overall balance of the bridge erecting machine. The critical threshold for the symmetry of the lateral track is ±5 millimeters. Excessive symmetry deviation may lead to jamming or even derailment during the lateral movement of the bridge erecting machine. Throughout the installation process of the bridge erecting machine, data from each module is collected in real time and transmitted to the control system. The control system compares and analyzes the collected data with the critical thresholds set in the risk factor layer, providing basic data support for subsequent safety assessments and ensuring that the initial installation state of the bridge erecting machine is safe and controllable.

[0076] Example 2:

[0077] The dynamic parameter layer is divided into multi-dimensional feature intervals based on fuzzy clustering algorithm, and fuzzy membership functions are set according to the parameter distribution characteristics of the dynamic parameter layer. The parameters of the dynamic parameter layer have different distribution characteristics; some parameters may exhibit a normal distribution, while others may exhibit a skewed distribution. These characteristics need to be fully considered when setting the fuzzy membership function. For displacement deviation parameters, their changes may be relatively continuous and have a certain fluctuation range; a Gaussian fuzzy membership function can be used to describe the degree to which they belong to a certain feature category within different value ranges. For angle offset parameters, since their value range is usually limited, a triangular or trapezoidal fuzzy membership function may be more suitable. The similarity matrix between different parameters is calculated using an iterative optimization algorithm. Common iterative optimization algorithms such as genetic algorithms and particle swarm optimization algorithms can be used. Taking a genetic algorithm as an example, a set of initial solutions is first randomly generated, and these solutions represent the similarity relationships between different parameters. Then, through operations such as selection, crossover, and mutation, these solutions are continuously optimized so that the generated similarity matrix can more accurately reflect the intrinsic relationship between the parameters. In the selection operation, the solution with higher fitness is selected to enter the next generation based on the fitness value of the solution; the crossover operation is to exchange some information between two solutions to generate a new solution; the mutation operation randomly changes some elements in the solution with a certain probability to prevent the algorithm from getting trapped in local optima.

[0078] The dynamic parameter layer is clustered based on its similarity matrix. Once the similarity matrix is ​​obtained, the parameters can be clustered according to the values ​​in the matrix. Hierarchical clustering algorithms can be used, starting with each parameter as a separate cluster and gradually merging highly similar clusters until a certain clustering stopping condition is met. The clustering stopping condition could be that the number of clusters reaches a preset value, or the similarity within a cluster reaches a certain threshold. In this way, the dynamic parameter layer is divided into different feature intervals, providing a more targeted data processing foundation for subsequent security evaluation.

[0079] In a large-scale highway bridge construction project, a specific model of highway bridge erecting machine was used for box girder erection. The initial safety assessment of this machine involved dividing the dynamic parameter layers into multi-dimensional feature intervals using a fuzzy clustering algorithm.

[0080] The dynamic parameter layer of this bridge erecting machine includes numerous parameters, such as the longitudinal displacement deviation, lateral torsion angle, and track levelness of the main beam positioning module, and the hydraulic pressure difference, telescopic speed synchronization rate, and vertical tilt angle of the outrigger synchronization module. These parameters exhibit different patterns of change during the bridge erecting machine's installation process. To achieve a more accurate safety assessment, it is necessary to rationally divide them into characteristic intervals.

[0081] First, a fuzzy membership function is defined based on the parameter distribution characteristics of the dynamic parameter layer. Taking longitudinal displacement deviation as an example, through extensive collection and analysis of longitudinal displacement deviation data from similar bridge erecting machine installation projects in the past, it was found that the data distribution approximates a normal distribution. Therefore, a Gaussian fuzzy membership function is used to describe the degree to which the longitudinal displacement deviation belongs to different feature categories. Let the longitudinal displacement deviation be... The Gaussian fuzzy membership function can be expressed as: ,in It is the center value of the function, determined based on the mean of historical data; The standard deviation reflects the dispersion of the data. In practical applications, the mean is calculated from the longitudinal displacement deviation data collected in the early stages of this project. (Unit: mm), standard deviation Thus, when the collected longitudinal displacement deviation is When the value is in millimeters, substituting it into the function yields the membership value of a certain feature category, which can be used to determine the relative degree of the displacement deviation in the safety assessment.

[0082] For the lateral torsion angle, since its range is limited and its variation is relatively gradual, a triangular fuzzy membership function is more suitable. Assume the range of the lateral torsion angle is... Based on engineering experience and safety standards, it is divided into three characteristic categories: safe range, warning range, and hazardous range. When the angle is... The time is within the safe range, and the corresponding triangular fuzzy membership function takes values ​​within this interval. ;exist and To define the warning range, the function value starts from... linearly decreasing to ;exist and Within the danger zone, the function value is When the actual lateral torsion angle collected is At that time, this function can be used to determine the degree to which a function belongs to the warning range, providing a basis for safety assessment.

[0083] Next, an iterative optimization algorithm is used to calculate the similarity matrix between different parameters. Here, Particle Swarm Optimization (PSO) is chosen. PSO is an optimization algorithm that simulates the foraging behavior of bird flocks. Each particle represents a potential solution and continuously adjusts its position in the solution space to find the optimal solution. In this application scenario, each particle represents an assumption about the similarity relationship between different parameters.

[0084] During algorithm initialization, random generation is performed. Each particle has a position vector containing the similarity values ​​for all parameter pairs for which similarity is to be calculated. For example, for the parameters longitudinal displacement deviation and hydraulic pressure difference, an element in the particle's position vector represents the assumed similarity value between them. Each particle is also assigned an initial velocity.

[0085] During the iteration process, particles adjust their speed and position based on their own historical best position and the group's historical best position. At each iteration, the objective function value is calculated for each particle based on the similarity relationships it represents. The objective function can be defined as maximizing the sum of similarities within each cluster and minimizing the sum of similarities between clusters after clustering all collected parameter data according to these similarity relationships. Through continuous iteration, the particles gradually move closer to the optimal solution. After several iterations, a relatively stable similarity matrix was obtained. For example, the final similarity matrix shows that the similarity between longitudinal displacement deviation and lateral torsion angle is... This indicates a certain correlation in the changing trends of these two parameters; while the similarity between the longitudinal displacement deviation and the synchronization rate of the expansion and contraction speed is... This indicates that the correlation between them is relatively weak.

[0086] Finally, the dynamic parameter layer is grouped by feature clustering based on the similarity matrix. A hierarchical clustering algorithm is used to achieve this. Hierarchical clustering is a cluster-based algorithm, which has two modes: agglomerative clustering (bottom-up) and divisive clustering (top-down). Here, agglomerative clustering is used.

[0087] Initially, each parameter is treated as a separate class. Based on the similarity matrix, the similarity between any two classes is calculated. For example, longitudinal displacement deviation and lateral torsional angle have high similarity, so they are first merged into a new class. Then, the similarity between the new class and other classes is recalculated, and high-similarity classes are merged again. During the merging process, information from each merge is recorded, forming a tree-structured clustering hierarchy. When the number of classes reaches a preset value, such as a preset division... When there are 1 class, the clustering process stops. Ultimately, the dynamic parameter layer is divided into 10 classes. The system is divided into several feature intervals: parameters related to the main beam position (including longitudinal displacement deviation, lateral torsion angle, etc.), outrigger hydraulic parameters (including hydraulic pressure difference, etc.), outrigger motion synchronization parameters (including telescopic speed synchronization rate, etc.), outrigger tilt parameters (including vertical tilt angle, etc.), and other comprehensive parameter intervals (including some parameters related to overall safety but with weaker correlation to the above categories). This feature clustering grouping makes subsequent analysis and safety evaluation of the dynamic parameter layer more targeted and efficient, enabling a more accurate assessment of the safety of the bridge erecting machine in its initial installation state.

[0088] Example 3:

[0089] Grey relational analysis was used to calculate the correlation coefficients of each parameter in the risk factor layer, and a benchmark reference sequence was selected for each parameter in the risk factor layer. The risk factor layer contains multiple parameters, and a benchmark reference sequence needs to be determined for each parameter. This benchmark reference sequence can be an ideal value sequence determined based on historical experience, design standards, or safety specifications. For the stress distribution parameters of the main beam, the benchmark reference sequence can be a sequence composed of the theoretical stress values ​​of various parts of the main beam under ideal working conditions. The correlation coefficients between each parameter sequence and the benchmark reference sequence were calculated. The correlation coefficients were calculated by comparing the differences between the actually collected parameter sequences and the benchmark reference sequences. The smaller the difference, the larger the correlation coefficient, indicating that the parameter is closer to the ideal state and the safety risk is relatively low; conversely, the larger the difference, the smaller the correlation coefficient, and the higher the safety risk. Factors such as the length of the parameter sequence and the trend of data change need to be considered when calculating the correlation coefficients. A correlation coefficient matrix was generated based on the average correlation coefficient. The correlation coefficients of each parameter were averaged to obtain an average value representing the degree of correlation between the parameter and the benchmark reference sequence. The average correlation coefficients of all parameters were combined to form the correlation coefficient matrix. This matrix can intuitively reflect the correlation between each parameter in the risk factor layer and the ideal state, providing an important basis for subsequent weight determination and safety evaluation.

[0090] Taking a highway bridge erecting machine used in a highway bridge construction project as an example, this machine is used to erect 30-meter prestressed concrete box girders, and its risk factors need to be assessed in the initial stage of installation.

[0091] Among the numerous parameters in the risk factor layer, the main girder stress is selected as a key parameter for analysis. First, a benchmark reference sequence for the main girder stress must be determined. By consulting the design documents of the bridge erecting machine, the theoretical stress values ​​at different locations on the main girder are obtained under standard working conditions, i.e., when the bridge erecting machine is in a stable installation state and lifting the rated weight box girder. For example, at the mid-span of the main girder, the standard stress value required by design should be stably maintained at around 200 MPa during installation; near the support point, the stress value is approximately 120 MPa. These theoretical stress values ​​are arranged in a certain order to form a benchmark reference sequence, assumed to be... The ellipsis here represents the theoretical stress values ​​at other key locations on the main beam.

[0092] Calculate the correlation coefficients between each parameter sequence and the benchmark reference sequence. During the bridge erection machine installation process, stress sensors installed at key locations on the main beam collect actual stress data at regular intervals (e.g., every 5 minutes) to form an actual stress sequence. For example, if the stress at the mid-span of the main beam is 210 MPa and the stress near the support point is 130 MPa at a certain moment, the actual stress sequence at that moment can be obtained. When calculating the correlation coefficient, the difference between corresponding points in the actual sequence and the reference sequence needs to be considered. For the mid-span of the main beam, the difference between the actual stress and the reference stress... , here This represents the difference between the first data point in the first actual sequence and the corresponding data point in the benchmark reference sequence. After calculating the differences between all corresponding points, the maximum value among all differences is found. and minimum value Introducing the resolution coefficient (The value is generally between 0 and 1, here it is taken as...) The correlation coefficient for each corresponding point is calculated using the correlation coefficient formula. Taking the mid-span position of the main beam as an example, the correlation coefficient... Substitute the previously calculated values ​​into the equation, assuming... , ,but Following the same method, the correlation coefficient between each point in the actual stress sequence and the corresponding point in the reference sequence was calculated, resulting in a correlation coefficient sequence. The ellipsis here represents the correlation coefficient in other positions.

[0093] A correlation coefficient matrix is ​​generated based on the average correlation coefficient. The average correlation coefficient of each actual stress sequence is then calculated to obtain the correlation coefficient between that sequence and the reference sequence. For example, for the preceding actual stress sequences... Its correlation coefficient sequence The average value is 0.7, which is the correlation coefficient between the actual stress sequence and the reference sequence. In practice, multiple actual stress sequences at different times are collected, such as... , And so on, and calculate their correlation coefficients with the benchmark reference sequence. Arrange all these correlation coefficients in a certain order to form a correlation coefficient matrix. Assume that after multiple data collections and calculations, the correlation coefficient matrix is ​​obtained. Each row in the matrix represents the correlation coefficient between an actual stress sequence and a reference sequence, and each column represents the correlation coefficient corresponding to data collected at different times. This correlation coefficient matrix provides a clear understanding of the degree of correlation between the actual stress of the main beam and the standard stress at different times. The closer the correlation coefficient is to 1, the closer the actual stress is to the stress under standard working conditions, and the relatively lower the safety risk. Conversely, the smaller the correlation coefficient, the greater the deviation of the actual stress from the standard working conditions, and the higher the safety risk.

[0094] Example 4:

[0095] The objective weights of each parameter in the dynamic parameter layer are determined using the entropy weight method, and the historical dataset of the dynamic parameter layer is transformed using a probability density function. The dynamic parameter layer has accumulated a large amount of historical data, which reflects the parameter changes of the bridge erecting machine under different installation conditions.

[0096] The historical data is transformed using probability density functions to represent the data's distribution characteristics in probabilistic form. Methods such as kernel density estimation can be used to estimate the probability density function of each parameter. The probability density function reveals the probability of a parameter occurring within different value ranges. The information entropy and coefficient of variation are then calculated for each parameter. Based on the probability density function, the information entropy value for each parameter is calculated. The information entropy value measures the degree of uncertainty in the parameter data; a higher information entropy value indicates a more dispersed data set for that parameter, potentially suggesting its greater importance in security assessment.

[0097] The difference coefficient is calculated, reflecting the degree of difference between different parameters. The objective weights of the dynamic parameter layer are generated by normalizing the difference coefficients. The difference coefficients are then normalized again so that the sum of the weights of all parameters is 1. This yields the objective weights of each parameter in the dynamic parameter layer. These objective weights reflect the relative importance of each parameter in the safety assessment, providing an objective basis for subsequent parameter fusion and safety evaluation.

[0098] In a bridge construction project of a newly built highway, a specific model of highway bridge erecting machine was used. When conducting a safety evaluation of its initial installation state, it is necessary to determine the objective weights of each parameter in the dynamic parameter layer using the entropy weight method. The implementation process is described in detail below based on the actual situation.

[0099] The dynamic parameter layer of this bridge erecting machine involves numerous parameters, such as the longitudinal displacement deviation of the main beam positioning module and the hydraulic pressure difference of the outrigger synchronization module. Before determining the objective weights, a large amount of historical data from the bridge erecting machine installation process was accumulated. This data was collected in several similar bridge construction projects using the same model of bridge erecting machine, covering parameter changes under different construction environments and operating procedures; the data is abundant and representative.

[0100] The historical dataset of the dynamic parameter layer is processed and converted into a probabilistic form to reflect the distribution characteristics of the parameters. Taking longitudinal displacement deviation data as an example, thousands of sets of longitudinal displacement deviation data were collected from past construction records. This data was organized, and the frequency of different displacement deviation values ​​was statistically analyzed. For example, during the statistical process, it was found that the displacement deviation occurred 200 times in the 10-20 mm range, and 150 times in the 20-30 mm range, etc. By calculating the proportion of the frequency of data in each range to the total data volume, the probability of longitudinal displacement deviation occurring in that range is approximated. In this way, the historical data of longitudinal displacement deviation is presented in a probabilistic form, allowing us to intuitively understand the likelihood of different displacement deviation values ​​occurring in actual construction.

[0101] Calculate the information entropy and difference coefficient for each parameter. For longitudinal displacement deviation, calculate the information entropy based on the previously obtained probability distribution. Information entropy can be understood as an indicator of the uncertainty or dispersion of data. If the longitudinal displacement deviation data is relatively concentrated, meaning most data is concentrated within a small range, then its information entropy value is small, indicating that the parameter's change is relatively stable. Conversely, if the data is relatively dispersed, distributed over a large range, then the information entropy value is large, meaning the parameter's uncertainty is high. During the calculation, the probability and number of each displacement deviation interval are considered to determine the information entropy value. Simultaneously, calculate the difference coefficient, which reflects the degree of difference between different parameters. When comparing the longitudinal displacement deviation and hydraulic pressure difference, determine their difference coefficients by analyzing their respective probability distributions and data fluctuation ranges. The larger the difference coefficient, the more significant the difference between the two parameters in the safety assessment, and the different their impacts on the assessment results.

[0102] Objective weights for the dynamic parameter layer are generated through difference coefficient normalization. The difference coefficients of all calculated parameters are processed to meet the weighting requirement, i.e., the sum of the weights of all parameters is 1. In this process, the objective weight of each parameter in the safety evaluation is determined according to a certain proportion based on the magnitude of its difference coefficient. For example, after calculation and normalization, the objective weight of longitudinal displacement deviation is determined to be 0.3, and the objective weight of hydraulic pressure difference is determined to be 0.25, etc. These objective weights reflect the relative importance of each parameter in the safety evaluation; the larger the objective weight, the more critical the parameter's role in evaluating the safety of the bridge erecting machine in its initial installation state. In subsequent safety evaluation processes, different parameters can be reasonably weighted based on these objective weights to more accurately assess the safety of the bridge erecting machine in its initial installation state.

[0103] Example 5:

[0104] By combining the correlation coefficient and entropy weight, a comprehensive weight for the dynamic parameter layer is generated. The correlation coefficient serves as a subjective weight factor, while the entropy weight serves as an objective weight factor. The correlation coefficient reflects the degree of correlation between the parameter and the ideal state, embodying a certain degree of subjective judgment; therefore, it can be used as a subjective weight factor. The entropy weight, calculated based on the objective distribution characteristics of the data, represents the objective importance of the parameter and serves as an objective weight factor. A Bayesian network model is used to probabilistically fuse the subjective and objective weights, generating the comprehensive weight for the dynamic parameter layer through the posterior probability distribution. The Bayesian network model is a graphical model based on probabilistic inference, capable of effectively integrating information from different sources. In this invention, the correlation coefficient and entropy weight are used as input information to construct a Bayesian network model. The posterior probability distribution under given subjective and objective weights is calculated using the Bayesian network inference algorithm. The comprehensive weight for the dynamic parameter layer is extracted from the posterior probability distribution. This comprehensive weight considers both subjective and objective factors, more comprehensively reflecting the importance of parameters in safety assessment, and laying a solid foundation for accurately calculating the risk index and determining the safety level.

[0105] In a city overpass construction project, a large highway bridge erecting machine was used. When conducting a safety evaluation of the initial installation state of this bridge erecting machine, this embodiment uses the following process to combine correlation coefficients and entropy weights to fuse parameters and generate a dynamic parameter layer comprehensive weight:

[0106] The dynamic parameter layer of this bridge erecting machine includes key parameters such as the longitudinal displacement deviation and lateral torsion angle of the main beam positioning module, and the hydraulic pressure difference of the outrigger synchronization module. In the preliminary work, the correlation coefficient of each parameter has been calculated using the grey relational analysis method, and the entropy weight of each parameter has been determined using the entropy weight method.

[0107] Taking longitudinal displacement deviation as an example, grey relational analysis shows that its correlation coefficient with the parameter sequence under ideal conditions is 0.75. This means that, subjectively, the longitudinal displacement deviation has a high degree of correlation with the ideal state and is of certain importance in safety assessment. Furthermore, the entropy weight of the longitudinal displacement deviation, calculated using the entropy weight method, is 0.3. The entropy weight is calculated based on the objective distribution characteristics of historical data, reflecting the objective importance of the longitudinal displacement deviation at the data level.

[0108] Using the correlation coefficient as a subjective weighting factor and the entropy weight as an objective weighting factor, a Bayesian network model is employed to probabilistically fuse the subjective and objective weights. The Bayesian network model is a directed acyclic graph where nodes represent random variables and edges represent dependencies between variables. In this case, the correlation coefficient and entropy weight of the vertical displacement deviation are used as input nodes, and together with the corresponding weight nodes of other parameters, a Bayesian network is constructed.

[0109] When constructing the Bayesian network, the conditional probability relationships between each node are determined based on professional knowledge and experience in the field of bridge erecting machines. For example, it is known that there is a certain correlation between longitudinal displacement deviation and lateral torsion angle in the actual operation of bridge erecting machines. When the longitudinal displacement deviation is large, the probability of abnormal lateral torsion angle will also increase. Through a large amount of historical data and expert experience, a conditional probability table between them is determined.

[0110] After constructing the Bayesian network, the Bayesian inference algorithm is used for computation. Based on Bayes' theorem, the Bayesian inference algorithm calculates the posterior probability distribution based on known conditional probabilities and evidence. In this example, through the inference of the Bayesian network, combined with information such as the correlation coefficient and entropy weight of the longitudinal displacement deviation, the posterior probability distribution of the longitudinal displacement deviation after considering subjective and objective factors is calculated. The comprehensive weight of the longitudinal displacement deviation is then extracted from this posterior probability distribution.

[0111] Assuming that the comprehensive weight of longitudinal displacement deviation is 0.4 after Bayesian network calculation and posterior probability analysis, this comprehensive weight no longer relies solely on subjective judgment (correlation coefficient) or purely objective data characteristics (entropy weight), but rather integrates information from both. Compared to using correlation coefficient or entropy weight alone, the comprehensive weight more comprehensively reflects the actual importance of longitudinal displacement deviation in safety assessment.

[0112] For other parameters such as lateral torsion angle and hydraulic pressure difference, the correlation coefficients and entropy weights are input into a Bayesian network for probability fusion in the same way to obtain their comprehensive weights. In this way, comprehensive weights are generated for each parameter in the dynamic parameter layer. These comprehensive weights play a crucial role in the subsequent calculation of the risk index of each structural layer of the bridge erecting machine and the overall risk score, enabling a more accurate determination of the safety level of the initial installation state of the highway bridge erecting machine and providing strong support for ensuring the safe installation of the bridge erecting machine.

[0113] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0114] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the safety of an initial state of installation of a highway bridge-erecting machine, characterized by, Includes the following steps: S1. Establish a multimodal safety evaluation model for highway bridge erecting machines, including structural layer, dynamic parameter layer and risk factor layer; S2. Divide the multi-dimensional feature intervals of the dynamic parameter layer based on the fuzzy clustering algorithm; S3. Use grey relational analysis to calculate the correlation coefficient of each parameter in the risk factor layer; S4. Determine the objective weights of each parameter in the dynamic parameter layer using the entropy weight method; S5. The correlation coefficient is used as the subjective weight factor, and the entropy weight is used as the objective weight factor. A Bayesian network model is used to probabilistically fuse the subjective and objective weights, and the comprehensive weight of the dynamic parameter layer is generated through the posterior probability distribution. S6. Real-time acquisition of multi-source sensor data during the installation process of the bridge erecting machine and discretization processing; S7. Calculate the risk index layer by layer to determine the safety level of the initial installation state of the highway bridge erecting machine; In step S1: The structural layer is divided into a main beam positioning module, a leg synchronization module, a support bearing module, and a whole machine balance module based on the physical topology of the bridge erecting machine's installation components. During the bridge erecting machine installation process, the dynamic parameter layer collects displacement deviation, angle offset, and stress distribution parameters of each module through a sensor network. The risk factor layer is based on a historical fault database to determine the safety parameter range for each module under different operating conditions; The calculation of the risk index in step S7 includes: The risk index for each structural layer is generated by multiplying the comprehensive weight of the dynamic parameter layer with the discretized data. A total risk score is generated by superimposing structural layer weight coefficients and risk indices, and safety levels are divided based on the score range. The dynamic parameter layer includes the following parameters: The longitudinal displacement deviation, lateral torsion angle, and track levelness of the main beam positioning module; Hydraulic pressure difference, extension / retraction speed synchronization rate, and vertical tilt angle of the outrigger synchronization module; The magnitude and distribution of stress on the support structure of the support module; The difference in support force between the front and rear outriggers of the overall balance module and the symmetry of the transverse track; Among them, the longitudinal displacement deviation is recorded by a laser rangefinder, showing the deviation of the main beam in the longitudinal direction relative to the ideal installation position; the lateral torsion angle is obtained by a dual-axis tilt sensor installed at both ends of the main beam, which measures the change of the torsion angle of the main beam on the horizontal plane in real time; the track levelness is monitored by multiple levels installed on the track, with a level set every 5 meters to collect the height difference data of the track, and then calculate the track levelness deviation. The hydraulic pressure difference is collected in real time by pressure sensors installed on the hydraulic cylinders of each outrigger to detect pressure differences between outriggers in a timely manner. The telescopic speed synchronization rate is calculated by collecting displacement data of the outriggers during the telescopic process by displacement sensors installed on the outrigger telescopic mechanism and combining it with the time interval to ensure that the telescopic actions of each outrigger are coordinated and consistent. The vertical tilt angle is monitored by tilt sensors installed on the outriggers. Stress sensors are installed at key stress-bearing parts of the support structure, including the connection between the column and the beam and the support point at the bottom of the support structure, to monitor the magnitude and distribution of stress on the support structure in real time. The difference in support force between the front and rear outriggers is measured in real time by weighing sensors installed at the bottom of the front and rear outriggers; laser alignment devices are installed on both sides of the transverse track to monitor the symmetry of the transverse track and calculate the track symmetry deviation.

2. The method for safety evaluation of the initial installation state of a highway bridge erecting machine according to claim 1, characterized in that, The step S2, which involves dividing the feature intervals based on the fuzzy clustering algorithm, includes: The fuzzy membership function is set according to the parameter distribution characteristics of the dynamic parameter layer; Calculate the similarity matrix between different parameters; The dynamic parameter layer is clustered and grouped based on the similarity matrix.

3. The method for safety evaluation of the initial installation state of a highway bridge erecting machine according to claim 1, characterized in that, The implementation steps of the grey relational analysis method in step S3 include: Select a benchmark reference sequence for each parameter in the risk factor layer; Calculate the correlation coefficient between each parameter sequence and the reference sequence; Generate a correlation coefficient matrix based on the mean of the correlation coefficients.

4. The method for safety evaluation of the initial installation state of a highway bridge erecting machine according to claim 1, characterized in that, The objective weight determination step of the entropy weight method in step S4 includes: The historical dataset of the dynamic parameter layer is transformed by probability density function, and the information entropy value and difference coefficient of each parameter are calculated. The objective weights of the dynamic parameter layer are generated by normalizing the difference coefficients.

5. The method for safety evaluation of the initial installation state of a highway bridge erecting machine according to claim 1, characterized in that, The discretization process in step S6 includes: Anomaly fluctuation detection and noise filtering are performed on multi-source sensor data. Data with different dimensions are converted into a unified discrete interval through the range standardization method, and a discrete distribution map of dynamic parameters is generated according to the time series.

6. The method for safety evaluation of the initial installation state of a highway bridge erecting machine according to claim 5, characterized in that, The steps for generating the discretized distribution map include: The standardized sensor data is divided into equal-length data segments according to time windows. The wavelet packet decomposition algorithm is used to extract the frequency domain energy features of each data segment. The feature difference matrix of the data segments in different time windows is calculated by Mahalanobis distance.

Images