Intelligent early warning device and method based on large-span open-web arch bridge construction risk assessment

Abstract

The invention relates to the technical field of construction risk assessment and early warning, and particularly discloses an intelligent early warning device and method based on construction risk assessment of a large-span open-web arch bridge, and the device comprises a risk related data collection module which comprehensively collects various risk related data in the construction of a large-span deck type open-web arch bridge in a mountainous area; the risk assessment algorithm screening module assesses the assessment accuracy of each combination mode of the risk assessment algorithm, and determines an optimal risk assessment algorithm in combination with the operation cost of all to-be-combined algorithms in each combination mode of the risk assessment algorithm; the risk assessment algorithm execution module comprehensively analyzes various risk related data based on an optimal risk assessment algorithm to obtain construction risk indexes; when the construction risk index reaches a risk threshold value, the intelligent early warning and response module gives an alarm based on an intelligent early warning device, and gives a targeted risk response suggestion based on the construction risk index; and intelligent assessment, early warning and response to the construction risk of the large-span open-web arch bridge are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of construction risk assessment and early warning, and in particular to an intelligent early warning device and method based on construction risk assessment of a large-span hollow arch bridge. Background Art

[0002] In the field of modern bridge engineering, long-span open-arch bridges have become an important choice in transportation infrastructure construction due to their unique structural advantages and excellent spanning capacity. However, the construction process of such bridges is complex and challenging, especially when they are constructed in mountainous environments. The complex and diverse topography, unstable geological conditions, and changeable climatic conditions in mountainous areas have brought great difficulties to the construction of long-span open-arch bridges. During the construction process, it is necessary not only to face the uncertainty of material properties and the complexity of construction technology, but also to deal with the influence of environmental factors such as temperature, humidity, and wind. In addition, the inconvenient transportation, limited construction sites, and difficulty in using large equipment in mountainous areas have further increased the difficulty and risk of construction. Under the current technical conditions, the assessment of the construction risk of long-span open-arch bridges mostly relies on traditional methods, which are mainly based on manual experience and limited data samples, lacking comprehensiveness and accuracy.

[0003] Existing risk assessment models and algorithms fail to fully integrate multi-source data, making it difficult to accurately capture various potential risk factors and precisely quantify construction risks. At the same time, in terms of risk monitoring and early warning, existing technical means are relatively backward, lacking real-time and intelligence, and are unable to timely and accurately perceive the changing trends of risks. This results in the formulation and implementation of risk response measures during the construction process, which are often not timely and effective, making it difficult to ensure the safety and quality of construction, and may also delay construction schedules and increase costs. Therefore, there is an urgent need for a more advanced, intelligent, and efficient intelligent early warning device based on large-span hollow arch bridge construction risk assessment to improve the level of construction risk management.

[0004] Therefore, the present invention proposes an intelligent early warning device and method based on the construction risk assessment of a large-span hollow arch bridge. Summary of the invention

[0005] The present invention provides an intelligent early warning device and method based on the risk assessment of large-span hollow arch bridge construction. Through the risk-related data collection module, various risk-related data can be comprehensively collected, providing a rich and comprehensive information basis for subsequent risk assessment. The risk assessment algorithm screening module determines the optimal risk assessment algorithm by evaluating the accuracy and operation cost of the algorithm, thereby improving the efficiency and accuracy of the assessment. The risk assessment algorithm execution module uses the optimal algorithm to comprehensively analyze the data to obtain the construction risk index, which can quantify the construction risk more accurately. The intelligent early warning and response module promptly issues an alarm when the risk index reaches the threshold, reminding relevant personnel to take measures to reduce risk losses. At the same time, targeted risk response suggestions are given, which helps to quickly formulate effective response strategies to ensure construction safety and smooth progress. Intelligent assessment, early warning and response to the construction risks of large-span hollow arch bridges are realized, and the level and efficiency of construction risk management are improved.

[0006] The present invention provides an intelligent early warning device based on risk assessment of large-span hollow arch bridge construction, comprising: The risk-related data collection module is used to comprehensively collect various risk-related data in the construction of large-span top-decker open-arch bridges in mountainous areas; The risk assessment algorithm screening module is used to evaluate the assessment accuracy of each combination of risk assessment algorithms based on a large number of manual construction risk assessment examples, and determine the best risk assessment algorithm by combining the computational costs of all algorithms to be combined under each combination of risk assessment algorithms. The risk assessment algorithm is a hybrid algorithm of neural network and fuzzy logic. The risk assessment algorithm execution module is used to comprehensively analyze various risk-related data based on the best risk assessment algorithm to obtain construction risk indicators; The intelligent early warning and response module is used to issue an alarm based on the intelligent early warning device when the construction risk index reaches the risk threshold. At the same time, it gives targeted risk response suggestions based on the construction risk index.

[0007] Preferably, the risk-related data collection module includes: The material performance acquisition submodule is used to acquire material performance data in real time based on material performance sensors; The construction process parameter acquisition submodule is used to obtain the structural displacement, structural stress, and structural strain in the construction process in real time as construction process parameter data based on the construction process sensor; The environmental parameter acquisition submodule is used to obtain environmental parameter data in real time based on environmental sensors; Among them, all types of risk-related data include: material performance data, construction process parameter data, and environmental parameter data.

[0008] Preferably, the risk assessment algorithm screening module includes: The algorithm accuracy assessment submodule is used to assess the assessment accuracy of each combination of risk assessment algorithms based on a large number of manual construction risk assessment examples, where the combination methods include series connection, parallel connection, and fusion; The operation cost calculation submodule is used to determine the algorithm architecture of all algorithms to be combined under each combination mode of the risk assessment algorithm, and calculate the total operation cost of the risk assessment algorithm under each combination mode based on the algorithm architecture of all algorithms to be combined under each combination mode; The optimal assessment algorithm determination submodule is used to determine the optimal risk assessment algorithm based on the assessment accuracy of each combination of the risk assessment algorithms and the total computational cost of the risk assessment algorithms under each combination.

[0009] Preferably, the algorithm accuracy assessment submodule includes: Algorithm architecture building unit, used to build a complete algorithm architecture for each combination of risk assessment algorithms; An algorithm accuracy testing unit, used to input all risk-related data contained in all manual construction risk assessment instances into the complete algorithm architecture of each combination mode of the risk assessment algorithm, and obtain the risk assessment results of each manual construction risk assessment instance under the complete algorithm architecture of each combination mode of the risk assessment algorithm; An assessment error calculation unit is used to average the deviations between the manual risk assessment results contained in all manual construction risk assessment instances and the risk assessment results of the corresponding manual construction risk assessment instances under the complete algorithm architecture of each combination of the risk assessment algorithm, so as to obtain the assessment error of each combination of the risk assessment algorithm; The evaluation accuracy calculation unit is used to calculate the evaluation accuracy of each combination of the risk evaluation algorithms based on the evaluation error of each combination of the risk evaluation algorithms.

[0010] Preferably, the operation cost calculation submodule includes: An algorithm architecture acquisition unit, used to acquire the algorithm architectures of all algorithms to be combined under each combination mode of the risk assessment algorithm; A single operation cost calculation unit, used for calculating each operation cost of each algorithm to be combined based on the algorithm architecture of each algorithm to be combined; A unit for determining a mixed method of single cost is used to determine a mixed calculation method of each operation cost under each combination method; A single mixed cost calculation unit is used to calculate the single mixed cost under each combination mode based on each operation cost of all algorithms to be combined under each combination mode of the risk assessment algorithm and the mixed calculation mode of the corresponding operation cost under the corresponding combination mode; The total operation cost calculation unit is used to calculate the total operation cost of the risk assessment algorithm in each combination mode based on each mixed operation cost and the corresponding calculation weight in each combination mode.

[0011] Preferably, the single operation cost calculation unit includes: A time cost calculation subunit, used to calculate the operation time cost required for each algorithm to be combined to process unit input data based on the algorithm architecture of each algorithm to be combined; A memory cost calculation subunit, used for calculating the computational memory cost required for each algorithm to be combined to process unit input data based on the algorithm architecture of each algorithm to be combined; Among them, the computational costs of all items include computational time cost and computational memory cost.

[0012] Preferably, the intelligent early warning and response module includes: The risk threshold acquisition submodule is used to obtain the risk thresholds corresponding to different construction stages and different construction requirements; The intelligent early warning submodule is used to issue an alarm based on the intelligent early warning device when the construction risk index reaches the risk threshold corresponding to the current construction stage and the current construction requirements; The risk response submodule is used to retrieve the risk response suggestion knowledge base based on construction risk indicators and actual construction conditions to obtain targeted risk response suggestions.

[0013] Preferably, the intelligent early warning device issues an alarm by one or more of sound, light, and display of text information, wherein the displayed text information includes the risk type, risk level, and risk consequence.

[0014] Preferably, the optimal evaluation algorithm determines the submodule, including: A comprehensive performance index calculation unit, used to calculate the comprehensive performance index of each combination of the risk assessment algorithm based on the assessment accuracy and corresponding calculation weight of each combination of the risk assessment algorithm, the total operation cost of the risk assessment algorithm under each combination and the corresponding calculation weight; The best algorithm screening unit is used to take the algorithm architecture corresponding to the combination method corresponding to the maximum comprehensive performance index among all the combination methods of the risk assessment algorithm as the best risk assessment algorithm.

[0015] The present invention provides an intelligent early warning method based on risk assessment of large-span open-web arch bridge construction, which is applied to any of the above intelligent early warning devices based on risk assessment of large-span open-web arch bridge construction, and comprises: S1: Comprehensively collect data related to various risks in the construction of large-span top-decker open-arch bridges in mountainous areas; S2: Based on a large number of manual construction risk assessment examples, the assessment accuracy of each combination of risk assessment algorithms is evaluated, and the computational costs of all algorithms to be combined under each combination of risk assessment algorithms are combined to determine the best risk assessment algorithm; S3: Comprehensively analyze various risk-related data based on the best risk assessment algorithm to obtain construction risk indicators, where the risk assessment algorithm is a hybrid algorithm of neural network and fuzzy logic; S4: When the construction risk index reaches the risk threshold, an alarm is issued based on the intelligent early warning device. At the same time, targeted risk response suggestions are given based on the construction risk index.

[0016] The beneficial effects of the present invention compared to the prior art are as follows: the risk-related data collection module can comprehensively collect various risk-related data, providing a rich and comprehensive information basis for subsequent risk assessment. The risk assessment algorithm screening module determines the best risk assessment algorithm by evaluating the accuracy and computational cost of the algorithm, thereby improving the efficiency and accuracy of the assessment. The risk assessment algorithm execution module uses the best algorithm to comprehensively analyze the data, obtain construction risk indicators, and can more accurately quantify construction risks. The intelligent early warning and response module will issue an alarm in time when the risk indicator reaches the threshold, reminding relevant personnel to take measures to reduce risk losses. At the same time, targeted risk response suggestions are given, which helps to quickly formulate effective response strategies to ensure construction safety and smooth progress. Intelligent assessment, early warning and response to the construction risks of large-span hollow arch bridges are realized, improving the level and efficiency of construction risk management.

[0017] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in this application document.

[0018] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 Schematic diagram of an intelligent early warning device based on risk assessment of large-span hollow arch bridge construction in an embodiment of the present invention; Figure 2 It is an information flow diagram during the execution of intelligent early warning based on risk assessment of large-span hollow arch bridge construction in an embodiment of the present invention; Figure 3The present invention is a flowchart of an intelligent early warning method based on risk assessment of large-span hollow arch bridge construction in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0021] Example 1 The present invention provides an intelligent early warning device based on the risk assessment of large-span hollow arch bridge construction. Figure 1 and Figure 2 ,include: The risk-related data collection module is used to comprehensively collect various risk-related data in the construction of large-span top-decker open-arch bridges in mountainous areas; The risk assessment algorithm screening module is used to evaluate the assessment accuracy of each combination of risk assessment algorithms based on a large number of manual construction risk assessment examples, and determine the best risk assessment algorithm by combining the computational costs of all algorithms to be combined under each combination of risk assessment algorithms. The risk assessment algorithm is a hybrid algorithm of neural network and fuzzy logic. The risk assessment algorithm execution module is used to comprehensively analyze various risk-related data based on the best risk assessment algorithm to obtain construction risk indicators; The intelligent early warning and response module is used to issue an alarm based on the intelligent early warning device when the construction risk index reaches the risk threshold. At the same time, it gives targeted risk response suggestions based on the construction risk index.

[0022] In this embodiment, in the hybrid algorithm, the neural network and the fuzzy logic can be combined in series, parallel or fusion, and the specific implementation process is as follows: (I) Series connection: 1.1 Basic idea: Connect the neural network and fuzzy logic system in sequence, and the output of the former system serves as the input of the latter system.

[0023] 1.2 Implementation process: Neural network first, fuzzy logic later: neural network is used to extract features and perform preliminary processing on raw data, and the abstract and refined feature information outputted is used as the input variable of fuzzy logic system. For example, in image recognition, neural network first extracts features such as edge and texture of the image, and fuzzy logic system then determines the image category based on them.

[0024] Fuzzy logic first, neural network later: The fuzzy logic system first performs fuzzification and preliminary rule inference on the data, and its output is used as the input of the neural network. For example, in risk assessment, the fuzzy logic system first makes a preliminary assessment based on fuzzy risk factors, and the neural network then makes an in-depth analysis based on the assessment results and other data to derive the final risk level.

[0025] (II) Parallel connection: 2.1 Basic idea: The neural network and fuzzy logic system work in parallel, each processing the input data independently, and then fusing the output results.

[0026] 2.2 Implementation process: Separate processing: The same input data is given to the neural network and the fuzzy logic system at the same time. The neural network uses learning and nonlinear mapping capabilities to process, and the fuzzy logic system processes based on fuzzy rules and reasoning mechanisms. For example, in fault diagnosis, the two give fault probability outputs separately.

[0027] Result fusion: Use the weighted average method to assign weights to the outputs of the two according to factors such as reliability and importance, and then take the weighted sum; or use the voting method. When the output is a discrete category, the final category is determined based on the voting results.

[0028] (III) Fusion method (taking fuzzy neural network as an example): 3.1 Basic idea: Integrate fuzzy logic into neural network structure and algorithm, or introduce neural network learning ability to improve the performance of fuzzy system.

[0029] 3.2 Implementation process: Fuzzification layer: Fuzzify the input data and map the precise input value to the membership value of the fuzzy set. For example, fuzzify "25℃" to the membership value of "moderate temperature".

[0030] Fuzzy rule layer: Define fuzzy rules based on domain knowledge and experience, present them in the form of “if…then…”, and set weights for each rule.

[0031] Reasoning layer: Reasoning is performed based on the input membership value and fuzzy rules, such as Mamdani reasoning method, Sugeno reasoning method, etc., to obtain fuzzy output results.

[0032] Defuzzification layer: converts fuzzy output results into precise values ​​or decisions, such as the centroid method, maximum membership method, etc.

[0033] Learning algorithm: Use neural network learning algorithms, such as back propagation algorithm, genetic algorithm, etc., to train and adjust fuzzy neural network parameters to optimize model accuracy and generalization ability.

[0034] In this embodiment, various risk-related data in the construction of large-span top-supported hollow arch bridges in mountainous areas refer to data related to various factors that may cause construction risks during the construction of large-span top-supported hollow arch bridges in mountainous environments. Including but not limited to material performance data used in construction (such as material strength, elastic modulus, etc.), construction process parameter data (such as structural displacement, stress, strain, etc.) and environmental parameter data (such as temperature, humidity, wind speed, etc.). For example, in the construction of a large-span top-supported hollow arch bridge in a mountainous area, the yield strength of the steel used, the displacement of the arch during construction, and the ambient temperature at the time.

[0035] In this embodiment, the manual construction risk assessment example is a specific case in which manual risk assessment is performed on the construction of a large-span, top-supported, open-web arch bridge in a specific mountainous area based on past experience and relevant standards. For example, in a similar mountainous bridge construction project in the past, manual risk assessment of structural stability, material quality, and other aspects during the construction process is actually recorded.

[0036] In this embodiment, the risk assessment algorithm is a calculation method or program for quantifying and analyzing the construction risk of a large-span, top-mounted, hollow arch bridge in a mountainous area, such as an algorithm for calculating the possibility and severity of construction risks based on techniques such as neural networks or fuzzy logic.

[0037] In this embodiment, all kinds of combinations of risk assessment algorithms include various forms of combining different risk assessment algorithms in series, parallel, fusion, etc. For example, using a neural network algorithm and a fuzzy logic algorithm in series, or using several different algorithms to process data in parallel.

[0038] In this embodiment, the best risk assessment algorithm is the algorithm most suitable for risk assessment of construction of large-span top-supported open-web arch bridges in mountainous areas after comprehensively considering factors such as assessment accuracy and computational cost. For example, after comparative analysis, it is found that an algorithm that integrates multiple technologies, has reasonable computational cost and high assessment accuracy is the best risk assessment algorithm.

[0039] In this embodiment, various risk-related data are comprehensively analyzed based on the best risk assessment algorithm to obtain a construction risk index: the selected best risk assessment algorithm is used to process and calculate various types of data collected related to construction risks to obtain a quantitative index that can reflect the degree of construction risk. For example, the best algorithm is used to analyze data such as material properties, construction technology and environment to obtain a value indicating that the construction risk is at a high, medium or low level.

[0040] In this embodiment, the construction risk index is a specific value or parameter used to measure the construction risk of a large-span top-decker open-arch bridge in a mountainous area. For example, it may be a value between 0 and 1, where 0 indicates an extremely low risk and 1 indicates an extremely high risk.

[0041] In this embodiment, risk threshold is a pre-set critical value used to determine whether the construction risk has reached a critical value that requires early warning measures. For example, the risk threshold is set to 0.8, and when the construction risk index exceeds 0.8, an early warning is triggered.

[0042] In this embodiment, the intelligent early warning device is a device or system that can automatically monitor construction risk indicators and issue an alarm when the risk threshold is reached or exceeded. For example, a set of devices including sensors, data processing modules and alarm modules can issue an alarm through sound, light, etc. when the risk exceeds the threshold.

[0043] In this example, targeted risk response suggestions are: measures and methods specifically proposed to reduce or eliminate risks based on specific construction risk indicators and actual construction conditions. For example, if the risk assessment shows that there may be problems with material quality, it is recommended to strengthen material inspection or change suppliers.

[0044] The beneficial effects of the above technologies are as follows: through the risk-related data collection module, various risk-related data can be comprehensively collected, providing a rich and comprehensive information basis for subsequent risk assessment. The risk assessment algorithm screening module determines the best risk assessment algorithm by evaluating the accuracy and computational cost of the algorithm, thereby improving the efficiency and accuracy of the assessment. The risk assessment algorithm execution module uses the best algorithm to comprehensively analyze the data, obtain construction risk indicators, and can quantify construction risks more accurately. The intelligent early warning and response module will issue an alarm in time when the risk indicator reaches the threshold, reminding relevant personnel to take measures to reduce risk losses. At the same time, targeted risk response suggestions are given, which helps to quickly formulate effective response strategies to ensure construction safety and smooth progress. Intelligent assessment, early warning and response to the construction risks of large-span hollow arch bridges have been realized, improving the level and efficiency of construction risk management.

[0045] Example 2 Based on Example 1, the risk-related data collection module includes: The material performance acquisition submodule is used to acquire material performance data in real time based on material performance sensors; The construction process parameter acquisition submodule is used to obtain the structural displacement, structural stress, and structural strain in the construction process in real time as construction process parameter data based on the construction process sensor; The environmental parameter acquisition submodule is used to obtain environmental parameter data in real time based on environmental sensors; Among them, all types of risk-related data include: material performance data, construction process parameter data, and environmental parameter data.

[0046] In this embodiment, the material performance sensor is a device used to monitor and obtain the relevant performance parameters of the materials used in the construction of the large-span top-bearing hollow arch bridge in mountainous areas in real time, such as a sensor capable of measuring the strength of steel and the compressive strength of concrete.

[0047] In this embodiment, material performance data refers to information collected by material performance sensors that reflects the properties of materials used in construction, such as the yield strength value of steel and the elastic modulus value of concrete.

[0048] In this embodiment, the construction process sensor is an instrument installed at the construction site for monitoring parameters related to the construction process, such as a sensor for measuring the casting speed of the arch ring.

[0049] In this embodiment, the structural displacement, structural stress, and structural strain during the construction process are: during the construction process, the position change (displacement), the internal force (stress), and the degree of shape change (strain) of the bridge structure. For example, how much the arch has sunk (displacement), the pressure on a certain part of the arch rib (stress), and the degree of tensile or compressive deformation (strain) of the bridge components.

[0050] In this embodiment, the environmental sensor is a device used to detect parameters related to the construction environment of a large-span, top-supported, open-arch bridge in a mountainous area, such as a sensor for measuring wind speed, temperature, and humidity.

[0051] In this embodiment, environmental parameter data refers to information about the construction environment acquired by the environmental sensor, such as specific temperature value, humidity percentage, wind speed, and other data.

[0052] The beneficial effects of the above technologies include: the material performance acquisition submodule obtains material performance data in real time through sensors, can grasp the material status in time, and provide basic data for risk assessment. The construction process parameter acquisition submodule collects key parameters in the construction process in real time, such as structural displacement, stress and strain, which helps to accurately evaluate the impact of construction technology on the structure. The environmental parameter acquisition submodule obtains environmental data in real time, taking into account the potential impact of environmental factors on construction risks. Comprehensive collection of data such as material properties, construction process parameters and environmental parameters makes risk-related data more comprehensive and accurate. It provides reliable data support for subsequent more accurate risk assessment and early warning, and improves the scientificity and reliability of construction risk assessment. It improves the comprehensiveness and real-time nature of data collection for risk assessment of large-span hollow arch bridge construction, which helps to better ensure construction safety.

[0053] Example 3 Based on Example 1, the risk assessment algorithm screening module includes: The algorithm accuracy assessment submodule is used to assess the assessment accuracy of each combination of risk assessment algorithms based on a large number of manual construction risk assessment examples, where the combination methods include series connection, parallel connection, and fusion; The operation cost calculation submodule is used to determine the algorithm architecture of all algorithms to be combined under each combination mode of the risk assessment algorithm, and calculate the total operation cost of the risk assessment algorithm under each combination mode based on the algorithm architecture of all algorithms to be combined under each combination mode; The optimal assessment algorithm determination submodule is used to determine the optimal risk assessment algorithm based on the assessment accuracy of each combination of the risk assessment algorithms and the total computational cost of the risk assessment algorithms under each combination.

[0054] In this embodiment, the algorithm architecture of all algorithms to be combined in each combination mode of the risk assessment algorithm is determined: the internal structure and organization of each individual algorithm involved in each combination mode of different risk assessment algorithms are clarified. For example, in the serial combination mode, the network layer number, node number and other structures of the neural network algorithm to be run first, and the rule number and logical structure of the fuzzy logic algorithm to be run later are determined.

[0055] In this embodiment, the total operation cost of the risk assessment algorithm in each combination mode is calculated to calculate the comprehensive resource cost required to complete the entire risk assessment process in a specific risk assessment algorithm combination mode. For example, it includes the sum of the time cost, memory usage cost, and possible hardware equipment loss cost required for the neural network algorithm and the fuzzy logic algorithm to run separately in a series mode.

[0056] The beneficial effects of the above technologies include: the algorithm accuracy assessment submodule evaluates the accuracy of each combination method through a large number of manual construction risk assessment examples, providing an accuracy basis for screening the best algorithm. The computational cost calculation submodule determines the algorithm architecture and calculates the total computational cost, which helps to consider resource consumption factors when selecting an algorithm. The optimal evaluation algorithm determination submodule comprehensively considers accuracy and computational cost, and can select the best risk assessment algorithm that strikes a balance between evaluation effect and efficiency. It improves the scientificity and rationality of risk assessment algorithm screening, ensuring that the selected algorithm is both accurate and cost-effective. It provides strong algorithm support for more accurate and efficient construction risk assessment, and improves the quality and efficiency of risk assessment. It enhances the comprehensiveness and effectiveness of large-span hollow arch bridge construction risk assessment algorithm screening, which is conducive to improving the level of construction risk management.

[0057] Example 4 Based on Example 3, the algorithm accuracy evaluation submodule includes: Algorithm architecture building unit, used to build a complete algorithm architecture for each combination of risk assessment algorithms; An algorithm accuracy testing unit, used to input all risk-related data contained in all manual construction risk assessment instances into the complete algorithm architecture of each combination mode of the risk assessment algorithm, and obtain the risk assessment results of each manual construction risk assessment instance under the complete algorithm architecture of each combination mode of the risk assessment algorithm; An assessment error calculation unit is used to average the deviations between the manual risk assessment results contained in all manual construction risk assessment instances and the risk assessment results of the corresponding manual construction risk assessment instances under the complete algorithm architecture of each combination of the risk assessment algorithm, so as to obtain the assessment error of each combination of the risk assessment algorithm; The evaluation accuracy calculation unit is used to calculate the evaluation accuracy of each combination of the risk evaluation algorithms based on the evaluation error of each combination of the risk evaluation algorithms.

[0058] In this embodiment, a complete algorithm architecture for each combination of risk assessment algorithms is constructed: a specific and comprehensive computing framework and process for each combination of risk assessment algorithms is constructed. For example, for the fusion combination of neural network and fuzzy logic, a complete architecture including data input layer, neural network layer, fusion layer, output layer, etc. is constructed.

[0059] In this embodiment, all risk-related data contained in all manual construction risk assessment instances are input into the complete algorithm architecture of each combination of risk assessment algorithms, and the risk assessment results of each manual construction risk assessment instance under the complete algorithm architecture of each combination of risk assessment algorithms are obtained: the risk-related data of many manual assessments are put into the architecture of each algorithm combination one by one to run, and the assessment conclusion of each instance under the algorithm combination is obtained. For example, the material properties, construction process, environment and other data of multiple past manual assessments are input into the series-connected neural network and fuzzy logic algorithm architectures to obtain their respective risk assessment values.

[0060] In this embodiment, the risk assessment results of each manual construction risk assessment instance under the complete algorithm architecture of each combination of risk assessment algorithms are: the risk assessment value or conclusion obtained after processing the data of a specific manual construction risk assessment instance through a specific risk assessment algorithm combination. For example, for a specific mountain bridge construction case, the construction risk index calculated under the algorithm architecture of parallel neural network and fuzzy logic is 0.4.

[0061] In this embodiment, the deviation between the manual risk assessment results contained in the manual construction risk assessment instance and the risk assessment results of the corresponding manual construction risk assessment instance under the complete algorithm architecture of each combination of the risk assessment algorithm is: the difference between the risk assessment conclusion drawn manually based on experience and the risk assessment conclusion of the same instance calculated by a specific algorithm architecture. For example, the construction risk index of a certain instance assessed manually is 0.4, while the construction risk index obtained by the algorithm is 0.45. The difference between the two is the deviation.

[0062] In this embodiment, the evaluation error of each combination of risk assessment algorithms is used to measure the degree of deviation from the actual situation when a certain combination of risk assessment algorithms is used to process manual construction risk assessment instances. For example, if the evaluation results of a certain algorithm combination for multiple instances differ greatly from the average manual evaluation results, it means that its evaluation error is large.

[0063] In this embodiment, the evaluation accuracy of each combination of the risk assessment algorithm is calculated based on the evaluation error of each combination of the risk assessment algorithm: by analyzing and calculating the evaluation error, the accuracy of each combination of the risk assessment algorithm close to the actual situation is obtained. For example, the smaller the evaluation error, the higher the evaluation accuracy; conversely, the larger the evaluation error, the lower the evaluation accuracy.

[0064] The beneficial effects of the above technologies include: the algorithm architecture building unit builds a complete algorithm architecture for each combination method, providing a basic framework for subsequent testing and evaluation. The algorithm accuracy testing unit inputs risk-related data into different architectures, obtains evaluation results, and comprehensively tests the performance of various combinations. The evaluation error calculation unit accurately quantifies the evaluation error by calculating the average deviation between the manual risk assessment results and the algorithm assessment results. The evaluation accuracy calculation unit derives the evaluation accuracy based on the evaluation error, providing an intuitive and reliable indicator for screening the optimal combination method. This step-by-step accuracy evaluation method improves the accuracy and scientificity of the evaluation. It can more accurately determine the advantages and disadvantages of various combinations of risk assessment algorithms and select the most suitable algorithm combination method. It improves the accuracy and reliability of the accuracy evaluation of the risk assessment algorithm for large-span hollow arch bridge construction, providing strong support for optimizing the risk assessment algorithm.

[0065] Example 5 Based on Example 3, the operation cost calculation submodule includes: An algorithm architecture acquisition unit, used to acquire the algorithm architectures of all algorithms to be combined under each combination mode of the risk assessment algorithm; A single operation cost calculation unit, used for calculating each operation cost of each algorithm to be combined based on the algorithm architecture of each algorithm to be combined; A unit for determining a mixed method of single cost is used to determine a mixed calculation method of each operation cost under each combination method; A single mixed cost calculation unit is used to calculate the single mixed cost under each combination mode based on each operation cost of all algorithms to be combined under each combination mode of the risk assessment algorithm and the mixed calculation mode of the corresponding operation cost under the corresponding combination mode; The total operation cost calculation unit is used to calculate the total operation cost of the risk assessment algorithm in each combination mode based on each mixed operation cost and the corresponding calculation weight in each combination mode.

[0066] In this embodiment, the algorithm architecture of all algorithms to be combined in each combination of risk assessment algorithms is obtained: the internal structure and operation mode of each specific algorithm involved in each combination of risk assessment algorithms are collected and clarified. For example, for the fusion combination of neural network and fuzzy logic, the architecture information such as the number of layers and nodes of the neural network and the number of rules of fuzzy logic is obtained.

[0067] In this embodiment, the cost of each operation of the algorithms to be combined refers to the various resource consumption costs generated during the operation of the individual algorithms to be combined, such as time cost, memory cost, etc. For example, the computing time required for a neural network algorithm to process a certain amount of data, or the cost corresponding to the memory size occupied by a fuzzy logic algorithm when running.

[0068] In this embodiment, the mixed calculation method of each operation cost in each combination mode is determined: the rules of how each operation cost interacts and is calculated in different algorithm combinations are clarified. For example, in a series combination, the cost of the first algorithm may be calculated first, and then the result is used as input to calculate the cost of the second algorithm; in a parallel combination, the costs of the two algorithms may be calculated separately and then added.

[0069] In this embodiment, based on the mixed calculation method of each operation cost of all algorithms to be combined under each combination of risk assessment algorithms and the corresponding operation cost under the corresponding combination, the single mixed operation cost under each combination is calculated: according to the operation cost of each algorithm and the cost calculation rules under the specific combination, the comprehensive operation cost of a single item under each combination is obtained. For example, for a certain series combination of neural network and fuzzy logic, according to the time cost of the two and the time cost calculation rules under the series combination, the single mixed operation cost of this series combination in terms of time is calculated.

[0070] In this embodiment, based on each mixed operation cost and the corresponding calculation weight under each combination mode, the total operation cost of the risk assessment algorithm under each combination mode is calculated: the overall operation cost of each single mixed operation cost under each combination mode is comprehensively considered, and combined with its importance in the total cost calculation (calculation weight, where the calculation weight can be preset), the overall operation cost of each algorithm combination mode is obtained. For example, assuming that the weight of the time cost is 0.4, the weight of the memory cost is 0.6, the time single mixed operation cost of a series combination mode is 10, and the memory single mixed operation cost is 20, then the total operation cost = 10 × 0.4 + 20 × 0.6 = 16.

[0071] The beneficial effects of the above technology include: the algorithm architecture acquisition unit accurately acquires the architecture of the algorithm to be combined, laying the foundation for subsequent cost calculation. The single operation cost calculation unit calculates each operation cost of each algorithm to be combined separately, making the cost analysis more detailed. The single cost mixed method determination unit clarifies the mixed calculation method of each cost under different combination methods to improve the rationality of the calculation. The single mixed cost calculation unit obtains the single mixed operation cost according to the mixed calculation method, which more accurately reflects the cost characteristics of the combination method. The total operation cost calculation unit comprehensively considers the mixed operation costs and calculation weights of each mixed operation, calculates the total operation cost, and provides an important economic indicator for algorithm selection. This comprehensive and detailed operation cost calculation method helps to fully consider economic factors when selecting risk assessment algorithms. It can more scientifically compare the operation costs of different combination methods and select the combination method of risk assessment algorithms with higher cost performance. It enhances the accuracy and effectiveness of operation cost calculation in the risk assessment of large-span hollow arch bridge construction, which is conducive to improving resource utilization efficiency and the economy of risk assessment.

[0072] Example 6 Based on Embodiment 5, the single operation cost calculation unit includes: A time cost calculation subunit, used to calculate the operation time cost required for each algorithm to be combined to process unit input data based on the algorithm architecture of each algorithm to be combined; A memory cost calculation subunit, used for calculating the computational memory cost required for each algorithm to be combined to process unit input data based on the algorithm architecture of each algorithm to be combined; Among them, the computational costs of all items include computational time cost and computational memory cost.

[0073] In this embodiment, unit input data refers to a piece of input data with a fixed data volume in the risk assessment algorithm operation. For example, assuming that the material performance data input to the algorithm for operation each time is fixed to 100 measurement values, the construction process parameter data is 50 groups, and the environmental parameter data is 20, this fixed combination of data volume constitutes a unit input data.

[0074] In this embodiment, the operation time cost required for each algorithm to be combined to process unit input data is calculated based on the algorithm architecture of each algorithm to be combined, including: When the algorithm to be combined is a feedforward neural network, the total number of neurons included in the input layer of the feedforward neural network, the total number of neurons included in the hidden layer, the total number of parameters, the number of bytes occupied by each parameter are determined based on the algorithm architecture of the feedforward neural network, and the time required for the feedforward neural network to perform a single multiplication operation on a unit of input data, the time required for the feedforward neural network to perform a single addition operation on a unit of input data, and the time required for the feedforward neural network to perform a single activation function calculation on a unit of input data are determined; When the algorithm to be combined is a recurrent neural network, the sequence length, input dimension, number of neurons, total number of parameters, number of bytes occupied by each parameter of the unit input data of the recurrent neural network are determined based on the algorithm architecture of the recurrent neural network, and the time required for the recurrent neural network to perform a single matrix multiplication operation on the unit input data and the time required for the recurrent neural network to perform a single activation function calculation on the unit input data are determined; When the algorithm to be combined is a fuzzy logic algorithm, the total number of fuzzy rules contained in the fuzzy logic algorithm, the total number of fuzzy sets contained in each fuzzy rule, the total number of discrete points contained in each fuzzy set, the number of bytes occupied by the membership value of each discrete point, the number of bytes occupied by each element are determined based on the algorithm architecture of the fuzzy logic algorithm, and the time required for each fuzzy set contained in each fuzzy rule to perform a single membership function calculation on unit input data, the synthesis operation time when the fuzzy logic algorithm uses each fuzzy rule for fuzzy reasoning, and the defuzzification time of the fuzzy logic algorithm are determined; Based on the total number of neurons included in the input layer of the feedforward neural network, the total number of neurons included in the hidden layer, the time required for the feedforward neural network to perform a single multiplication operation on unit input data, the time required for the feedforward neural network to perform a single addition operation on unit input data, the time required for the feedforward neural network to perform a single activation function calculation on unit input data, the sequence length of unit input data input into the recurrent neural network, the input dimension, the number of neurons, the time required for the recurrent neural network to perform a single matrix multiplication operation on unit input data, the time required for the recurrent neural network to perform a single activation function calculation on unit input data, the total number of fuzzy rules included in the fuzzy logic algorithm, the total number of fuzzy sets included in each fuzzy rule, the time required for each fuzzy set included in each fuzzy rule to perform a single membership function calculation on unit input data, the synthesis operation time when the fuzzy logic algorithm uses each fuzzy rule for fuzzy reasoning, and the defuzzification time of the fuzzy logic algorithm, the operation time cost required for each algorithm to be combined to process unit input data is calculated: In the formula, is the computational time cost required to process unit input data when the algorithm to be combined is a feedforward neural network, is the total number of neurons contained in the input layer of the feedforward neural network, is the total number of neurons contained in the hidden layer of the feedforward neural network, It is the time required for a feedforward neural network to perform a single multiplication operation on a unit input data. It is the time required for the feedforward neural network to perform a single addition operation on the unit input data. It is the time required for a feedforward neural network to perform a single activation function calculation on a unit input data. is the computational time cost required to process unit input data when the algorithm to be combined is a recurrent neural network, is the sequence length of the unit input data of the input recurrent neural network, is the input dimension of the recurrent neural network, is the number of neurons in the recurrent neural network, is the time required for a recurrent neural network to perform a single matrix multiplication on unit input data. It is the time required for the recurrent neural network to calculate a single activation function for a unit input data. is the computational time cost required to process unit input data when the algorithm to be combined is a fuzzy logic algorithm, is the total number of fuzzy rules contained in the fuzzy logic algorithm, The fuzzy logic algorithm The total number of fuzzy sets contained in the fuzzy rules, For the The fuzzy rule contains The time required for a fuzzy set to perform a single membership function calculation on a unit input data is The fuzzy logic algorithm adopts The synthetic operation time when fuzzy rules are used for fuzzy reasoning, is the defuzzification time of the fuzzy logic algorithm.

[0075] In this embodiment, the computational memory cost required for each algorithm to be combined to process unit input data is calculated based on the algorithm architecture of each algorithm to be combined, including: Based on the total number of parameters of feedforward neural networks and recurrent neural networks, the number of bytes occupied by each parameter, the total number of discrete points contained in each fuzzy set, the number of bytes occupied by the membership value of each discrete point, and the number of bytes occupied by each element In the formula, is the computational memory cost required to process unit input data when the algorithm to be combined is a feedforward neural network or a recurrent neural network, is the total number of parameters of the feedforward neural network or recurrent neural network, is the number of bytes occupied by each parameter of a feedforward neural network or a recurrent neural network, is the computational memory cost required to process unit input data when the algorithm to be combined is a fuzzy logic algorithm, The number of bytes occupied by the membership value of each discrete point, For the The fuzzy rule contains The total number of discrete points contained in the fuzzy sets, The number of bytes occupied by each element.

[0076] The beneficial effects of the above technologies include: the time cost calculation subunit calculates the operation time cost of processing unit input data through the algorithm architecture, which helps to evaluate the operation efficiency of the algorithm. The memory cost calculation subunit calculates the operation memory cost of processing unit input data, taking into account the consumption of hardware resources by the algorithm. The comprehensive operation time cost and memory cost comprehensively cover the main resource consumption factors in the algorithm execution process. Detailed and key indicators are provided for more accurate calculation of single operation costs. It can more accurately measure the resource requirements of each algorithm to be combined, providing a more reliable basis for selecting the optimal algorithm combination method. It improves the comprehensiveness and accuracy of single operation cost calculation in the risk assessment of large-span hollow arch bridge construction, which is conducive to optimizing algorithm selection and resource allocation.

[0077] Example 7 Based on Example 1, the intelligent early warning and response module includes: The risk threshold acquisition submodule is used to obtain the risk thresholds corresponding to different construction stages and different construction requirements; The intelligent early warning submodule is used to issue an alarm based on the intelligent early warning device when the construction risk index reaches the risk threshold corresponding to the current construction stage and the current construction requirements; The risk response submodule is used to retrieve the risk response suggestion knowledge base based on construction risk indicators and actual construction conditions to obtain targeted risk response suggestions.

[0078] In this embodiment, the construction stage refers to different time periods or links in the construction process of a large-span top-bearing hollow arch bridge in a mountainous area, which have different work contents and characteristics, such as the foundation construction stage, the arch ring casting stage, the bridge deck paving stage, etc.

[0079] In this embodiment, construction requirements are specific regulations and standards for the quality, progress, safety, etc. of the construction process of a large-span, top-supported, hollow arch bridge in mountainous areas. For example, the construction requires the arch ring to be poured within a specific time, or the concrete strength of the arch ring must meet a specific standard.

[0080] In this embodiment, the risk thresholds corresponding to different construction stages and different construction requirements are obtained: the critical values ​​that are considered to be risks requiring early warnings are collected and determined under each specific construction time period and specific construction standard. For example, in the foundation construction stage, the risk threshold may be set to 0.5; for construction links that require high precision, the risk threshold may be more stringent.

[0081] In this embodiment, actual construction conditions refer to various conditions actually occurring at the construction site, including construction progress, quality of completed work, problems encountered, etc. For example, in actual construction, the progress of arch ring pouring is slower than planned, or the concrete strength of certain parts does not meet expectations.

[0082] In this embodiment, the risk response suggestion knowledge base is an information base that stores response methods and suggestions for various possible construction risk conditions. For example, the knowledge base contains suggestions such as changing suppliers when material quality problems occur and adding construction personnel when construction progress lags behind.

[0083] The beneficial effects of the above technologies include: the risk threshold acquisition submodule can obtain the corresponding risk threshold according to different construction stages and requirements, making the early warning more targeted and reasonable. The intelligent early warning submodule will issue an alarm in time when the construction risk index reaches the threshold, and can quickly remind relevant personnel to pay attention to the risk situation. The risk response submodule obtains targeted suggestions by searching the knowledge base, providing guidance for quickly taking effective response measures. The different stages and requirements of construction are comprehensively considered to improve the accuracy and adaptability of early warning and response. It helps to take timely measures to reduce risks and ensure the smooth progress of construction and the safety of personnel. It enhances the intelligent early warning and response capabilities of large-span hollow arch bridge construction risks, and improves the level and efficiency of construction risk management.

[0084] Example 8 Based on Example 7, the intelligent warning device issues an alarm through one or more of sound, light, and displayed text information, wherein the displayed text information includes the risk type, risk level, and risk consequence.

[0085] In this embodiment, text information is displayed: relevant content presented in text form in the intelligent early warning device, such as text descriptions such as "construction risk warning, risk type, risk level, and risk consequences" displayed on the screen.

[0086] In this embodiment, the risk type can be obtained by inputting various risk-related data into a pre-prepared risk type diagnosis model; For the determination of risk level, the threshold range of risk indicators can be pre-set. For example, if the risk indicator is lower than a certain value, it is a low risk level; if it is within a certain range, it is a medium risk level; and if it exceeds a certain value, it is a high risk level. The risk level is determined by comparing the actual construction risk indicators with these thresholds; The determination of risk consequences needs to be based on the specific risk type and level, and reference should be made to previous construction experience and relevant standards and specifications. For example, for high-level structural risks, the possible consequence is serious damage or even collapse of the bridge structure; for medium-level environmental risks, the consequence may be a certain delay in the construction progress; For example, if the construction risk indicator shows that the material strength is far below the design requirements, the risk type can be determined as material quality risk and the risk level is high. The risk consequence may be the need to replace a large amount of materials and cause serious delays in the construction period.

[0087] In this embodiment, the risk type, risk level, and risk consequence are respectively: Risk type: a description of the classification of risks that may occur during construction, such as structural risk, environmental risk, etc. For example, in the construction of a large-span, top-supported, open-arch bridge in a mountainous area, possible risk types include progress risk caused by insufficient material supply, foundation settlement risk caused by unstable geological conditions, etc.

[0088] Risk level: A classification used to measure the severity of a risk, usually high, medium, low, etc. For example, if a construction link has a large deviation, it is assessed as a high risk level.

[0089] Risk consequences: specific results and impacts that may occur after the risk occurs, such as arch collapse, construction delays, cost increases, etc.

[0090] The beneficial effects of the above technologies include: issuing alarms in a variety of ways, such as sound, light, and text messages, increasing the likelihood that the alarm will be noticed. Sound alarms can attract people's attention in noisy environments. Light alarms can also work at long distances or in poor visibility. Displaying text messages can clearly convey detailed information such as risk type, level, and consequences, allowing relevant personnel to have a more comprehensive understanding of the risk situation. The combination of multiple alarm methods improves the effectiveness of alarms and the accuracy of information conveyed. It helps relevant personnel respond quickly and accurately and take appropriate measures to deal with risks. It improves the effectiveness and practicality of intelligent early warning of construction risks of large-span hollow arch bridges.

[0091] Example 9 Based on Example 3, the optimal evaluation algorithm determines the submodule, including: A comprehensive performance index calculation unit, used to calculate the comprehensive performance index of each combination of the risk assessment algorithm based on the assessment accuracy and corresponding calculation weight of each combination of the risk assessment algorithm, the total operation cost of the risk assessment algorithm under each combination and the corresponding calculation weight; The best algorithm screening unit is used to take the algorithm architecture corresponding to the combination method corresponding to the maximum comprehensive performance index among all the combination methods of the risk assessment algorithm as the best risk assessment algorithm.

[0092] In this embodiment, the evaluation accuracy and the corresponding calculation weight: the evaluation accuracy is a measure of the accuracy of the risk assessment algorithm, and the calculation weight is the relative importance given to the evaluation accuracy in the comprehensive calculation. For example, the evaluation accuracy may be 80%, and the corresponding calculation weight is 0.6, indicating that it has a high proportion of importance in the comprehensive performance calculation.

[0093] In this embodiment, the total computing cost and the corresponding computing weight: the total computing cost is the sum of the resource consumption required to run the risk assessment algorithm, and the corresponding computing weight reflects its importance in the overall consideration. For example, if the total computing cost is 100 and the computing weight is 0.4, it means that its importance is relatively low when comprehensively considered.

[0094] In this embodiment, based on the evaluation accuracy and corresponding calculation weight of each combination of the risk assessment algorithm, the total computing cost of the risk assessment algorithm in each combination and the corresponding calculation weight, the comprehensive performance index of each combination of the risk assessment algorithm is calculated: the evaluation accuracy and the total computing cost are comprehensively considered, and weighted calculation is performed according to the respective calculation weights to obtain a value that can reflect the comprehensive performance of the algorithm combination. For example, if the evaluation accuracy is 75%, the weight is 0.7, the total computing cost is 80, the weight is 0.3, and the comprehensive performance index = 75% × 0.7 + 80 × 0.3.

[0095] In this embodiment, the comprehensive performance index of each combination of risk assessment algorithms is a quantitative value used to measure the comprehensive performance of a certain combination of risk assessment algorithms in terms of accuracy and resource consumption. For example, the comprehensive performance index may be 0.65, indicating the comprehensive performance of the combination in terms of accuracy and cost.

[0096] In this embodiment, the algorithm architecture corresponding to the combination method corresponding to the maximum comprehensive performance index among all the combinations of risk assessment algorithms is regarded as the best risk assessment algorithm: from the comprehensive performance indexes of all possible algorithm combinations, the one with the largest value is selected, and its corresponding algorithm architecture is determined as the best choice. For example, if there are three combinations, and their comprehensive performance indexes are 0.5, 0.6, and 0.7 respectively, then the combination method corresponding to 0.7 is the best risk assessment algorithm.

[0097] The beneficial effects of the above technologies include: the comprehensive performance index calculation unit comprehensively considers the evaluation accuracy and the total computing cost and their corresponding calculation weights, and calculates the comprehensive performance index of each combination method, making the evaluation more comprehensive and objective. The best algorithm screening unit compares the comprehensive performance indicators of all combination methods, selects the algorithm architecture corresponding to the maximum comprehensive performance index as the best risk assessment algorithm, and ensures that the selected algorithm achieves the best balance in accuracy and economy. This scientific determination method improves the reliability and practicality of the selected best risk assessment algorithm. It can provide the most suitable algorithm for the construction risk assessment of large-span hollow arch bridges, and improve the quality and efficiency of the assessment. It helps to more accurately assess construction risks, reduce the possibility of misjudgment, and save computing resources. It enhances the scientificity and effectiveness of the determination of the best algorithm in the construction risk assessment of large-span hollow arch bridges, and provides strong support for construction risk management.

[0098] Example 10 The present invention provides an intelligent early warning method based on the risk assessment of large-span open-web arch bridge construction, which is applied to any one of the intelligent early warning devices based on the risk assessment of large-span open-web arch bridge construction in embodiments 1 to 9, with reference to Figure 3 ,include: S1: Comprehensively collect data related to various risks in the construction of large-span top-decker open-arch bridges in mountainous areas; S2: Based on a large number of manual construction risk assessment examples, the assessment accuracy of each combination of risk assessment algorithms is evaluated, and the computational costs of all algorithms to be combined under each combination of risk assessment algorithms are combined to determine the best risk assessment algorithm; S3: Comprehensively analyze various risk-related data based on the best risk assessment algorithm to obtain construction risk indicators, where the risk assessment algorithm is a hybrid algorithm of neural network and fuzzy logic; S4: When the construction risk index reaches the risk threshold, an alarm is issued based on the intelligent early warning device. At the same time, targeted risk response suggestions are given based on the construction risk index.

[0099] The beneficial effects of the above technology include: Step S1 can comprehensively collect various risk-related data, providing a rich and comprehensive information basis for subsequent risk assessment. Step S2 determines the best risk assessment algorithm by evaluating the accuracy and computational cost of the algorithm, thereby improving the efficiency and accuracy of the assessment. Step S3 uses the best algorithm to comprehensively analyze the data to obtain construction risk indicators, which can more accurately quantify construction risks. Step S4 promptly issues an alarm when the risk indicator reaches the threshold, reminding relevant personnel to take measures to reduce risk losses. At the same time, targeted risk response suggestions are given, which helps to quickly formulate effective response strategies to ensure construction safety and smooth progress. Intelligent assessment, early warning and response to the construction risks of large-span hollow arch bridges are realized, improving the level and efficiency of construction risk management.

[0100] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An intelligent early warning device based on risk assessment of large-span hollow arch bridge construction, characterized in that: include: The risk-related data collection module is used to comprehensively collect various risk-related data in the construction of large-span top-decker open-arch bridges in mountainous areas; The risk assessment algorithm screening module is used to evaluate the assessment accuracy of each combination of risk assessment algorithms based on a large number of manual construction risk assessment examples, and determine the best risk assessment algorithm by combining the computational costs of all algorithms to be combined under each combination of risk assessment algorithms. The risk assessment algorithm is a hybrid algorithm of neural network and fuzzy logic. The risk assessment algorithm execution module is used to comprehensively analyze various risk-related data based on the best risk assessment algorithm to obtain construction risk indicators; The intelligent early warning and response module is used to issue an alarm based on the intelligent early warning device when the construction risk index reaches the risk threshold. At the same time, it gives targeted risk response suggestions based on the construction risk index.

2. The intelligent early warning device based on construction risk assessment of large-span hollow arch bridges according to claim 1 is characterized in that: Risk-related data collection module, including: The material performance acquisition submodule is used to acquire material performance data in real time based on material performance sensors; The construction process parameter acquisition submodule is used to obtain the structural displacement, structural stress, and structural strain in the construction process in real time as construction process parameter data based on the construction process sensor; The environmental parameter acquisition submodule is used to obtain environmental parameter data in real time based on environmental sensors; Among them, all types of risk-related data include: material performance data, construction process parameter data, and environmental parameter data.

3. The intelligent early warning device based on the construction risk assessment of a large-span hollow arch bridge according to claim 1 is characterized in that: Risk assessment algorithm screening module, including: The algorithm accuracy assessment submodule is used to assess the assessment accuracy of each combination of risk assessment algorithms based on a large number of manual construction risk assessment examples, where the combination methods include series connection, parallel connection, and fusion; The operation cost calculation submodule is used to determine the algorithm architecture of all algorithms to be combined under each combination mode of the risk assessment algorithm, and calculate the total operation cost of the risk assessment algorithm under each combination mode based on the algorithm architecture of all algorithms to be combined under each combination mode; The optimal assessment algorithm determination submodule is used to determine the optimal risk assessment algorithm based on the assessment accuracy of each combination of the risk assessment algorithms and the total computational cost of the risk assessment algorithms under each combination.

4. The intelligent early warning device based on construction risk assessment of large-span hollow arch bridges according to claim 3 is characterized in that: Algorithm accuracy evaluation submodule, including: Algorithm architecture building unit, used to build a complete algorithm architecture for each combination of risk assessment algorithms; An algorithm accuracy testing unit, used to input all risk-related data contained in all manual construction risk assessment instances into the complete algorithm architecture of each combination mode of the risk assessment algorithm, and obtain the risk assessment results of each manual construction risk assessment instance under the complete algorithm architecture of each combination mode of the risk assessment algorithm; An assessment error calculation unit is used to average the deviations between the manual risk assessment results contained in all manual construction risk assessment instances and the risk assessment results of the corresponding manual construction risk assessment instances under the complete algorithm architecture of each combination of the risk assessment algorithm, so as to obtain the assessment error of each combination of the risk assessment algorithm; The evaluation accuracy calculation unit is used to calculate the evaluation accuracy of each combination of the risk evaluation algorithms based on the evaluation error of each combination of the risk evaluation algorithms.

5. The intelligent early warning device based on construction risk assessment of large-span hollow arch bridges according to claim 3 is characterized in that: Operation cost calculation submodule, including: An algorithm architecture acquisition unit, used to acquire the algorithm architectures of all algorithms to be combined under each combination mode of the risk assessment algorithm; A single operation cost calculation unit, used for calculating each operation cost of each algorithm to be combined based on the algorithm architecture of each algorithm to be combined; A unit for determining a mixed method of single cost is used to determine a mixed calculation method of each operation cost under each combination method; A single mixed cost calculation unit is used to calculate the single mixed cost under each combination mode based on each operation cost of all algorithms to be combined under each combination mode of the risk assessment algorithm and the mixed calculation mode of the corresponding operation cost under the corresponding combination mode; The total operation cost calculation unit is used to calculate the total operation cost of the risk assessment algorithm in each combination mode based on each mixed operation cost and the corresponding calculation weight in each combination mode.

6. The intelligent early warning device based on construction risk assessment of large-span hollow arch bridges according to claim 5 is characterized in that: Single operation cost calculation unit, including: A time cost calculation subunit, used to calculate the operation time cost required for each algorithm to be combined to process unit input data based on the algorithm architecture of each algorithm to be combined; A memory cost calculation subunit, used for calculating the computational memory cost required for each algorithm to be combined to process unit input data based on the algorithm architecture of each algorithm to be combined; Among them, the computational costs of all items include computational time cost and computational memory cost.

7. The intelligent early warning device based on construction risk assessment of large-span hollow arch bridges according to claim 1 is characterized in that: Intelligent early warning and response modules, including: The risk threshold acquisition submodule is used to obtain the risk thresholds corresponding to different construction stages and different construction requirements; The intelligent early warning submodule is used to issue an alarm based on the intelligent early warning device when the construction risk index reaches the risk threshold corresponding to the current construction stage and the current construction requirements; The risk response submodule is used to retrieve the risk response suggestion knowledge base based on construction risk indicators and actual construction conditions to obtain targeted risk response suggestions.

8. The intelligent early warning device based on construction risk assessment of large-span hollow arch bridges according to claim 7 is characterized in that: The intelligent early warning device issues an alarm through one or more of sound, light, and displayed text information, wherein the displayed text information includes the risk type, risk level, and risk consequence.

9. The intelligent early warning device based on construction risk assessment of large-span hollow arch bridges according to claim 3 is characterized in that: The best evaluation algorithm determines the submodules, including: A comprehensive performance index calculation unit, used to calculate the comprehensive performance index of each combination of the risk assessment algorithm based on the assessment accuracy and corresponding calculation weight of each combination of the risk assessment algorithm, the total operation cost of the risk assessment algorithm under each combination and the corresponding calculation weight; The best algorithm screening unit is used to take the algorithm architecture corresponding to the combination method corresponding to the maximum comprehensive performance index among all the combination methods of the risk assessment algorithm as the best risk assessment algorithm.

10. An intelligent early warning method based on risk assessment of large-span hollow arch bridge construction is characterized in that: The intelligent early warning device based on the risk assessment of the construction of a large-span hollow arch bridge as described in any one of claims 1 to 9 comprises: S1: Comprehensively collect data related to various risks in the construction of large-span top-decker open-arch bridges in mountainous areas; S2: Based on a large number of manual construction risk assessment examples, the assessment accuracy of each combination of risk assessment algorithms is evaluated, and the computational costs of all algorithms to be combined under each combination of risk assessment algorithms are combined to determine the best risk assessment algorithm; S3: Comprehensively analyze various risk-related data based on the best risk assessment algorithm to obtain construction risk indicators, where the risk assessment algorithm is a hybrid algorithm of neural network and fuzzy logic; S4: When the construction risk index reaches the risk threshold, an alarm is issued based on the intelligent early warning device. At the same time, targeted risk response suggestions are given based on the construction risk index.

Images