Monitoring and early warning method and system for urban railway engineering construction

Abstract

The invention relates to a monitoring and early warning method and system for urban railway engineering construction, and solves the technical problem that safe and stable operation of key equipment cannot be effectively guaranteed due to the fact that judgment alarm is too single only according to an operation parameter threshold value of the equipment, and the method comprises the steps: employing a preset analytic hierarchy process for a risk assessment factor set, obtaining weights of different factors; calculating the risk values of the single factors one by one according to a preset different factor risk value evaluation method, and calculating the comprehensive risk value of the key equipment by adopting a preset comprehensive risk value calculation method according to the obtained risk values of the single factors and the weights of the different factors; and according to the category of the key equipment and the corresponding relationship between the risk value interval range in which the comprehensive risk value falls and the early warning processing scheme, analyzing and determining the early warning processing scheme, and executing the early warning processing scheme. The method has the following effects that the risk of key equipment is accurately evaluated, the early warning scheme is scientifically determined, and the engineering safety is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of urban rail engineering construction, and in particular to a monitoring and early warning method and system for urban rail engineering construction. Background Art

[0002] With the development of urban rail engineering construction, the safe and stable operation of various key equipment on the construction site is of vital importance. Monitoring and early warning methods play an indispensable role in ensuring the status of equipment, reducing failure losses, and ensuring the smooth progress of the project. By collecting and analyzing the operating data of key equipment and establishing a scientific and reasonable evaluation and early warning system, potential risks can be effectively dealt with.

[0003] At present, in the field of equipment monitoring and early warning, some methods only collect and analyze equipment operation data. Some technologies only focus on the threshold judgment of the equipment's own operating parameters and issue an alarm when the parameters exceed the set range.

[0004] The current technologies have the following drawbacks: On the one hand, in terms of data processing, they only simply collect and analyze operating data, which is not deep enough, and the complex relationships and potential values ​​hidden in the data are not fully explored, making it difficult to fully grasp the true status of the equipment. On the other hand, in terms of early warning mechanisms, it is too simple to judge the alarm based only on the threshold of the equipment's own operating parameters. This method is prone to false alarms and missed alarms, because equipment failures may have hidden dangers when the parameters do not exceed the threshold, or the parameters exceeding the threshold may only be a short-term fluctuation and not a real failure. It is impossible to accurately respond to potential risks and cannot effectively ensure the safe and stable operation of key equipment. Summary of the invention

[0005] In order to accurately assess the risks of key equipment, scientifically determine early warning plans, and ensure project safety, this application provides a monitoring and early warning method and system for urban rail engineering construction.

[0006] In the first aspect, the present application provides a monitoring and early warning method for urban rail engineering construction, which adopts the following technical solution: A monitoring and early warning method for urban rail engineering construction, comprising: Collect and obtain the operating data of various key equipment at the construction site; According to the mapping relationship between the categories of key equipment and the key features related to fault diagnosis, analyze and determine the key features related to fault diagnosis of various key equipment, and extract the key features related to fault diagnosis from the collected operation data of various key equipment; Use a preset specific equipment fault diagnosis model that matches the key equipment category to analyze and output the fault category distribution probability and fault severity that matches the key equipment category; Obtain the remaining factor information of different key equipment, and construct a risk assessment factor set based on the failure category distribution probability and failure severity of different key equipment and the remaining factor information; For the risk assessment factor set, the preset analytic hierarchy process is used to obtain the weights of different factors; According to the preset risk value assessment method for different factors, the risk value of each single factor is calculated one by one, and according to the risk value of each single factor and the weights of different factors, the preset comprehensive risk value calculation method is used to calculate the comprehensive risk value of key equipment; According to the correspondence between the category of key equipment and the risk value range that the comprehensive risk value falls into and the early warning processing plan, the early warning processing plan is analyzed and determined, and the early warning processing plan is implemented.

[0007] By adopting the above technical solution, by collecting the operation data of key equipment, extracting features based on the mapping relationship between equipment category and key fault features, and then using a specific fault diagnosis model to output the fault category and severity, and combining other factors to build a risk assessment factor set. Use the hierarchical analysis method to determine the weights, calculate the single and comprehensive risk values, and finally determine and execute the early warning plan based on the equipment category and the comprehensive risk value interval. This method realizes the whole process from data collection to fault diagnosis, risk assessment and effective early warning, which can comprehensively and accurately grasp the equipment status, ensure the safe operation of equipment, and reduce the adverse effects on engineering construction.

[0008] Optionally, the analysis outputs the fault category distribution probability and fault severity matching the key equipment category including: Use a preset specific equipment fault diagnosis model that matches the key equipment category to analyze and output the fault category distribution probability and fault severity that matches the key equipment category. Simultaneously, extract common features from the operating data of various types of key equipment, use a common fault diagnosis model, and analyze and output the fault category distribution probability and fault severity that matches the various types of key equipment. The voting method is used to count the votes predicted by the specific equipment fault diagnosis model and the general fault diagnosis model for each fault category, and the fault category with the most votes is selected as the most likely fault type, and its relative probability is calculated; The results of the specific equipment fault diagnosis model and the general fault diagnosis model are integrated according to the probabilities of different models to construct a comprehensive fault diagnosis model. The comprehensive fault diagnosis model is used to analyze and output the fault category distribution probability and fault severity that match the key equipment category.

[0009] By adopting the above technical solution, the fault category distribution probability and severity are output respectively by presetting a specific equipment fault diagnosis model that matches the key equipment category and a general fault diagnosis model that extracts the general characteristics of the operating data. The voting method is used to determine the most likely fault type and calculate the relative probability, making the results more scientific and reliable. The two models are then integrated to construct a comprehensive fault diagnosis model for analysis and output. This method fully utilizes the advantages of different models, reduces the error of a single model, can more accurately diagnose key equipment faults, provide accurate data support for subsequent work, and ensure the smooth progress of engineering construction.

[0010] Optionally, obtain other factor information of different key equipment, and construct a risk assessment factor set based on the fault category distribution probability and fault severity of different key equipment and other factor information, including: According to the correspondence between the preset fault categories, key equipment categories and associated factors, the associated factors associated with the fault categories output by the comprehensive fault diagnosis model are analyzed; Obtain the equipment structure integrity factors and the synergy factors of each subsystem, and integrate the correlation factors, equipment structure integrity factors and the synergy factors of each subsystem to construct a risk assessment factor set.

[0011] By adopting the above technical solution, by analyzing the fault category correlation factors output by the comprehensive fault diagnosis model, the relevant elements that affect the fault can be considered more comprehensively. The comprehensive fault diagnosis model, equipment structural integrity factors and subsystem synergy factors are integrated to construct a risk assessment factor set, so that the risk assessment covers the equipment's own conditions and system synergy, and is no longer limited to a single fault diagnosis result, so as to more accurately assess the risks of key equipment in the construction of urban rail projects and provide a more reliable basis for subsequent risk response measures.

[0012] Optionally, for the risk assessment factor set, a preset hierarchical analysis method is used to obtain the weights of different factors, including: A hierarchical model consisting of a target layer, a criterion layer, and a factor layer is established, wherein the target layer is an accurate and comprehensive risk assessment of key equipment under specific working conditions, the criterion layer includes the core criterion of faults, the equipment operation and maintenance criterion, the environmental impact criterion, and the equipment structure and system criterion, and the factor layer includes factors under the core criterion of faults, factors under the equipment operation and maintenance criterion, factors under the environmental impact criterion, and factors under the equipment structure and system criterion; For the criterion layer and the factor layer, the corresponding judgment matrix construction method is adopted to construct the judgment matrix of the criterion layer and the judgment matrix of the factor layer; According to the judgment matrix of the criterion layer and the judgment matrix of the factor layer, the weights of different factors are calculated by using the corresponding weight vector calculation method.

[0013] By adopting the above technical solution, first, a structural model with three layers including goals, criteria and factors is established to clarify the evaluation direction and various criteria and factors, making the evaluation more organized. Secondly, a judgment matrix of the criterion layer and the factor layer is constructed to scientifically quantify the relationship between the elements. Finally, the weights of different factors are obtained by calculating the weight vector to make the risk assessment more accurate. Overall, this method comprehensively considers multiple influencing factors, provides a scientific basis for accurately evaluating the comprehensive risks of key equipment under specific working conditions, and guides subsequent decision-making.

[0014] Optionally, a preset comprehensive risk value calculation method is used to calculate the comprehensive risk value of key equipment, including: Substitute the obtained single factor risk value and the weights of different factors into the formula of the preset comprehensive risk value calculation method to calculate the comprehensive risk value, where the formula of the preset comprehensive risk value calculation method is: ,in, is the comprehensive risk value, is the total number of risk factors, Indicates A single factor risk value, Indicates The weight of each risk factor.

[0015] By adopting the above technical solution, by combining the single factor risk value with the weight, the risk status of key equipment can be fully and accurately reflected. The weights of different factors are considered, the impact of important risk factors is highlighted, and a single factor is avoided from dominating the assessment results. At the same time, the clear calculation formula makes the risk assessment process operational and standardized, providing a scientific basis for the risk management of key equipment, helping to take targeted measures in a timely manner to reduce risks and ensure the safe and stable operation of key equipment.

[0016] Optionally, a preset comprehensive risk value calculation method is used to calculate the comprehensive risk value of key equipment, including: According to the corresponding relationship between the categories of key equipment and fault trees, analyze and determine the fault trees corresponding to different categories of key equipment; Use fault trees corresponding to different categories of critical equipment to analyze the interaction paths between different risk factors; According to the corresponding relationship between the interaction paths and the interaction influence coefficients between different risk factors, the influence coefficients between different risk factors are analyzed and determined; Substitute the obtained single factor risk value, the weights of different factors and the influence coefficients between different risk factors into the formula of the preset comprehensive risk value calculation method to calculate the comprehensive risk value; The formula for calculating the preset comprehensive risk value is as follows: ; in, is the comprehensive risk value, is the total number of risk factors, Indicates A single factor risk value, Indicates A single factor risk value, Indicates The weight of the risk factors, Is a factor and factors The influence coefficient between them.

[0017] By adopting the above technical solution and determining the corresponding relationship between equipment and fault tree, the interaction path of risk factors can be clearly presented. The influence coefficient is accurately analyzed, and the single factor risk value and the weights of different factors are combined and substituted into the formula to calculate the comprehensive risk value. The whole process is logically clear, comprehensive and systematic, which can accurately quantify the comprehensive risks of key equipment in the construction of urban rail projects, provide a reliable basis for subsequent risk warning and processing, ensure the stable operation of key equipment in the project, and reduce potential losses.

[0018] Optionally, according to the corresponding relationship between the category of key equipment and the risk value range into which the comprehensive risk value falls and the early warning processing plan, the early warning processing plan is analyzed and determined to include: According to the obtained comprehensive risk value, the change rate of the comprehensive risk value within the preset time window is analyzed and obtained; Analyze whether the change rate of the comprehensive risk value within the preset time window exceeds the preset upward trend threshold; If not, maintain the original setting; If yes, the early warning treatment plan is determined based on the result information that the comprehensive risk value is in an upward trend, the category of key equipment, the risk value interval range that the comprehensive risk value falls into, and the correspondence between the early warning treatment plan.

[0019] By adopting the above technical solution, by analyzing the trend of changes in the comprehensive risk value, especially paying attention to the upward trend, the risk dynamics can be detected in time. If it has not increased, the original settings are maintained to avoid excessive intervention; if it is an upward trend, the plan is determined by combining the correspondence between the key equipment category and the risk value range and the early warning plan, making the early warning processing more targeted and scientific. This is conducive to making reasonable responses to key equipment risks in the construction of urban rail projects, ensuring the safe operation of equipment, and reducing losses caused by equipment failures and the impact on project progress.

[0020] In the second aspect, the present application provides a monitoring and early warning system for urban rail engineering construction, which adopts the following technical solution: A monitoring and early warning system for urban rail engineering construction, comprising a memory, a processor and a program stored in the memory and executable on the processor, wherein the program can implement the monitoring and early warning method for urban rail engineering construction as described in the first aspect when loaded and executed by the processor. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the overall process of a monitoring and early warning method for urban rail engineering construction according to an embodiment of the present application.

[0022] Figure 2 It is a flowchart of another embodiment of the present application for analyzing and outputting the distribution probability and fault severity of fault categories that match the key equipment categories.

[0023] Figure 3 It is a flowchart of another embodiment of the present application for constructing a risk assessment factor set.

[0024] Figure 4 This is a flowchart of another embodiment of the present application that uses a preset hierarchical analysis method to obtain weights of different factors for a risk assessment factor set. DETAILED DESCRIPTION

[0025] The present application is further described in detail below in conjunction with the accompanying drawings.

[0026] Reference Figure 1 , a monitoring and early warning method for urban rail engineering construction disclosed in this application includes: Step S100, collecting and acquiring operation data of various key equipment at the construction site.

[0027] Among them, key equipment refers to equipment that has a vital impact on the progress, quality and safety of the urban rail project construction. Failure of these equipment may lead to project delays, increased costs or safety accidents. For example, shield machines, cranes, concrete mixers, track laying machines, etc. Shield machines are used to dig tunnels, and their normal operation directly determines the progress of tunnel construction; cranes are responsible for lifting heavy objects, and if they fail, they may cause heavy objects to fall, endangering the safety of personnel and equipment.

[0028] Operation data refers to various parameter information generated by key equipment during operation. These parameters can reflect the operating status of the equipment, including but not limited to the equipment's operating temperature, pressure, speed, current, voltage, displacement, vibration and other physical quantities. For example, the cutter head speed, propulsion speed, hydraulic system pressure of the shield machine; the lifting capacity, lifting height, amplitude, operating speed and other data of the crane.

[0029] The acquisition of operating data involves sensor acquisition, data collection, and data storage and management.

[0030] Among them, sensor collection mainly involves installing corresponding sensors at key parts of key equipment to collect data. For shield machines, a speed sensor is installed on the cutterhead drive motor to obtain cutterhead speed data; a pressure sensor is installed on the pipeline of the hydraulic system to measure the pressure of the hydraulic oil; a displacement sensor is installed at the propulsion cylinder to monitor the propulsion distance and speed of the shield machine. For cranes, speed sensors and torque sensors are installed on the motor of the lifting mechanism to obtain data related to lifting speed and lifting force; an angle sensor is installed on the lifting arm to measure the amplitude of the lifting arm; lifting weight data is obtained at the lifting weight limiter; and a speed sensor is installed at the wheel of the walking mechanism to monitor the walking speed of the crane.

[0031] Data acquisition is to build a data acquisition system to collect the data collected by each sensor in a centralized manner. The system can transmit sensor data to data acquisition terminals (such as industrial computers, data acquisition cards, etc.) through wired (such as cables) or wireless (such as Wi-Fi, Bluetooth, ZigBee, etc.) communication methods. The data acquisition system needs to have certain data processing capabilities, such as preliminary filtering of the collected data to remove noise interference and ensure data quality.

[0032] The process of data storage and management is as follows: the collected data needs to be stored in a suitable storage medium, such as a hard disk, solid-state drive, etc. A database management system (such as MySQL, SQLServer, Oracle, etc.) can be used to manage the data, and different types of data from different devices can be stored according to a certain structure (such as data tables, data fields). At the same time, a data backup mechanism is established to back up data regularly to prevent data loss. In addition, in order to facilitate data query and retrieval, the data needs to be indexed and classified, such as classified storage according to device name, collection time, data type, etc.

[0033] Step S200, according to the mapping relationship between the categories of key equipment and the key features related to fault diagnosis, analyze and determine the key features related to fault diagnosis of each type of key equipment, and extract the key features related to fault diagnosis from the collected operation data of each type of key equipment.

[0034] Mapping relationship between key equipment categories and key features related to fault diagnosis: This is a pre-established relationship that specifies the corresponding relationship between key parameters or status features related to different types of key equipment that can reflect the fault situation when a fault occurs. For example, for a shield machine, there is a specific mapping between features such as the degree of tool wear, changes in oil temperature in the hydraulic system, and current fluctuations in the electrical system and the types of faults that may occur in the shield machine; for a crane, features such as the wear of the wire rope, the braking distance of the brake, and the signal stability of the electrical control system are associated with the fault diagnosis of the crane. This mapping relationship is determined based on the working principle of the equipment, historical fault data, and expert experience.

[0035] Key features related to fault diagnosis: refers to important parameters or status performance that can effectively indicate whether key equipment has a fault, as well as the type and severity of the fault. These features are crucial for timely and accurate diagnosis of equipment faults. For example, abnormally high temperature of a certain component of the equipment, vibration amplitude beyond the normal range, reduced operating efficiency, etc. may all be key features related to the fault.

[0036] The process of extracting key features related to fault diagnosis from the collected operation data of various key equipment is as follows: First, based on the mapping relationship between the categories of key equipment and the key features related to fault diagnosis, the key features corresponding to different categories of key equipment are clarified. For example, for a certain type of construction equipment, its key features may include specific operating parameters, temperature change trends, vibration amplitudes, etc.

[0037] Then, the collected operation data of this type of key equipment is analyzed and screened. Data processing techniques, such as data filtering and feature extraction algorithms, are used to find the data portion corresponding to the determined key features from a large amount of operation data. For example, if the key feature is a specific operation parameter, the value of the parameter is extracted from the operation data; if it is a temperature change trend, it may be necessary to analyze and calculate the temperature data over a period of time to obtain trend information. In this way, the key features related to fault diagnosis can be accurately extracted.

[0038] Data filtering mainly involves filtering the effective data range. The effective data range is set according to the normal operating range of key equipment and the requirements of fault diagnosis. Data outside this range may be abnormal data or invalid data, which can be filtered out.

[0039] The feature extraction algorithm can be feature extraction based on statistical analysis or feature extraction based on signal processing.

[0040] Feature extraction based on statistical analysis can be done in the following ways: 1. Mean and standard deviation: For continuous operating data, such as temperature and pressure, the mean and standard deviation of the data over a period of time can be calculated. The mean reflects the central trend of the data, and the standard deviation reflects the degree of dispersion of the data. If the mean or standard deviation changes significantly, it may mean that the operating state of the equipment has changed. For example, the oil temperature data of an engine, under normal circumstances, the mean and standard deviation are relatively stable. If the mean of the oil temperature suddenly rises and the standard deviation increases, it may indicate that the engine is at risk of overheating failure. 2. Peak and valley values: For vibration, current and other data, their peak and valley values ​​can be detected. The size and frequency of peak and valley values ​​can be used as features for fault diagnosis. For example, in the current data of the motor, if large peak currents appear frequently, it may be that the motor is overloaded or there is a short circuit or other faults. 3. Trend analysis: By performing linear regression or other trend analysis methods on the data over a period of time, the trend of data changes can be obtained. If the trend is rising, falling or fluctuating greatly, it may be a sign of equipment failure. For example, for the wear condition of key equipment, it is possible to determine whether the degree of wear is gradually increasing by monitoring the trend of its displacement data.

[0041] Feature extraction based on signal processing can be done in the following ways: 1. Fourier transform: For periodic signals, such as vibration signals, Fourier transform can be used to convert time domain signals into frequency domain signals and analyze the frequency components of the signals. Different faults may produce characteristic frequencies within a specific frequency range, and the faults can be diagnosed by detecting these characteristic frequencies. For example, for rotating machinery, an unbalanced fault will produce a large vibration amplitude at a specific speed frequency. This frequency component can be detected by Fourier transform to determine whether an unbalanced fault exists. 2. Wavelet transform: Wavelet transform can analyze signals in both the time domain and the frequency domain, and has a good processing effect on non-stationary signals. It can be used to detect mutation points, singular values, etc. in the signal, which may be signs of a fault. For example, in the fault diagnosis of electrical equipment, wavelet transform can detect instantaneous mutations in the current signal, which may be caused by faults such as short circuits or open circuits.

[0042] Step S300: using a preset specific equipment fault diagnosis model that matches the key equipment category, analyzing and outputting the fault category distribution probability and fault severity that matches the key equipment category.

[0043] The preset specific equipment fault diagnosis model that matches the key equipment category refers to a mathematical model or algorithm developed specifically for a certain type of key equipment. It is built based on the physical structure, working principle, a large amount of historical fault data, and domain expert knowledge of the equipment. Different types of key equipment (such as shield machines, cranes, etc.) have their own corresponding specific equipment fault diagnosis models, which can accurately analyze the unique failure modes and related parameters of the equipment, and output fault information that matches the equipment category.

[0044] Fault category distribution probability: refers to the probability of occurrence of different types of faults that may occur in the equipment. For example, for a shield machine, its fault categories may include cutter head tool failure, hydraulic system failure, electrical system failure, etc. The fault category distribution probability is the probability value of each fault type occurring in the current equipment operating state, such as the probability of cutter head tool failure is 0.3, the probability of hydraulic system failure is 0.2, etc. The sum of these probability values ​​is 1.

[0045] Fault severity: used to measure the degree of impact of equipment failure on the operation of the equipment itself and the entire project construction. Generally, it can be divided into different levels such as minor, moderate, and severe. For example, a slight wear of the shield machine cutter can be considered a minor fault, while a serious deformation of the cutter head that makes it impossible to dig normally is a serious fault; for a crane, a small vibration of the lifting mechanism may be a minor fault, but a broken wire rope is a serious fault.

[0046] Using a preset specific equipment fault diagnosis model that matches the key equipment category, analyzing and outputting the fault category distribution probability and fault severity that matches the key equipment category involves the following steps, which are as follows: The model input parameters involve the following: The input of the specific equipment fault diagnosis model is the data related to the operating status of key equipment, which has been collected and processed in steps S100 and S200. Taking the shield machine as an example, the input parameters may include the cutter head speed, propulsion speed, hydraulic system pressure, flow and temperature, electrical system current and voltage, tool wear degree, etc. For cranes, the input parameters may include lifting weight, lifting height, lifting speed, amplitude, wind speed, wire rope tension, electrical system parameters, etc.

[0047] The model principles and algorithms involved are as follows: 1. Methods based on physical models: For some equipment, mathematical models can be established based on their physical principles. For example, the cutter head cutting force model of the shield machine calculates the cutting force through parameters such as the geometric shape of the tool, the mechanical properties of the rock and soil, the cutter head rotation speed and the propulsion speed. When the cutting force changes abnormally, combined with the tool wear model, the probability and severity of the tool failure can be inferred. The hydraulic system can establish a pressure-flow model based on the principles of fluid mechanics, and judge the fault situation by monitoring the deviation between the actual pressure and flow and the theoretical model. 2. Data-driven methods: Using a large amount of historical fault data and normal operation data, machine learning algorithms (such as neural networks, support vector machines, etc.) are used to train fault diagnosis models. Taking a crane as an example, the parameters such as lifting weight, lifting height, wind speed in the historical data can be used as input features, and the fault category and severity can be used as output labels to train the model. When new data is input, the model can output the probability and severity of the fault category distribution in the current state. 3. Expert system-based methods: The knowledge and experience of domain experts are converted into rules to build an expert system. For example, for a shield machine, if expert experience shows that the probability of hydraulic system failure increases when the hydraulic oil temperature exceeds a certain threshold and the pressure fluctuation exceeds a certain range, and the severity of the failure can be determined based on the size of the temperature and pressure deviations, these rules are encoded into the expert system for fault diagnosis.

[0048] Model output results: The fault category distribution probability output by the model is a probability vector, and each element corresponds to the probability of a fault type. For example, the output of the shield machine fault diagnosis model may be [probability of cutter head failure 0.3, probability of hydraulic system failure 0.2, probability of electrical system failure 0.1, probability of other failures 0.4]. The severity of the fault is output in the form of a grade (such as minor, moderate, severe) or a quantitative value (such as a value between 0-1, 0 means no fault, 1 means extremely severe fault).

[0049] Step S400, obtaining other factor information of different key equipment, and constructing a risk assessment factor set according to the fault category distribution probability and fault severity of different key equipment and other factor information.

[0050] Other factors: refers to all kinds of information related to key equipment that may affect its risk status, in addition to the probability of fault category distribution and fault severity obtained through the analysis of key equipment operation data. For example, the age of the equipment. A newly commissioned crane and a crane that has been in use for many years may have different overall risk status due to their different years of use, even if the current operation data shows similar fault conditions. Another example is the equipment maintenance record. Equipment that undergoes regular and standardized maintenance may have relatively lower risks, while equipment that lacks good maintenance records may have higher risks. It also includes the working environment of the equipment. If the crane works for a long time in severe weather conditions (such as areas with frequent heavy rain and strong winds) or complex geological conditions (such as rugged terrain, soft soil, etc.), the risks it faces will increase accordingly.

[0051] Risk assessment factor set: is a set that integrates all factors related to the risk assessment of key equipment. These factors together form the basis for a comprehensive assessment of the risk of key equipment. The elements in the risk assessment factor set cover fault-related information (fault category distribution probability and fault severity) and other related factors. Through the comprehensive analysis of these factors, the risk level of key equipment can be accurately measured.

[0052] Equipment service life: This can be accurately determined through the equipment's purchase records, commissioning time, and other information. For example, by checking the equipment purchase contract signing date and the time point when the equipment was first put into use as recorded on the construction site, the difference between these two time points can clearly determine the equipment's service life.

[0053] Maintenance records: Construction units usually have special equipment maintenance records, which record in detail the time, content, parts and components of each maintenance of key equipment. By consulting these records, you can obtain complete maintenance records of the equipment.

[0054] Working environment information: On the one hand, real-time weather information can be obtained through meteorological monitoring equipment at the construction site, and the common weather characteristics of the area can be understood in combination with long-term meteorological statistical data to determine the weather environment conditions faced by the equipment. On the other hand, for working environment information below the ground, such as geological conditions, geological survey reports can be used. These reports describe in detail the soil quality, stratigraphic structure, etc. underground at the construction site, thereby clarifying the geological working environment in which the equipment is located.

[0055] Step S500: using a preset analytic hierarchy process to obtain weights of different factors for a risk assessment factor set.

[0056] AHP: This is a decision-making method that decomposes decision-related elements into goals, criteria, plans, and other levels, and conducts qualitative and quantitative analysis on this basis. In this method, it is used to assign reasonable weights to different factors in the risk assessment factor set to reflect the relative importance of each factor in assessing the risk of key equipment. For example, when assessing the risk of a shield machine, the severity of the fault may have a greater impact on the overall risk than the age of the equipment in some cases. The purpose of AHP is to determine the quantitative relationship of this relative importance.

[0057] The process of obtaining the weights of different factors involves the following process, which are as follows: Constructing a hierarchical model: First, determine the target layer, i.e., the comprehensive risk assessment of key equipment; then determine the criterion layer, including the risk assessment factors such as the probability of fault category distribution, fault severity, equipment service life, maintenance records, and working environment obtained in step S400; and finally, the solution layer. If the risk assessment is for multiple key equipment of the same type, each equipment can be used as a solution. For example, for multiple cranes at a construction site, each crane can be analyzed as a solution.

[0058] Construct a judgment matrix: Through expert evaluation, historical data statistical analysis and other methods, the relative importance of each factor in the criterion layer is compared pairwise. For example, invite experienced engineering and technical experts, equipment maintenance experts, etc. to judge the relative importance of fault severity and equipment service life in risk assessment based on their professional knowledge and experience. If the expert believes that the fault severity is much more important than the equipment service life in risk assessment, the corresponding value can be assigned according to a certain scaling method (such as 1-9 scaling method).

[0059] Calculate the weight vector and perform consistency check: Calculate the weight vector of each factor through mathematical calculation methods (such as the eigenvector method, etc.). After the calculation is completed, a consistency check is required to ensure the rationality of the judgment matrix. If the consistency check fails, the judgment matrix needs to be readjusted until it passes. For example, when calculating the weights of each risk assessment factor of a crane, if the calculated weight vector fails the consistency check, it is necessary to re-examine whether the relative importance judgment given by the expert is reasonable, and if necessary, readjust the judgment matrix.

[0060] Step S600, according to the preset risk value assessment method for different factors, the risk value of each single factor is calculated one by one, and according to the risk value of each single factor and the weights of different factors, the preset comprehensive risk value calculation method is used to calculate the comprehensive risk value of the key equipment.

[0061] Single factor risk value: refers to the risk quantification value determined for a single risk assessment factor when assessing the risk of key equipment. Each factor has its own impact on the equipment risk, and this impact is converted into a numerical value through a specific method for comprehensive assessment. For example, for the factor of the severity of the crane's fault, its single factor risk value reflects the individual contribution of the severity of the fault to the overall risk of the crane.

[0062] The calculation of single factor risk value can be based on historical data and empirical formulas. Specifically, for some common key equipment and risk factors, a large amount of data accumulated in previous similar projects can be used to summarize empirical formulas to calculate single factor risk values. For example, for a certain fault category of a crane, an empirical formula for fault category and risk value is established based on the number of times the fault has occurred in history, the degree of accident loss caused, and other data. Assuming that the higher the frequency of a certain fault and the greater the historical accident loss, the higher the single factor risk value corresponding to the fault category.

[0063] Comprehensive risk value: A quantitative value of the overall risk of key equipment obtained by comprehensively considering all risk assessment factors and their weights. It comprehensively reflects the risk level of key equipment under the combined effects of multiple factors such as current operating status, failure conditions, service life, maintenance and working environment. For example, the comprehensive risk value of a shield machine combines the failure risk of each key component of the shield machine, the wear risk caused by the length of use, and the working environment risk under different geological conditions.

[0064] The comprehensive risk value calculation can adopt the weighted summation method. Specifically, the weights of different factors obtained in step S500 are used to multiply the single factor risk value of each factor by its weight, and then all the products are added to obtain the comprehensive risk value of the key equipment. For example, it is known that the single factor risk value of the fault severity of the crane is 30, and its weight is 0.4; the single factor risk value of the equipment service life is 20, and its weight is 0.3; the single factor risk value of the working environment is 15, and its weight is 0.3. Then the comprehensive risk value of the crane = 30×0.4+20×0.3+15×0.3=22.5. In this way, the influence of various factors is combined to comprehensively evaluate the risks of key equipment.

[0065] Step S700, analyzing and determining the early warning processing scheme according to the corresponding relationship between the category of the key equipment and the risk value interval range into which the comprehensive risk value falls and the early warning processing scheme, and executing the early warning processing scheme.

[0066] Risk value interval range: Different intervals are divided according to the size of the comprehensive risk value. Each interval represents a different degree of risk level. For example, it can be divided into a low risk interval (0-30), a medium risk interval (31-70), and a high risk interval (71-100).

[0067] Early warning and handling plan: Pre-set response measures for different risk value intervals to reduce equipment risks and avoid accidents. For example, for low risks, it may be necessary to strengthen daily inspections; for medium risks, it may be necessary to arrange for professional personnel to conduct focused inspections and prepare spare equipment; for high risks, it may be necessary to suspend equipment operation for comprehensive maintenance.

[0068] The process of determining the risk value range of the comprehensive risk value is as follows: after calculating the comprehensive risk value of the key equipment, compare it with the pre-set risk value range division standard. For example, if the comprehensive risk value of a crane is calculated to be 50, it can be known that it is in the medium risk range according to the set low, medium and high risk ranges.

[0069] The process of determining the early warning treatment plan is as follows: It is determined based on the company's internal safety management regulations, industry standards, and previous experience in handling similar projects. For example, in the construction industry, for hoisting equipment in a high-risk state, there is usually a standard handling process, including stopping operations, evacuating surrounding personnel, conducting a comprehensive inspection of the equipment, and troubleshooting. These processes can serve as the basis for the early warning treatment plan.

[0070] Assume that there is a shield machine at a construction site, and its comprehensive risk value is calculated to be 75 through the previous steps. First, according to the category of key excavation equipment and the fact that 75 is in the high-risk range (71-100), according to the pre-established early warning treatment plan, the shield machine operation needs to be stopped immediately, and a professional shield machine maintenance team needs to be organized to conduct a comprehensive inspection of the machine, including the wear of the cutter, the hydraulic system, the electrical system and other key parts. At the same time, the construction environment in the tunnel needs to be re-evaluated. After ensuring that all risk factors are effectively controlled, a decision can be made whether to restart the shield machine operation.

[0071] Reference Figure 2 , the analysis outputs the fault category distribution probability and fault severity matching the key equipment categories, including: Step S310, using a preset specific equipment fault diagnosis model that matches the key equipment category, analyzes and outputs the fault category distribution probability and fault severity that matches the key equipment category. Simultaneously, common features in the operating data of various types of key equipment are extracted, and a common fault diagnosis model is used to analyze and output the fault category distribution probability and fault severity that matches the various types of key equipment.

[0072] Specifically, specific equipment fault diagnosis models: fault diagnosis models specially built for different types of key equipment. These models are built based on information such as the structure, working principle, operation characteristics, and a large amount of historical fault data of specific equipment. For example, for a specific key equipment such as a shield machine, its fault diagnosis model will take into account the data and characteristics closely related to the operation of the shield machine, such as the shield machine cutter head speed, thrust pressure, and sealing system parameters, to accurately diagnose the possible fault types of the shield machine.

[0073] Common features: These are some common characteristic attributes in the operation data of various key equipment, such as the operating temperature, operating current, vibration frequency, etc. of the equipment. These features can be used as the basis for judging whether the equipment is operating normally in different types of key equipment.

[0074] Universal fault diagnosis model: A fault diagnosis model established based on the common features in the operating data of various key equipment. It is not targeted at a specific type of equipment, but a universal diagnosis model applicable to a variety of equipment, which can make a preliminary judgment on equipment failure from a macro perspective.

[0075] The construction and application of a specific equipment fault diagnosis model involves the following contents, as follows: 1. Model construction: Collect a large amount of operating data of a specific category of key equipment in normal operation and fault state. These data can come from previous engineering projects, test data of equipment manufacturers, and real-time monitoring data on site. Then, use machine learning algorithms (such as neural networks, support vector machines, etc.) or rule-based expert system methods to build models based on the professional knowledge and data characteristics of the equipment. For example, for a crane, by collecting normal operating data under different lifting weights and different working ranges, as well as data from previous situations such as brake failures and wire rope wear failures, a specific fault diagnosis model for the crane is trained using a neural network algorithm. 2. Output analysis: Input the currently collected operating data of a specific key equipment into the corresponding specific equipment fault diagnosis model. The model outputs the fault category distribution probability and fault severity that match the key equipment category through internal algorithms and parameter calculations. For example, input the current operating data of a shield machine into the shield machine fault diagnosis model, and the model outputs a cutter head failure probability of 30% (fault category distribution probability) and a medium severity (fault severity).

[0076] Universal feature extraction: Filter out universal features from the operating data of various key equipment. This can be determined through statistical analysis of data from different types of equipment and the experience of domain experts. For example, through analysis, it is found that whether it is a shield machine, a crane or other key equipment, the operating temperature outside the normal range may indicate a failure, so the operating temperature is a universal feature. Process all the collected operating data of key equipment to extract the values ​​of these universal features.

[0077] Construction and output analysis of general fault diagnosis model: Using the extracted general feature data, a general fault diagnosis model is also constructed using machine learning or other appropriate methods. The general feature values ​​in the current key equipment operation data are input into the model, and the model outputs the distribution probability and fault severity of the matching fault categories of various key equipment. For example, the model determines the probability of a key equipment failure and the approximate severity of the failure based on general features such as excessive operating temperature and abnormal vibration.

[0078] Step S320 , using a voting method, counting the number of votes predicted by the specific equipment fault diagnosis model and the general fault diagnosis model for each fault category, selecting the fault category with the most votes as the most likely fault type, and calculating its relative probability.

[0079] Voting method: A simple method to combine the prediction results of multiple models. In this scenario, the specific equipment fault diagnosis model and the general fault diagnosis model are regarded as different "voters", and the prediction of each model on the fault category is regarded as a "vote", and the most likely fault category is determined according to the number of votes.

[0080] Voting statistics involve the following process: For each fault category, the prediction results of the specific equipment fault diagnosis model and the general fault diagnosis model are recorded separately (considered as voting). For example, if the specific equipment fault diagnosis model predicts fault category A and the general fault diagnosis model also predicts fault category A, then fault category A gets 2 votes; if the specific equipment fault diagnosis model predicts fault category B and the general fault diagnosis model predicts fault category A, then fault category A gets 1 vote and fault category B gets 1 vote. Traverse all fault categories and count the votes for each fault category.

[0081] Determine the most likely fault type and calculate relative probability: Select the fault category with the most votes as the most likely fault type. For example, in the above example, if fault category A has the most votes, then A is the most likely fault type. Calculate relative probability: Divide the number of votes for the most likely fault type by the total number of votes to get its relative probability. Assuming the total number of votes is 3 (two models predict 3 different fault categories, each model casts 1 vote), and fault category A has 2 votes, then the relative probability of fault category A is 2 / 3.

[0082] Step S330, according to the probabilities of different models, the results of the specific equipment fault diagnosis model and the general fault diagnosis model are integrated to form a comprehensive fault diagnosis model, and the comprehensive fault diagnosis model is used to analyze and output the fault category distribution probability and fault severity that match the key equipment category.

[0083] Comprehensive fault diagnosis model: A new model built by integrating the specific equipment fault diagnosis model and the general fault diagnosis model. It combines the advantages of both and can more accurately analyze and output the fault category distribution probability and fault severity that match the key equipment category.

[0084] The fusion of model results involves the following processes, which are as follows: 1. Determine the weight: According to the probabilities obtained by the specific equipment fault diagnosis model and the general fault diagnosis model in step S320, determine the weights of the two models in the fusion process. Generally speaking, the model with a higher probability is given a larger weight, and the model with a lower probability is given a smaller weight. For example, if the prediction probability of a certain fault category by the specific equipment fault diagnosis model is 70%, and the prediction probability of the fault category by the general fault diagnosis model is 40%, then the weight of the specific equipment fault diagnosis model can be set to 0.7, and the weight of the general fault diagnosis model can be set to 0.3. 2. Weighted fusion: For each fault category, the fault category distribution probability and fault severity output by the specific equipment fault diagnosis model and the general fault diagnosis model are weighted and fused respectively. The specific calculation formula is as follows: Fault category distribution probability after fusion = Fault category distribution probability of the specific equipment fault diagnosis model × weight of the specific equipment fault diagnosis model + Fault category distribution probability of the general fault diagnosis model × weight of the general fault diagnosis model. Fault severity after fusion = Fault severity of the specific equipment fault diagnosis model × weight of the specific equipment fault diagnosis model + Fault severity of the general fault diagnosis model × weight of the general fault diagnosis model.

[0085] The comprehensive fault diagnosis model is used to analyze and output the fault category distribution probability and fault severity that match the key equipment category. These output results combine the advantages of the specific equipment fault diagnosis model and the general fault diagnosis model to more accurately reflect the fault status of the key equipment. For example, the output result may be "Key equipment fault category distribution probability: mechanical failure 65%, electrical failure 25%, hydraulic failure 10%; fault severity is above medium."

[0086] Reference Figure 3 , obtain the remaining factor information of different key equipment, and construct a risk assessment factor set based on the failure category distribution probability and failure severity of different key equipment and the remaining factor information, including: Step S410: Analyze the associated factors associated with the fault category output by the comprehensive fault diagnosis model according to the preset correspondence between the fault category, the key equipment category and the associated factors.

[0087] Related factors: Other conditions or factors associated with specific fault categories and key equipment categories that may affect the occurrence and development of faults and the extent to which the fault affects equipment and construction. For example, for the TBM tool wear fault category, rock hardness, TBM propulsion speed, etc. are related factors.

[0088] The process of establishing the correspondence between preset fault categories, key equipment categories and associated factors involves the following: 1. Data collection: Collect technical information of different types of key equipment, including instructions, maintenance manuals, technical specifications, etc. provided by equipment manufacturers, to understand the structure, performance and working principle of the equipment. Collect historical fault data in urban rail engineering construction, including the type of equipment where the fault occurred, the fault category, the operating parameters and environmental conditions when the fault occurred, etc. Consult industry experts and engineers to obtain their experience and knowledge in actual engineering, and understand the common faults and associated factors of different key equipment. 2. Analysis and collation: Classify and collate the collected data, and classify them according to key equipment categories, fault categories and associated factors. For example, divide the fault data of all cranes into categories such as lifting mechanism faults and braking system faults, and sort out the associated factors corresponding to each fault category. Use data analysis methods, such as correlation analysis and regression analysis, to find out the potential relationship between fault categories and associated factors. For example, by analyzing historical data, it is found that when a crane lifting mechanism fault occurs, the degree of wear of the wire rope has a high correlation with the probability of the fault occurring. 3. Establish a corresponding relationship table: According to the results of analysis and collation, establish a corresponding relationship table of preset fault categories, key equipment categories and related factors. It can be stored and managed in the form of tables, databases, etc., so as to facilitate quick query and use in practical applications.

[0089] The specific steps for analyzing the correlation factors are as follows: 1. Obtain the output of the comprehensive fault diagnosis model: First, obtain the fault category information of the key equipment from the comprehensive fault diagnosis model. This fault category is obtained by the model based on the operating data and feature analysis of the key equipment. 2. Determine the key equipment category: According to the equipment where the fault occurs, determine the key equipment category to which it belongs. For example, if the fault occurs on a crane, then the key equipment category is the crane category. 3. Find the corresponding relationship: Based on the determined fault category and key equipment category, find the factors associated with it in the preset corresponding relationship table. By querying the table or database, quickly find the correlation factors corresponding to the specific fault category and key equipment category. 4. Analyze the impact of the correlation factors: Further analyze the found correlation factors to understand their impact on the occurrence and development of the fault and on the equipment and construction. For example, analyze the impact mechanism of rock hardness on the wear failure of the shield machine cutter and how to reduce this impact by adjusting the construction parameters.

[0090] Step S420, obtaining the equipment structure integrity factor and the synergy factor of each subsystem, and integrating the correlation factor, the equipment structure integrity factor and the synergy factor of each subsystem to construct a risk assessment factor set.

[0091] Integration of comprehensive fault diagnosis model: Directly incorporate the fault category distribution probability and fault severity output by the comprehensive fault diagnosis model into the risk assessment factor set. For example, the comprehensive fault diagnosis model of a shield machine outputs a cutter failure probability of 30% and a medium severity, and these data are put into the set as one of the factors. The model output results can be obtained through the programs and algorithms that previously built and run the comprehensive fault diagnosis model. These programs collect data in real time, analyze and output corresponding results during the operation of the equipment.

[0092] Equipment structural integrity factors: Involves the integrity of each component of key equipment and its connection structure. For example, for a shield machine, it includes whether the cutter head, shield, propulsion system and other structural components have cracks, deformation, wear, etc. These factors directly affect the stability and normal operation of the equipment and are a key part of risk assessment.

[0093] The integration of equipment structural integrity factors is as follows: 1. Inspection method: a combination of regular manual inspection and automated monitoring equipment is used. For large key equipment, such as shield machines, during manual inspection, engineers will visually inspect each structural component of the equipment and use measuring tools (such as calipers to measure the wear of components) according to the predetermined inspection route and standards. At the same time, sensors installed on the equipment (such as strain gauges to measure structural stress and displacement sensors to detect component displacement) will collect data in real time and feed it back to the monitoring system through the data transmission system. 2. Information recording and organization: record and organize the results of manual inspections and automated monitoring data. For example, after each inspection, engineers will record the wear of the structural components and the presence or absence of cracks in the equipment maintenance file; the automated monitoring system will store the data collected by the sensors in a database in a time series. After these data are processed and analyzed, the information related to the structural integrity of the equipment is incorporated into the risk assessment factor set in a quantitative or qualitative form.

[0094] Synergy factors of each subsystem: Key equipment is usually composed of multiple subsystems, such as the lifting subsystem, operation subsystem, and electrical control subsystem of a crane. The synergy factors of each subsystem refer to the ability of these subsystems to cooperate and coordinate with each other. If there is a problem with the coordination between subsystems, it may cause equipment failure or abnormal operation.

[0095] The integration of the synergy factors of each subsystem is as follows: 1. Subsystem performance test: Regularly conduct individual performance tests and joint operation tests on subsystems. Taking the crane as an example, the lifting subsystem is tested for its lifting speed, lifting weight and other performance indicators, and the running subsystem is tested for its walking speed, braking distance, etc. In the joint test, observe the coordination when the lifting and walking operations are carried out simultaneously. These tests can be completed by professional testing equipment and instruments, such as using high-precision weighing sensors to test the lifting weight and laser rangefinders to measure the braking distance. 2. Data analysis and synergy evaluation: Evaluate the synergy between subsystems by analyzing the data of the subsystems in individual and joint operation tests. For example, if it is found in the joint test that the crane is stuck in walking during the lifting process, find out the problems that may affect the synergy by analyzing the relevant parameters of the lifting and running subsystems (such as motor current, speed control signal, etc.). The evaluation results of the synergy of these subsystems are included in the risk assessment factor set in a certain form (such as qualitative descriptions such as good synergy, general, poor, or quantitative methods such as synergy coefficient).

[0096] Reference Figure 4 ,For the risk assessment factor set, the preset hierarchical analysis method is used to obtain the weights of different factors, including: Step S510, establishing a hierarchical model including a target layer, a criterion layer and a factor layer.

[0097] Among them, the target layer is the precise and comprehensive risk assessment of key equipment under specific working conditions; the criterion layer includes core fault criteria, equipment operation and maintenance criteria, environmental impact criteria, equipment structure and system criteria; the factor layer includes factors under core fault criteria, factors under equipment operation and maintenance criteria, factors under environmental impact criteria, and factors under equipment structure and system criteria.

[0098] Target layer: It is the highest level of the entire hierarchy and represents the ultimate goal we want to achieve. This refers to the precise and comprehensive risk assessment of key equipment under specific working conditions, that is, a comprehensive measurement of the risk level of the equipment under current working conditions.

[0099] Criteria layer: Below the target layer, it is several main criteria obtained by decomposing the target. The core criterion of failure focuses on the content directly related to equipment failure; the equipment operation and maintenance criterion focuses on the factors related to the daily operation and maintenance of the equipment; the environmental impact criterion considers the impact of the external environment of the equipment on its risk; the equipment structure and system criterion (here it may be "equipment structure and system criterion") mainly involves the aspects related to the equipment's own structure and internal systems. These criteria provide a basis for comprehensive risk assessment from different angles.

[0100] Failure core criteria: By analyzing previous key equipment failure cases, we summarize the key aspects directly related to the failure itself and summarize them into failure core criteria. These cases include failure records of the same type of equipment in different engineering projects, failure statistics provided by equipment manufacturers, etc. For example, a large number of shield machine failure cases show that cutter head failure and hydraulic system failure are common problems that have a significant impact on equipment operation, which are all related to the failure core criteria.

[0101] Equipment operation and maintenance criteria: Determined based on the equipment operation manual, maintenance procedures and engineering practice experience. The equipment operation manual specifies the parameter range and operation process for normal operation of the equipment. From this information, factors related to operation can be extracted, such as the equipment's operating speed and load conditions. The maintenance procedures specify the equipment's maintenance cycle and maintenance content, such as regular replacement of wire ropes for cranes and lubrication of joints. These relevant information constitute the factors of the equipment operation and maintenance criteria.

[0102] Environmental impact criteria: determined through analysis of construction site environmental monitoring data and environmental assessment reports. Environmental factors at the construction site include meteorological conditions (temperature, humidity, wind speed, etc.) and geological conditions (soil type, groundwater level, etc.). For example, in a high temperature environment, the heat dissipation of equipment may affect its performance; in soft soil conditions, the foundation stability of equipment such as cranes may be affected. These environmental factors are included in the environmental impact criteria.

[0103] Equipment structure and system criteria: Determined based on the equipment design drawings and technical specifications. The design drawings show the various structural components of the equipment and how they are connected, and the technical specifications describe in detail the functions and parameters of each subsystem of the equipment. For example, the strength of the shield structure of the shield machine, the coordination of the propulsion system and the cutterhead system, and other structural and system-related contents all fall under this criterion.

[0104] Factor layer: It is a further subdivision of the criterion layer. Each criterion contains multiple specific factors. For example, under the core criterion of fault, there may be factors such as fault type and fault severity; under the equipment operation and maintenance criterion, there may be factors such as equipment operation time and maintenance frequency.

[0105] Factors under the core criteria of faults: Determined by combining the output of the fault diagnosis model and the fault analysis report. For example, the fault category distribution probability and fault severity output by the fault diagnosis model are important factors. At the same time, refer to the location and fault mode of the fault in the fault analysis report to further refine the factors, such as whether the electrical equipment fault is a short circuit fault or an open circuit fault.

[0106] Factors under the equipment operation and maintenance criteria: obtained from equipment operation records and maintenance records. Operation records include equipment operation time, operation frequency, load changes, etc.; maintenance records include the time of each maintenance, maintenance items, replaced parts, etc. For example, a long equipment operation time may increase the risk of failure, and a low maintenance frequency may result in potential equipment problems not being discovered in time. These are all factors under this criterion.

[0107] Factors in the environmental impact criteria: Based on environmental monitoring data and research results on the impact of the environment on the equipment. Such as the duration of extreme temperatures and humidity changes in meteorological data, and changes in soil bearing capacity in geological data. For different key equipment, the degree of impact of environmental factors varies, and specific factors need to be determined based on the characteristics of the equipment. For example, for equipment operating in the open air, meteorological factors are more critical; for equipment with a foundation fixed underground, geological factors have a greater impact.

[0108] Factors under the equipment structure and system criteria: obtained from the detailed inspection report of the equipment and the system performance test data. The detailed inspection report includes the flaw detection results of the equipment structure components, wear measurement data, etc.; the system performance test data reflects the operating parameters of each subsystem, such as the voltage stability of the electrical system, the transmission efficiency of the mechanical transmission system, etc. These data can determine factors such as the integrity of the equipment structure and the coordination of each subsystem.

[0109] Step S520: For the criterion layer and the factor layer, a judgment matrix of the criterion layer and a judgment matrix of the factor layer are constructed by using the corresponding judgment matrix construction method.

[0110] Judgment matrix: In the hierarchical analysis method, the judgment matrix is ​​a matrix used to express the relative importance of each factor in the same level relative to a factor in the previous level. By comparing the factors pairwise, their relative importance in the process of achieving the goal is determined.

[0111] The construction of the judgment matrix at the criterion level generally adopts the historical data reference method. The specific process is as follows: 1. Collect historical risk case data: Collect risk cases of key equipment in previous urban rail engineering construction, including accident records caused by equipment failure, equipment problems caused by environmental factors, and losses caused by improper equipment operation and maintenance. 2. Analyze the impact of each criterion in the case: Statistically calculate the role of the core criterion of failure, equipment operation and maintenance criterion, environmental impact criterion, and equipment structure and system criterion in each case. For example, through analysis, it is found that in 50% of equipment failure cases, the core criterion of failure plays a leading role; in 30% of cases, the equipment operation and maintenance criterion is the main factor. 3. Construct a judgment matrix: According to the frequency and importance of each criterion in historical cases, a judgment matrix is ​​constructed according to certain quantitative rules. If the impact of the core criterion of failure is significantly higher than that of the equipment operation and maintenance criterion, a corresponding larger value (such as 5 or 7, etc.) can be assigned in the judgment matrix according to the difference in the impact degree.

[0112] Because the factor layer is subdivided based on the criterion layer, a judgment matrix should be constructed for each factor under each criterion. For example, factors under the core criterion of faults, such as fault type and fault severity, are compared in pairs; factors under the criterion of equipment operation and maintenance, such as equipment operation time and maintenance frequency, are compared in pairs.

[0113] Expert evaluation can also be used to let experts compare and score factors under the same criteria. For example, under the equipment operation and maintenance criteria, for the two factors of equipment operation time and maintenance frequency, experts judge whether the equipment operation time is "more important" or "less important" relative to the maintenance frequency based on the actual operation status of the equipment and the impact of risks, and then quantify it. At the same time, adjust the expert scores based on data and experience from actual projects. If the actual data shows that the risk of failure increases sharply after the equipment operation time exceeds a certain limit, while the impact of maintenance frequency on the risk is relatively slow, then when constructing the judgment matrix, the importance of equipment operation time will be highlighted accordingly.

[0114] Step S530, according to the judgment matrix of the criterion layer and the judgment matrix of the factor layer, the weights of different factors are calculated by using the corresponding weight vector calculation method.

[0115] Among them, weight vector: In the hierarchical analysis method, the weight vector is a vector that represents the relative importance of each factor in the entire evaluation system. It is obtained by calculating the judgment matrix, and its element value reflects the weight of the corresponding factor to the target.

[0116] The criterion layer weight vector calculation method can use the eigenvalue method, as follows: 1. Calculate the maximum eigenvalue of the judgment matrix and its corresponding eigenvector: For the criterion layer judgment matrix, use mathematical calculation methods (such as numerical calculation methods such as power method) to find its maximum eigenvalue and corresponding eigenvector. For example, for a 4×4 criterion layer judgment matrix (corresponding to the core fault criterion, equipment operation and maintenance criterion, environmental impact criterion, equipment structure and system criterion), the maximum eigenvalue (assuming it is λ_max) and a 4-dimensional eigenvector are obtained by calculation. 2. Normalize the eigenvector to obtain the weight vector: Normalize the obtained eigenvector so that the sum of its elements is 1. Assuming that the obtained eigenvector is [V1, V2, V3, V4], the elements of the weight vector are calculated as: W_i=V_i / (V1+V2+V3+V4), where i=1, 2, 3, 4. The [W1, W2, W3, W4] obtained in this way is the weight vector of each factor (i.e., each criterion) in the criterion layer, and each element represents the weight of the corresponding criterion in the risk assessment of key equipment.

[0117] The calculation method of the factor layer weight vector (calculated separately for each factor layer under each criterion) is similar to the calculation process of the criterion layer, as follows: Taking the factors under the fault core criterion as an example, the eigenvalue method or the sum-product method can also be used to calculate the weight vector for its judgment matrix. Assuming that there are factors such as fault category, fault severity, and fault frequency under the fault core criterion, after constructing its judgment matrix, its weight vector is calculated by the above calculation method. If the eigenvalue method is used, first calculate the maximum eigenvalue and corresponding eigenvector of the judgment matrix, and then normalize it to obtain the weight vector. This weight vector represents the relative importance of factors such as fault category, fault severity, and fault frequency under the fault core criterion.

[0118] Among them, one way to calculate the comprehensive risk value of key equipment using the preset comprehensive risk value calculation method is as follows: substitute the obtained single factor risk value and the weights of different factors into the formula of the preset comprehensive risk value calculation method to calculate the comprehensive risk value, wherein the formula of the preset comprehensive risk value calculation method is ,in, is the comprehensive risk value, is the total number of risk factors, Indicates A single factor risk value, Indicates The weight of each risk factor.

[0119] Among them, another way to calculate the comprehensive risk value of key equipment using the preset comprehensive risk value calculation method is as follows: Step S610: Analyze and determine the fault trees corresponding to the key equipment of different categories according to the corresponding relationship between the key equipment categories and the fault trees.

[0120] Fault tree: Fault tree is a graphical tool used to analyze the causes of system failures and their interrelationships. It takes the undesirable events (top events) of the system as the starting point, and connects the various direct and indirect causes (bottom events) that may cause the top events through logic gates (such as AND gates, OR gates, etc.), forming an inverted tree structure, which clearly shows the various possible ways and causal relationships of the failure. In this scenario, the fault trees constructed for different categories of key equipment can help us gain a deeper understanding of the various situations in which the equipment may fail and the relationship between related factors.

[0121] Determining the correspondence between key equipment categories and fault trees can be based on equipment types and functional characteristics. The specific process is as follows: First, determine the corresponding fault tree based on the specific type of key equipment (such as shield machines, cranes, concrete mixers, etc.) and their unique functional characteristics. Different types of equipment face different failure conditions and risk factors due to differences in their structure, working principle, and operating environment. For example, shield machines are mainly used to dig tunnels, and their key components include cutter heads, propulsion systems, hydraulic systems, etc. Then the fault tree constructed for shield machines will revolve around the possible failures of these key components, such as wear of the cutter head, failure of the propulsion system, and leakage of the hydraulic system as bottom events that may cause shield machine failures (top events), and use logic gates to represent the relationship between them.

[0122] Step S620: using fault trees corresponding to different types of key equipment to analyze interaction paths between different risk factors.

[0123] Interaction paths between risk factors: refers to the specific ways in which various risk factors (such as equipment failure, environmental impact, operational errors, etc.) interact and relate to each other during the operation of key equipment, thus causing changes in the overall risk status of the equipment. By analyzing these paths through the fault tree, we can clearly understand how different risk factors work together to affect the normal operation of the equipment.

[0124] The analysis method of the interaction path between different risk factors can be adopted from the top event of the fault tree to the bottom analysis method. The specific process is as follows: 1. Determine the starting point of the risk factor corresponding to the top event: Take the top event of the fault tree corresponding to the key equipment (that is, the equipment has a serious failure or cannot operate normally and other undesirable situations) as the starting point. For example, for the fault tree of the shield machine, the top event may be "shield machine excavation interruption", which is the starting risk condition for our analysis of the interaction path. 2. Trace the path along the logic gate: Starting from the top event, according to the logic gates set in the fault tree (such as AND gates, OR gates, etc.), gradually trace the connection relationship between each risk factor downward. If it is an AND gate, it means that only when all the bottom events connected to it (that is, specific risk factors) occur at the same time, will the upper layer event occur; if it is an OR gate, as long as one of the bottom events occurs, it will cause the upper layer event to occur. For example, in the fault tree of a shield machine, the top event "shield machine excavation interruption" may be connected to "cutter head failure" and "propulsion system failure" through an AND gate, which means that only when the cutter head failure and propulsion system failure occur at the same time will the shield machine excavation be interrupted. This is a clear risk factor interaction path. 3. Record complete path information: In the process of tracing along the logic gate, record in detail each risk factor passed and the connection logic relationship between them to form a complete interaction path description. For example, for the fault tree of a crane, there may be such a path: the top event "crane cannot lift normally" is connected to "wire rope breakage" or "motor failure" through an OR gate, and then "wire rope breakage" is connected to "wire rope excessive wear" and "wire rope uneven force" through an AND gate. In this way, the relationship from the top event to the specific risk factors and the progressive relationship between them, that is, the interaction path, is fully recorded.

[0125] The analysis method of the interaction path between different risk factors can be adopted to analyze the association of risk factors at different levels. The specific process is as follows: Identify risk factors at different levels: In the fault tree, risk factors are distributed at different levels, from the macro level near the top event to the micro level near the bottom event. For example, in the fault tree of a concrete mixer, the top event "concrete mixer cannot mix properly" is at a higher level, while the "mixing blade wear" and "motor overload" connected to it belong to the middle level, and the "mixing blade material is not good" and "motor heat dissipation is poor" are lower-level risk factors.

[0126] Determine cross-level interaction paths: Analyze how risk factors at different levels influence each other. For example, starting from the top event of "concrete mixer cannot mix properly" at a higher level, it may pass through the middle level "mixing blade wear", and "mixing blade wear" has a causal relationship with the lower level "poor mixing blade material", thus forming a cross-level interaction path: top event → middle level risk factor → lower level risk factor. Through such analysis, we can fully understand the complex relationship between risk factors at different levels and the path by which they work together to cause equipment failure.

[0127] Step S630: analyzing and determining the influence coefficients between different risk factors according to the corresponding relationship between the interaction paths and the influence coefficients of the interaction between different risk factors.

[0128] Impact coefficient: A value used to quantify the degree of influence of the interaction between different risk factors on the overall risk status of the equipment. It reflects the weight relationship of a certain risk factor's contribution to the final comprehensive risk value when it interacts with other risk factors.

[0129] The process of analyzing and determining the influence coefficients between different risk factors is as follows: Collect relevant historical data: Comprehensively collect all kinds of historical failure data involving key equipment, including detailed records of equipment failures in previous engineering construction. These records should cover information such as the time and location of the failure, the risk factors involved, the severity of the failure, and the losses caused (such as maintenance costs, construction delays, etc.). For example, for a certain type of shield machine, records of multiple failures in the past were collected, one of which involved risk factors such as "cutter head failure" and "poor geological conditions", and the specific circumstances such as how much the maintenance cost increased and how long the construction delay was caused by the failure were recorded.

[0130] Analyze the co-occurrence and impact of risk factors: From the collected historical data, sort out the situations where different risk factors appear at the same time (co-occurrence), and analyze the impact on the equipment in each co-occurrence. The impact can be measured by some quantitative indicators, such as the proportion of economic losses caused by failures to the total value of the equipment, the proportion of construction delays to the total construction period, etc. For example, statistics show that "cutter head failure" and "poor geological conditions" co-occur many times, and each co-occurrence will lead to a significant increase in maintenance costs and serious construction delays. By calculating the proportion of increased maintenance costs to the total value of the equipment and the proportion of construction delays to the total construction period, it is determined that the impact on the equipment in this co-occurrence situation is high.

[0131] Adjust the impact coefficient based on historical data: Based on the above analysis of the co-occurrence and impact of risk factors, adjust the impact coefficient initially determined based on the fault tree structure. If it is found that the co-occurrence frequency of two risk factors is high and the impact caused by each co-occurrence is large, then the impact coefficient of their interaction should be appropriately increased; conversely, if the co-occurrence frequency is low and the impact is small, then the impact coefficient should be appropriately reduced. For example, the impact coefficient of the interaction between "cutter head failure" and "poor geological conditions" was initially determined to be 0.3 based on the fault tree structure, but through the analysis of historical data, it was found that they had a high co-occurrence frequency and each co-occurrence caused serious losses. In this case, the impact coefficient can be adjusted to 0.4 or a higher value to more accurately reflect their actual impact on equipment risks.

[0132] Step S640, substituting the obtained single factor risk value, the weights of different factors and the influence coefficients between different risk factors into the formula of the preset comprehensive risk value calculation method to calculate the comprehensive risk value.

[0133] The formula for calculating the preset comprehensive risk value is as follows: ; in, is the comprehensive risk value, is the total number of risk factors, Indicates A single factor risk value, Indicates A single factor risk value, Indicates The weight of the risk factors, Is a factor and factors The influence coefficient between them.

[0134] According to the corresponding relationship between the categories of key equipment and the risk value range that the comprehensive risk value falls into and the early warning treatment plan, the early warning treatment plan is analyzed and determined to include: Step S710: analyzing the change rate of the comprehensive risk value within a preset time window according to the obtained comprehensive risk value.

[0135] Specifically, the preset time window is a time period set for analyzing the trend of changes in the comprehensive risk value. For example, you can choose the past week, month, or other appropriate time lengths.

[0136] The process of analyzing the comprehensive risk value change rate is as follows: 1. Determine the start time and end time. According to the length of the preset time window, determine the start time and end time of the analysis. For example, if the preset time window is one week, then the start time is the current time minus one week, and the end time is the current time. 2. Obtain comprehensive risk value data: Obtain the comprehensive risk value data of key equipment within the start time and end time range. This data can be obtained from previous risk assessment records or through a real-time monitoring system. 3. Calculate the change rate. The change rate of the comprehensive risk value can be calculated using the following formula: Change rate = (comprehensive risk value at the end time - comprehensive risk value at the start time) / comprehensive risk value at the start time. For example, if the comprehensive risk value at the start time is 50 and the comprehensive risk value at the end time is 60, then the change rate is 20%.

[0137] Step S720, analyzing whether the change rate of the comprehensive risk value within the preset time window exceeds the preset rising trend threshold. If not, execute step S730; if yes, execute step S740.

[0138] Specifically, the rising trend threshold is a pre-set value used to determine whether the rate of change of the comprehensive risk value indicates that the risk is on an upward trend. The rising trend threshold is usually determined based on experience, historical data, or expert judgment. It can be a fixed value or a range. For example, according to historical data, when the rate of change of the comprehensive risk value exceeds 10%, it usually means that the risk is on an upward trend, so the rising trend threshold can be set to 10%.

[0139] Step S730, maintaining the original settings.

[0140] Step S740, analyzing and determining the early warning processing scheme according to the result information that the comprehensive risk value is in an upward trend, the category of key equipment, the risk value interval range into which the comprehensive risk value falls, and the corresponding relationship between the early warning processing scheme.

[0141] Information about the result of the comprehensive risk value being in an upward trend: This refers to the information obtained by analyzing the change rate of the comprehensive risk value within the preset time window and determining that it exceeds the preset upward trend threshold. This indicates that the risk of key equipment is increasing and requires further attention and processing.

[0142] The risk value range that the comprehensive risk value falls into: The comprehensive risk value is divided into different intervals, each of which represents a different risk level. For example, it can be divided into a low risk interval (0-30), a medium risk interval (31-60), a high risk interval (61-100), etc. Different intervals correspond to different early warning processing plans.

[0143] Early warning treatment plan: Pre-planned response strategies and measures for different risk conditions of key equipment, aiming to reduce equipment risks, avoid failures or reduce losses caused by failures. It is determined based on factors such as equipment category, comprehensive risk value and its changing trend.

[0144] The analysis and determination of the early warning treatment plan is as follows: query the corresponding relationship table (or storage form such as database) based on the pre-established result information of the comprehensive risk value being in an upward trend, the category of key equipment and the risk value range in which the comprehensive risk value falls and the early warning treatment plan. For example, for shield machines, when the comprehensive risk value is in the medium risk range and is on an upward trend, the early warning treatment plan may include increasing the monitoring frequency of key parts such as the cutter head to once an hour, arranging professionals to conduct detailed inspections of the cutter head and propulsion system, preparing spare tools, etc. For cranes, when the comprehensive risk value is in the medium risk range and is on an upward trend, the early warning treatment plan may include shortening the inspection cycle of the brake system, performing flaw detection on the wire rope, limiting the lifting weight, etc.

[0145] Furthermore, considering that there may be multiple preset time windows with different lengths, according to the corresponding relationship between the risk value interval range into which the comprehensive risk value falls and the early warning processing scheme, the early warning processing scheme is analyzed and determined to include: Step S7a0: analyzing the change rate of the comprehensive risk value within preset time windows of different lengths according to the obtained comprehensive risk value.

[0146] The time window is determined as follows: according to the characteristics of key equipment and the expected time scale of risk changes, multiple time windows of different lengths are preset. For example, for some small key equipment with fast operating status changes (such as electric hoists), shorter time windows can be set, such as 1 hour, 3 hours and 6 hours; for large and complex key equipment (such as shield machines), because their operating parameters are relatively stable and the fault development cycle may be long, longer time windows can be set, such as 12 hours, 1 day and 3 days. The length of these time windows should cover the range of equipment changes from short-term fluctuations to long-term trend changes, so as to fully capture the changes in risk values. At the same time, the setting of time windows should be reasonable and operational to avoid excessive data noise due to too short windows or loss of sensitivity to short-term sudden risks due to too long windows.

[0147] The change rate calculation process for different time windows is as follows: traverse the comprehensive risk value data sequence in chronological order, and calculate the corresponding change rate for each preset time window length. Take a case where 1 hour, 3 hours and 6 hours time windows are set at the same time as an example: for the 1 hour time window, starting from the first data point of the data sequence, select two data points (starting point and ending point) every 1 hour, and calculate the change rate between them. Similarly, calculate the change rate of all 1 hour time windows in the entire data sequence. For the 3 hour time window, starting from the first data point of the data sequence, select two data points every 3 hours to calculate the change rate, and complete the change rate calculation for all 3 hour time windows. For the 6 hour time window, the above method is also used for calculation to obtain the corresponding change rate sequence.

[0148] Step S7b0, analyzing whether the change rates of the comprehensive risk value in time windows of different lengths exceed the rising trend threshold of the corresponding preset time windows of different lengths.

[0149] Among them, for time windows of different lengths, the rising trend threshold is determined based on the category of key equipment, operating characteristics, and past engineering experience. For example, for large-scale shield machines with relatively stable operation, the rising trend threshold of the comprehensive risk value change rate may be set to a relatively low value, such as 10%-15%, within a longer time window (such as 1 day or 3 days). This is because the complex structure and high cost of the shield machine determine its high sensitivity to risk changes, and even a small increase in long-term risk needs attention.

[0150] For some relatively small, simple-structured and relatively easy-to-recover critical equipment (such as small cranes), the rising trend threshold in a longer time window can be appropriately increased, such as 20%-30%. For a shorter time window (such as 1 hour or 3 hours), regardless of the type of equipment, the rising trend threshold is usually set lower to quickly capture the sharp changes in risks in the short term, and can generally be set at around 5%-10%.

[0151] The process of comparing the change rate and threshold of different time windows one by one is as follows: after calculating the change rate of the comprehensive risk value in each time window of different lengths, compare the change rate of each time window with the corresponding preset rising trend threshold one by one. For example, for a monitoring system with 1 hour, 3 hours and 6 hours time windows, compare the change rate of the 1 hour time window with the threshold of the 1 hour time window, the change rate of the 3 hour time window with the threshold of the 3 hour time window, and the change rate of the 6 hour time window with the threshold of the 6 hour time window. This comparison is performed independently, and the analysis results of each time window will serve as one of the bases for subsequent warning processing decisions.

[0152] The process of recording the comparison results is as follows: During the comparison process, it is necessary to record whether the change rate in each time window exceeds the threshold. This can be expressed as a Boolean value (such as "yes" or "no"), or a digital code (such as 0 for not exceeding, 1 for exceeding). For example, if the change rate in a 1-hour time window exceeds the threshold, it is recorded as "yes", and if the change rate in a 3-hour time window does not exceed the threshold, it is recorded as "no", and so on. These records will be used in subsequent steps to comprehensively judge and determine the early warning processing plan.

[0153] Step S7c0: if no, maintain the original setting.

[0154] Step S7d0, if yes, then analyze and determine the early warning processing plan based on the correspondence between the time window exceeding the corresponding preset length time window's rising trend threshold, the category of key equipment, the risk value interval range in which the comprehensive risk value falls, and the early warning processing plan.

[0155] Specifically, step S7d0 involves integrating relevant information, as follows: 1. Summarize time window data that exceeds the threshold: Collect all time window information where the rate of change of the comprehensive risk value exceeds the corresponding preset upward trend threshold. This information includes the length of the time window (such as 1 hour, 3 hours, etc.) and the change of the risk value within the time window (such as the size of the rate of change, the characteristics of the change trend, etc.). 2. Associate key equipment category information: Clarify what kind of key equipment it is, such as a shield machine, a crane, or other specific types of equipment. Different categories of equipment have different structures, functions, and operating characteristics, which determine the source and manifestation of the risk. 3. Determine the comprehensive risk value interval: Determine the interval range of the comprehensive risk value of the current key equipment (such as a low risk interval, a medium risk interval, or a high risk interval). This information helps to assess the severity of the risk.

[0156] Risk analysis in the time window dimension: For each time window that exceeds the threshold, analyze the speed and pattern of risk change. If the risk value rises sharply in a short time window, it may mean that the equipment is facing sudden and serious problems; while a slow rise in a longer time window may imply long-term accumulated potential risks. For example, if the comprehensive risk value change rate reaches 20% in a 1-hour time window, this may be a very urgent signal, indicating that the equipment may have a serious failure or be affected by external sudden factors; and if the risk value rises at a rate of 3% per day in a 3-day time window, it may be the result of long-term wear and tear of the equipment or the gradual influence of environmental factors.

[0157] Analyze risk sources in combination with key equipment characteristics: Analyze potential factors that may lead to increased risks based on the categories of key equipment. Taking the shield machine as an example, if the comprehensive risk value increases, possible risk sources include cutter head tool wear, abnormal propulsion system pressure, sealing system leakage, etc.; for cranes, it may be wire rope wear, brake system failure, electrical control system abnormality, etc. By combining the risk changes within the time window, it can be further determined which risk sources are more critical in the current situation. For example, if the risk of the shield machine increases in a short time window and is accompanied by abnormal vibration, it may be that the cutter head tool is suddenly damaged; if the risk of the crane increases in a long time window and is accompanied by a longer braking distance, it may be a chronic failure of the braking system.

[0158] Determining the early warning treatment plan involves the following: 1. Querying the preset correspondence: Based on the pre-established correspondence table or rule base between the key equipment category, the comprehensive risk value range, and the time window exceeding the threshold and the early warning treatment plan, find the matching early warning treatment plan. These correspondences are constructed based on a large amount of historical data, engineering experience, simulation experiments, and expert knowledge. 2. Formulate targeted measures: Formulate a specific early warning treatment plan based on the query results. The plan should cover a series of measures from equipment inspection and maintenance to possible operational adjustments or emergency responses. These measures should take into account multiple factors such as the urgency of the risk, the importance of the equipment, and the impact on the construction progress.

[0159] Suppose we are monitoring and providing early warning for tunnel boring machines in urban rail construction.

[0160] Case 1: The comprehensive risk value is in the medium risk range (assuming it is 40-60), and the comprehensive risk value change rate exceeds the respective rising trend thresholds within the 1-hour and 3-hour time windows, with a 15% change rate in 1 hour and a 10% change rate in 3 hours. According to the corresponding relationship, the early warning processing plan may include: 1. Immediately suspend the shield machine's propulsion operation and check the wear of the cutterhead tool, especially to check whether there are signs of tool chipping or severe wear. 2. Check the shield machine's propulsion system parameters, including hydraulic system pressure, jack stroke, etc., to see if there are abnormal pressure fluctuations or stroke deviations. 3. Perform a quick check of the sealing system to see if there is any leakage of sealing grease, especially at the shield tail seal. 4. Increase the monitoring frequency of key parts (such as the temperature and vibration of the cutterhead drive motor) from once every 30 minutes to once every 10 minutes, and arrange for technicians to analyze the monitoring data in real time.

[0161] Case 2: The comprehensive risk value is in the high-risk range (assuming it is 60-80), and the comprehensive risk value change rate exceeds the threshold within the 6-hour time window, with a change rate of 8%. The early warning processing plan in this case may be: 1. Stop the shield machine and evacuate the relevant personnel on the construction site to a safe area. 2. Organize a professional maintenance team to conduct a comprehensive inspection of the shield machine, including detailed inspection and fault diagnosis of all key subsystems such as the cutter head, propulsion system, segment assembly machine, and electrical system. 3. Check the geological conditions and environmental changes around the construction area to determine whether external factors (such as groundwater changes and stratum collapse) have caused the risk to increase. If necessary, take corresponding geological reinforcement measures. 4. Develop a detailed maintenance and rectification plan based on the inspection results. Only after ensuring that the risks are effectively controlled can the shield machine be restarted and its operating status be closely monitored.

[0162] Based on the same inventive concept, an embodiment of the present invention provides a monitoring and early warning system for urban rail engineering construction, including a memory and a processor, wherein the memory stores a program that can be executed on the processor to implement the following Figures 1 to 4 Procedure for either method.

[0163] The embodiments of this specific implementation method are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Therefore, all equivalent changes made based on the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A monitoring and early warning method for urban rail engineering construction, characterized in that: include: Collect and obtain the operating data of various key equipment at the construction site; According to the mapping relationship between the categories of key equipment and the key features related to fault diagnosis, analyze and determine the key features related to fault diagnosis of various key equipment, and extract the key features related to fault diagnosis from the collected operation data of various key equipment; Use a preset specific equipment fault diagnosis model that matches the key equipment category to analyze and output the fault category distribution probability and fault severity that matches the key equipment category; Obtain the remaining factor information of different key equipment, and construct a risk assessment factor set based on the failure category distribution probability and failure severity of different key equipment and the remaining factor information; For the risk assessment factor set, the preset analytic hierarchy process is used to obtain the weights of different factors; According to the preset risk value assessment method for different factors, the risk value of each single factor is calculated one by one, and according to the risk value of each single factor and the weights of different factors, the preset comprehensive risk value calculation method is used to calculate the comprehensive risk value of key equipment; According to the correspondence between the category of key equipment and the risk value range that the comprehensive risk value falls into and the early warning processing plan, the early warning processing plan is analyzed and determined, and the early warning processing plan is implemented.

2. A monitoring and early warning method for urban rail engineering construction according to claim 1, characterized in that: The analysis outputs the fault category distribution probability and fault severity matching the key equipment categories, including: Use a preset specific equipment fault diagnosis model that matches the key equipment category to analyze and output the fault category distribution probability and fault severity that matches the key equipment category. Simultaneously, extract common features from the operating data of various types of key equipment, use a common fault diagnosis model, and analyze and output the fault category distribution probability and fault severity that matches the various types of key equipment. The voting method is used to count the votes predicted by the specific equipment fault diagnosis model and the general fault diagnosis model for each fault category, and the fault category with the most votes is selected as the most likely fault type, and its relative probability is calculated; The results of the specific equipment fault diagnosis model and the general fault diagnosis model are integrated according to the probabilities of different models to construct a comprehensive fault diagnosis model. The comprehensive fault diagnosis model is used to analyze and output the fault category distribution probability and fault severity that match the key equipment category.

3. A monitoring and early warning method for urban rail engineering construction according to claim 2, characterized in that: Obtain the remaining factor information of different key equipment, and construct a risk assessment factor set based on the failure category distribution probability and failure severity of different key equipment and the remaining factor information, including: According to the correspondence between the preset fault categories, key equipment categories and associated factors, the associated factors associated with the fault categories output by the comprehensive fault diagnosis model are analyzed; Obtain the equipment structure integrity factors and the synergy factors of each subsystem, and integrate the correlation factors, equipment structure integrity factors and the synergy factors of each subsystem to construct a risk assessment factor set.

4. A monitoring and early warning method for urban rail engineering construction according to claim 3, characterized in that: For the risk assessment factor set, the preset hierarchical analysis method is used to obtain the weights of different factors, including: A hierarchical model consisting of a target layer, a criterion layer, and a factor layer is established, wherein the target layer is an accurate and comprehensive risk assessment of key equipment under specific working conditions, the criterion layer includes the core criterion of faults, the equipment operation and maintenance criterion, the environmental impact criterion, and the equipment structure and system criterion, and the factor layer includes factors under the core criterion of faults, factors under the equipment operation and maintenance criterion, factors under the environmental impact criterion, and factors under the equipment structure and system criterion; For the criterion layer and the factor layer, the corresponding judgment matrix construction method is adopted to construct the judgment matrix of the criterion layer and the judgment matrix of the factor layer; According to the judgment matrix of the criterion layer and the judgment matrix of the factor layer, the weights of different factors are calculated by using the corresponding weight vector calculation method.

5. A monitoring and early warning method for urban rail engineering construction according to any one of claims 1 to 4, characterized in that: The preset comprehensive risk value calculation method is used to calculate the comprehensive risk value of key equipment, including: Substitute the obtained single factor risk value and the weights of different factors into the formula of the preset comprehensive risk value calculation method to calculate the comprehensive risk value, where the formula of the preset comprehensive risk value calculation method is: ,in, is the comprehensive risk value, is the total number of risk factors, Indicates A single factor risk value, Indicates The weight of each risk factor.

6. A monitoring and early warning method for urban rail engineering construction according to any one of claims 1 to 4, characterized in that: The preset comprehensive risk value calculation method is used to calculate the comprehensive risk value of key equipment, including: According to the corresponding relationship between the categories of key equipment and fault trees, analyze and determine the fault trees corresponding to different categories of key equipment; Use fault trees corresponding to different categories of critical equipment to analyze the interaction paths between different risk factors; According to the corresponding relationship between the interaction paths and the interaction influence coefficients between different risk factors, the influence coefficients between different risk factors are analyzed and determined; Substitute the obtained single factor risk value, the weights of different factors and the influence coefficients between different risk factors into the formula of the preset comprehensive risk value calculation method to calculate the comprehensive risk value; The formula for calculating the preset comprehensive risk value is as follows: ； in, is the comprehensive risk value, is the total number of risk factors, Indicates A single factor risk value, Indicates A single factor risk value, Indicates The weight of the risk factors, Yes Factor and factors The influence coefficient between them.

7. A monitoring and early warning method for urban rail engineering construction according to claim 1, characterized in that: According to the corresponding relationship between the categories of key equipment and the risk value range that the comprehensive risk value falls into and the early warning treatment plan, the early warning treatment plan is analyzed and determined to include: According to the obtained comprehensive risk value, the change rate of the comprehensive risk value within the preset time window is analyzed and obtained; Analyze whether the change rate of the comprehensive risk value within the preset time window exceeds the preset upward trend threshold; If not, maintain the original setting; If yes, the early warning treatment plan is analyzed and determined based on the result information that the comprehensive risk value is in an upward trend, the category of key equipment, the risk value interval range that the comprehensive risk value falls into, and the correspondence between the early warning treatment plan.

8. A monitoring and early warning system for urban rail engineering construction, characterized in that: It includes a memory, a processor, and a program stored in the memory and executable on the processor. When the program is loaded and executed by the processor, it can implement a monitoring and early warning method for urban rail engineering construction as described in any one of claims 1 to 7.

Images