Cable group safety early warning method based on space-time prediction in stay cable replacement process, medium and equipment

Abstract

The invention discloses a space-time prediction-based cable group safety early warning method in a stay cable replacement process, a medium and equipment in the technical field of bridge operation and maintenance. The space-time prediction-based cable group safety early warning method comprises the following steps: acquiring a current unloading process of a to-be-disassembled stay cable in an overall unloading process and a distance between the to-be-disassembled stay cable and a monitored stay cable in a cable group; inputting into a pre-trained cable force time-space variation prediction model established based on a space-time Gaussian modeling method to obtain a cable group predicted cable force of a next unloading process; calculating a cable group risk state probability and a cable group danger state probability; and determining a safety risk state according to the cable group risk state probability and the cable group dangerous state probability. According to the method, the process data of cable force grading unloading of the disassembled cable is utilized to quickly model and predict the stress of the cable group in the subsequent unloading process, and the unique characteristics that the space-time Gaussian process can learn online and can obtain a probability prediction result are utilized, so that the method is effectively applied to structure safety state early warning and emergency disposal in the cable changing process of the cable system bridge cable changing process.

Description

Technical Field

[0001] This invention relates to a method, medium, and equipment for early warning of cable group safety during the replacement process of cable stays based on spatiotemporal prediction, belonging to the field of bridge operation and maintenance technology. Background Technology

[0002] When the lifespan of the stay cables of a cable-stayed bridge exceeds the minimum design service life, appropriate replacement of the stay cables is necessary to ensure the long-term operational safety of the structure. Before implementing stay cable replacement, it is often necessary to analyze the structural condition to determine whether the structural safety status can support the need for cable replacement. The current common practice is to establish a finite element model of the bridge based on the bridge design data and inspection reports to analyze the structural condition of the cable replacement, confirming that the safety of each component of the bridge structure can meet the specifications even with the missing cable, and then guiding construction based on the design calculation results. However, this method mainly relies on theoretical calculations, and the changes in the structural condition of the bridge structure during the actual cable replacement process usually deviate from the theoretical results, which may lead to safety risks in the actual implementation process in some cases. Construction monitoring controls the structural safety status by monitoring multiple structural responses during the cable replacement process and comparing them with the design theoretical values. When the theoretical model calculation results are inaccurate, it is difficult to directly judge the structural condition through monitoring data. On the other hand, monitoring data has a lag. During construction, the cable force of the cable to be removed is generally unloaded in stages. The structural response can only be obtained after unloading reaches a certain level, thus making it impossible to provide timely warnings and take emergency measures in advance when risks occur. Therefore, there is an urgent need to invent a method that can predict and judge structural safety based on data from the monitoring stage of the cable replacement process.

[0003] Current engineering data prediction and assessment methods often rely on fitting and predicting data from a single sensor, followed by inferring structural status using prediction models that combine data from multiple sensors. These methods employ multinomial prediction, Bayesian machine learning, and alternative modeling to establish mapping models between the forces of the removed cables and those of adjacent cables. When the forces of the removed cables change, the model is trained using already collected data to predict subsequent data development. However, these methods typically suffer from low accuracy due to multi-step predictions and the independence of prediction results between multiple sensors (based on response spatial patterns). Furthermore, the responses of the removed cables and adjacent cables are coupled, and their future data development patterns are not independent. Therefore, building separate prediction models for each sensor while ignoring spatial coupling effects is not only computationally expensive but also results in low prediction accuracy. On the other hand, assessments of structural safety after complete cable unloading based on prediction data still rely on judging the most unfavorable location, lacking a comprehensive assessment method based on the overall structural response of the component group. All these problems make it difficult to timely and accurately assess the structural safety status of bridges during cable replacement, leading to certain safety risks during construction. Therefore, a method that can comprehensively predict and evaluate the overall structure using the monitoring responses of the component group is urgently needed. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide a method, medium and equipment for early warning of cable group safety during the cable replacement process based on spatiotemporal prediction.

[0005] To solve the above-mentioned technical problems, the present invention is implemented using the following technical solution.

[0006] In a first aspect, this invention discloses a method for early warning of cable group safety during the replacement process of stay cables based on spatiotemporal prediction, comprising:

[0007] The current unloading progress of the cable to be dismantled during the overall unloading process is obtained, as well as the distance between the cable to be dismantled and the monitored cable in the cable group; the cable group refers to several cable-stayed cables with monitored cable forces adjacent to the cable to be dismantled.

[0008] The current unloading process of the cable to be dismantled and the distance between the cable to be dismantled and the monitored cable in the cable group are input into a pre-trained cable force time-space change prediction model based on the spatiotemporal Gaussian modeling method to obtain the predicted cable force of the cable group in the next unloading process.

[0009] The probability of a cable group risk state and the probability of a cable group danger state are calculated based on the predicted cable force of the cable group in the next unloading process and the total cable bearing capacity of each cable monitored in the cable group.

[0010] The safety risk state is determined based on the probability of the risk state of the cable group and the probability of the dangerous state of the cable group.

[0011] Furthermore, the training process of the cable force time-space variation prediction model established based on the spatiotemporal Gaussian modeling method includes:

[0012] Monitoring data was collected during the unloading process of the first k stay cables. This monitoring data included the distance D between each monitored cable and the cable being dismantled, the unloading progress S, and the cable force F(S,D) of each monitored cable. S , , Let be the distance between the nth monitored cable and the dismantled cable. To uninstall to the m-th process, Let N be the cable force of the nth monitored cable when unloading to the mth process, where n∈[1,N], m∈[1,M], N represents the number of monitored cables other than the dismantled cables, and M represents the total number of unloading stages;

[0013] A training set is constructed based on the monitoring data from the unloading process of the first k cable stays;

[0014] Perform a spacetime Gaussian model, represented as:

[0015] ;

[0016] in, For F(S,D), it represents a potential spatiotemporal process; Obeying a Gaussian process , represented as ; The covariance function describing temporal and spatial correlation is decomposed into , Let the time covariance function be... It is the spatial covariance function; It follows a Gaussian distribution with zero mean; Variables belonging to the same category For any two points in the domain, Variables belonging to the same category Any two points in the domain;

[0017] Define time state vector , The state-space model is then expressed as:

[0018] ;

[0019] ;

[0020] In the formula, Indicates the predicted value; F(t, representing the cable force F of each monitored cable measured at time t) Potential spatiotemporal processes; All are noise variables that follow a Gaussian distribution with a mean of 0; This is the state transition matrix; To select the matrix; For a moment The state vector;

[0021] Based on the training set, the state-space model is trained using the Kalman filter recursive formula, and the state transition matrix is ​​updated. Selection matrix and noise variables The parameters in the table are used to obtain the time at time t. The expression is obtained at time t. The expression is a trained cable force time-space variation prediction model established based on the spatiotemporal Gaussian modeling method.

[0022] Furthermore, the process of determining the cable force F(S,D) of each monitored cable includes:

[0023] Accelerometers were used to obtain acceleration data of the dismantled cable and several adjacent cables;

[0024] The acceleration data were subjected to a short-time Fourier transform to obtain the time variation of the frequencies of several adjacent cables;

[0025] The temporal variation results of the frequencies of the adjacent cables are subjected to synchronous compression transformation to improve the temporal and frequency resolution of the cable frequencies.

[0026] Based on the improved time and frequency resolution, the multiple relationship between the cable frequencies of each time point is determined. The frequency order is identified based on the multiple relationship between the cable frequencies. The cable force of each monitored cable is calculated based on the frequency order.

[0027] Furthermore, the time covariance function adopts an exponential kernel function, expressed as:

[0028] ;

[0029] In the formula, The standard deviation parameter of the kernel function. It is a natural exponential function. For scale parameters;

[0030] The spatial covariance function employs a squared exponential kernel function.

[0031] Furthermore, the formula for calculating the total bearing capacity of the cable is as follows:

[0032] ;

[0033] In the formula, This represents the total bearing capacity of the cable structure. This is the Daniel effect coefficient. p The number of steel wires. For the tensile strength of the steel wire, This is the standard cross-sectional area of ​​a single steel wire.

[0034] Furthermore, the formula for calculating the probability of the risk state of the search group is as follows:

[0035] ;

[0036] In the formula, P 1 represents the probability of a risky state in the search group. P [ a , b This represents the total load-bearing capacity of each cable in the cable group. N r Predicted cable forces for each cable-stayed bridge F ratio N r / F The probability value that belongs to the interval [a, b), where a and b are preset boundary ranges, and a < b. n Let i be the number of cable roots in the cable group, i∈[1, n].

[0037] Furthermore, the formula for calculating the probability of a dangerous state in a rope group is as follows:

[0038] ;

[0039] In the formula, P 2 represents the probability of a dangerous state in a search group. P [ 0 , a This represents the total load-bearing capacity of each cable in the cable group. N r Predicted cable forces for each cable-stayed bridge F ratio N r / F The probability value that belongs to the interval [0, a) is given by max, which means taking the maximum value.

[0040] Further, determining the safety risk state based on the probability of the cable group risk state and the probability of the cable group danger state includes:

[0041] when and If so, the search group as a whole is safe in the corresponding process;

[0042] when and In this case, the entire search group is in a high-risk state during the corresponding process;

[0043] when At that time, there is a high risk of cable breakage in the cable group during the corresponding process;

[0044] P0 is the upper limit of the probability value.

[0045] In a second aspect, the present invention also discloses a computer-readable storage medium storing one or more programs, the one or more programs including instructions that, when executed by a computing device, cause the computing device to perform any of the methods described in the first aspect.

[0046] Thirdly, the present invention also provides a computer device, comprising,

[0047] One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods described in the first aspect.

[0048] The beneficial effects achieved by this invention are as follows:

[0049] In cable-stayed bridge replacement construction, structural monitoring and evaluation often rely too heavily on theoretical calculations, lacking sufficient understanding of the actual structural condition. There is a lack of methods for rapid evaluation and early warning of the overall safety of the bridge cable group during replacement, leading to unknown structural safety risks. To address this, the spatiotemporal Gaussian process-based cable group safety early warning method described in this invention rapidly models and predicts the cable group stress during subsequent unloading processes using process data from the graded unloading of the dismantled cables. Leveraging the unique characteristics of spatiotemporal Gaussian processes—their ability to learn online and yield probabilistic predictions—this method proposes a timely assessment and evaluation of the overall cable group safety, effectively applicable to early warning and emergency response for structural safety conditions during cable-stayed bridge replacement. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating the present invention. Detailed Implementation

[0051] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0052] Example 1, as Figure 1As shown in this embodiment, a method for early warning of cable group safety during the replacement process of stay cables based on spatiotemporal prediction is introduced. The cable to be dismantled is unloaded in multiple unloading processes, from 0% to 100% (complete removal of cable force). During this process, the cable force of several adjacent stay cables (i.e., forming a cable group) is monitored. When unloading begins, the change in the cable force of the cable to be dismantled can reflect the unloading process S=x% (x∈[0,100]). For each unloading process, the cable force value of the cable group can be obtained. Using each unloading process S and the distance D from each cable in the cable group to the cable to be dismantled as independent variables, and the cable force of each cable in the cable group as the dependent variable, a spatiotemporal Gaussian model can be established. Based on this model, under the current unloading process x%, the training data is the cable force data in the process from 0 to x%. This model can predict the cable force of the cable group in the process from (x+1)% to 100%, and thus predict the risk status of the cable body.

[0053] The specific process of this method includes:

[0054] The current unloading progress of the cable to be dismantled during the overall unloading process is obtained, as well as the distance between the cable to be dismantled and the monitored cable in the cable group; the cable group refers to several cable-stayed cables with monitored cable forces adjacent to the cable to be dismantled.

[0055] The current unloading process of the cable to be dismantled and the distance between the cable to be dismantled and the monitored cable in the cable group are input into a pre-trained cable force time-space change prediction model based on the spatiotemporal Gaussian modeling method to obtain the predicted cable force of the cable group in the next unloading process.

[0056] The probability of a cable group risk state and the probability of a cable group danger state are calculated based on the predicted cable force of the cable group in the next unloading process and the total cable bearing capacity of each cable monitored in the cable group.

[0057] The safety risk state is determined based on the probability of the risk state of the cable group and the probability of the dangerous state of the cable group.

[0058] The training process of the cable force time-space variation prediction model established based on the spatiotemporal Gaussian modeling method includes:

[0059] Monitoring data was collected during the k cable unloading processes before unloading. This monitoring data included the distance D between each monitored cable and the cable to be dismantled, the unloading progress S, and the cable force F(S,D) of each monitored cable. S , , Let be the distance between the nth monitored cable and the dismantled cable. To uninstall to the m-th process, Let N be the cable force of the nth monitored cable when unloading to the mth process, where n∈[1,N], m∈[1,M], N represents the number of monitored cables other than the dismantled cables, and M represents the total number of unloading stages;

[0060] A training set was constructed based on the monitoring data of the first k dismantled cables during the unloading process of the cable group;

[0061] Performing spatiotemporal Gaussian modeling includes:

[0062] ;

[0063] Among them, the cable force F(S, D) of each monitored cable is measured. F(S, D) is a potential spacetime process that obeys a Gaussian process. , represented as , The covariance function describing temporal and spatial correlation is decomposed into , Let the time covariance function be... It is the spatial covariance function; It follows a Gaussian distribution with zero mean; Variables belonging to the same category For any two points in the domain, Variables belonging to the same category Any two points in the domain;

[0064] Define time state vector , The state-space model is then expressed as:

[0065] ;

[0066] ;

[0067] In the formula, F(t, representing the cable force F of each monitored cable measured at time t) Potential spatiotemporal processes; All are noise variables that follow a Gaussian distribution with a mean of 0; This is the state transition matrix; To select the matrix; For a moment The state vector;

[0068] Based on the acquired training data, the state-space model can be trained and the state transition matrix updated using the Kalman filter recursive formula. Selection matrix and noise variables The parameters in the table are used to obtain the time at time t. The expression is the trained cable force time-space variation prediction model established based on the spatiotemporal Gaussian modeling method.

[0069] The process of determining the cable force F(S,D) of each monitored cable includes:

[0070] Accelerometers were used to obtain acceleration data of the dismantled cable and several adjacent cables;

[0071] The acceleration data were subjected to a short-time Fourier transform to obtain the time variation of the frequencies of several adjacent cables;

[0072] The temporal variation results of the frequencies of the adjacent cables are subjected to synchronous compression transformation to improve the temporal and frequency resolution of the cable frequencies.

[0073] Based on the improved time and frequency resolution, the multiple relationship between the cable frequencies of each time point is determined. The frequency order is identified based on the multiple relationship between the cable frequencies. The cable force of each monitored cable is calculated based on the frequency order.

[0074] The time covariance function adopts an exponential kernel function, expressed as:

[0075] ;

[0076] In the formula, The standard deviation parameter of the kernel function. It is a natural exponential function. For scale parameters;

[0077] The spatial covariance function employs a squared exponential kernel function.

[0078] The formula for calculating the total bearing capacity of the cable is:

[0079] ;

[0080] In the formula, This represents the total bearing capacity of the cable structure. This is the Daniel effect coefficient. For the tensile strength of the steel wire, This is the standard cross-sectional area of ​​a single steel wire.

[0081] The formula for calculating the probability of the risk state of the search group is:

[0082] ;

[0083] In the formula, Let the probability of the risk state of the search group be denoted as . P [ a ,b This represents the total load-bearing capacity of each cable in the cable group. N r Predicted cable forces for each cable-stayed bridge F ratio N r / F (Also known as the safety factor of the cable) is the probability value that falls within the interval [a, b), where a and b are preset boundary ranges, and a < b. According to industry knowledge, it is generally considered that... , representing the cable safety factor within this range where the cable is in a risky state, i∈[1, n]. N r It is a constant value, F is a prediction, and it follows a Gaussian distribution. N r / F It itself follows a distribution, and then the probability of belonging to [a, b) is calculated from this distribution.

[0084] The formula for calculating the probability of a dangerous state in a group of ropes is:

[0085] ;

[0086] In the formula, Let be the probability of a dangerous state in a search group. P [0, a This represents the total load-bearing capacity of each cable in the cable group. N r Predicted cable forces for each cable-stayed bridge F ratio N r / F The probability value belonging to the interval [0, a), i∈[1, n], is generally considered to be The value represents the safety factor of the cable, indicating that the cable is considered to be in a dangerous state within this range. N r It is a constant value, F is a prediction, and it follows a Gaussian distribution. N r / F It itself follows a distribution, and then the probability of belonging to [0, a) is calculated from this distribution.

[0087] The step of determining the safety risk state based on the probability of the cable group risk state and the probability of the cable group danger state includes:

[0088] when and If so, the search group as a whole is safe in the corresponding process;

[0089] when and In this case, the entire search group is in a high-risk state during the corresponding process;

[0090] when At that time, there is a high risk of cable breakage in the cable group during the corresponding process.

[0091] Example 2, based on the same inventive concept as Example 1, introduces a method for early warning of cable group safety during cable replacement based on spatiotemporal prediction, including:

[0092] Before performing spatiotemporal modeling of the cable group forces, it is necessary to first identify the real-time changes in cable force. This invention employs a method based on synchronous compression transform and real-time cable frequency order identification to acquire cable force in real time. Specifically, accelerometers are used to acquire acceleration data of the dismantled cable and several adjacent cables; short-time Fourier transform is used to obtain the time variation results of each cable frequency; synchronous compression transform is used to improve the time and frequency resolution of the cable frequencies; finally, the frequency order is identified based on the multiple relationship between cable frequencies, and the real-time cable force is calculated. Thus, continuous change data of cable group forces during the unloading process can be collected.

[0093] Secondly, the distance between each monitored cable and the cable being dismantled is used as a reference. Spatial independent variable, unloading process progress As the independent variable of time, cable force The basic dataset that constitutes the spatiotemporal Gaussian process is the dependent variable. Here, 'n' represents the number of monitored cables excluding the dismantled cable, and 'm' represents the total number of unloading stages. The distance between each monitored cable and the dismantled cable can be directly measured. The unloading progress is determined by dividing the cable force data at different times of the dismantled cable by the initial cable force. To predict and assess structural safety during the cable replacement and unloading process, it is necessary to use data from before the cables are unloaded to 100% to infer data from subsequent unloading to 100%. Here, we assume that unloading to 100% is used... The previous data was used as the training dataset, that is... We need to predict the spooling data results for the remaining mk uninstallation processes.

[0094] Furthermore, based on the existing training dataset, a spatiotemporal Gaussian modeling method is used to establish the temporal-spatial variation law of cable force, and to predict the cable force of each monitored cable during the subsequent unloading to 100% process. The specific steps of spatiotemporal Gaussian modeling are as follows:

[0095] ;

[0096] Among them, the measured cable force data variables , Its potential spacetime process obeys a Gaussian process. ,Right now , The covariance function, which describes the temporal and spatial correlation, can be decomposed into: Generally, the time covariance function is chosen to be an exponential kernel function. The spatial covariance function is selected based on the data characteristics. The cable force object targeted by this invention has relatively smooth characteristics, so the squared exponential kernel function can be selected. The distribution follows a zero-mean Gaussian distribution, representing spatiotemporal noise. Based on this, a certain moment is defined... state vector Then the state-space model is written as

[0097] ;

[0098] ;

[0099] In the above formula All are noise variables that follow a Gaussian distribution with a mean of 0; This is called the state transition matrix; This is called the selection matrix. Finally, the out-of-training time steps can be obtained using the Kalman filter recursive formula. The spatial data prediction results, i.e. ,here This represents the mean of the predicted results. The variance representing the prediction result, or prediction uncertainty, can be estimated using the maximum likelihood method for each parameter in the state-space model.

[0100] After obtaining the predicted results for each monitored cable during the subsequent unloading process, an early warning is issued using the specifications regarding the cable's load-bearing capacity. The following section uses unloading to... Predicting after training a spatiotemporal Gaussian model using process data Results of cable group forces during the process For example, the total load-bearing capacity of each cable can be calculated based on the number of wires, wire strength, and Daniel effect of the cable body.

[0101] ;

[0102] in Represents the total load-bearing capacity of the cable structure; The value represents the Daniel effect coefficient, which is 0.8 when the number of steel wires exceeds 200, 0.9 when the number of wires is between 100 and 200, and 1.0 when the number of wires is less than 100. This refers to the number of steel wires in the cable. The tensile strength of the steel wire; This represents the standard cross-sectional area of ​​a single steel wire. The total load-bearing capacity of any monitored cable can be calculated using this formula. It is generally believed that stay cables should maintain a certain safety factor, meaning the ratio of the total load-bearing capacity to the actual force on the cable needs to be kept within a certain limit. At that time, the cable was considered safe; when At that time, it was believed that the cable body posed some risks but was generally safe; when At that time, the cord was considered to be in a high-risk state; when At that time, the cable structure is considered to be in a critical state, with the risk of breakage and failure imminent. Therefore, for monitoring cable groups, a probability of risk state for the cable group is defined. The mean probability of the search group

[0103] ;

[0104] Similarly, define the probability of dangerous states in a search group. The maximum fracture failure probability of the cable group

[0105] ;

[0106] Ultimately, it can be achieved by giving Set an upper limit for probability values To evaluate the safety status of the cable group during the cable replacement process, when and Then it is believed that in the first The overall security of the process cluster; when and Then it is believed that the search group is in the first... The process as a whole is in a high-risk state, and attention needs to be paid to the security status of the cluster after complete uninstallation; when At any time, regardless Is it greater than All of these indicate that there is a high risk of cable breakage in the cable group, and it is necessary to control the cable replacement and unloading process and take timely emergency measures.

[0107] The early warning method in this embodiment takes the change of cable force in the cable group during the unloading process as the object, uses a spatiotemporal Gaussian process to model and predict the cable force, and uses a probabilistic method of cable group safety coefficient to determine the risk and danger state of the cable group under the future development of cable force, so as to achieve timely early warning.

[0108] In this embodiment's early warning method, the cable force of the adjacent cable group to the dismantled cable is obtained by collecting cable acceleration data through an accelerometer and then using a synchronous compression transformation algorithm based on the frequency method. This enables real-time acquisition of cable force changes, providing timely data for model training during construction. The time-frequency transformation algorithm here can also be a multi-step iterative synchronous compression algorithm or other ideal time-frequency transformation methods.

[0109] The cable force prediction method proposed in the early warning method of this embodiment is a spatiotemporal Gaussian modeling method. It uses the cable unloading progress and the distance of a single cable in the cable group from the cable to be dismantled as time and space independent variables, respectively, and the cable force of each cable in the cable group as the dependent variable for modeling. Finally, it can realize the prediction of the cable force of each cable in the subsequent stage of unloading and obtain the probability prediction result.

[0110] The early warning method in this embodiment defines the overall risk state of the cable group and the dangerous state of individual cable groups based on the probability results predicted by the spatiotemporal Gaussian process and the safety factor of the cable bearing capacity. The overall safety of the cable group is determined by designing the upper limit of probability.

[0111] Example 3, based on the same inventive concept as other examples, describes a computer-readable storage medium storing one or more programs, the one or more programs including instructions that, when executed by a computing device, cause the computing device to perform the method described in Example 1.

[0112] Example 4, based on the same inventive concept as other examples, describes a computer device, including,

[0113] One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for performing the method described in Embodiment 2.

[0114] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0115] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0116] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0117] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0118] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for early warning of cable group safety in a stay cable replacement process based on space-time prediction, characterized in that, The method comprises: obtaining the current unloading process of the cable to be replaced in the overall unloading process and the distance between the cable to be replaced and the monitored cables in the cable group; the cable group is a plurality of adjacent monitored cables in the cable group; inputting the current unloading process of the cable to be replaced in the overall unloading process and the distance between the cable to be replaced and the monitored cables in the cable group into a cable force time-space change prediction model established based on a pre-trained space-time Gaussian modeling method to obtain the predicted cable force of the cable group in the next unloading process; calculating the cable group risk state probability and the cable group dangerous state probability based on the predicted cable force of the cable group in the next unloading process and the total cable force of each cable in the monitored cables in the cable group; determining the safety risk state according to the cable group risk state probability and the cable group dangerous state probability.

2. The method of claim 1, wherein, The training process of the cable force time-space change prediction model based on the space-time Gaussian modeling method comprises: The monitoring data in the unloading process of the first k cables are collected, and the monitoring data include the distance D of each monitored cable from the cable to be removed, the unloading process progress S, and the cable force F(S, D) of each monitored cable; D , S , , is the distance of the nth monitored cable from the cable to be removed, is the unloading to the mth process, is the cable force of the nth monitored cable when unloading to the mth process, n∈[1, N], m∈[1, M], N represents the number of monitored cables except the cable to be removed, and M represents the total unloading stage number; constructing a training set according to the monitoring data in the unloading process of the first k cables; performing space-time Gaussian modeling, which is expressed as: ； wherein, is a spatio-temporal process underlying F(S,D); obeys a Gaussian process , denoted as ; is a covariance function describing the temporal and spatial correlation, decomposed as , is a temporal covariance function, is a spatial covariance function; is a 0-mean Gaussian distribution; is any two points in the domain of the same variable is any two points in the domain of the same variable is any two points in the domain of the same variable is any two points in the domain of the same variable Definition of the time instant state vector , The state space model is then represented as: ； ； wherein, represents a predicted value; represents the cable force F(t, ) of each monitored cable measured at time t, which is a potential spatio-temporal process; are noise variables subject to a 0-mean Gaussian distribution; is a state transition matrix; is a selection matrix; is a state vector at time . Based on the training set, the state space model is trained by using Kalman filter recursion formula to update the state transition matrix , selection matrix and noise variable parameters to obtain the expression of t time , and the expression of t time obtained is the time-space variation prediction model of cable force established by the trained space-time Gaussian modeling method.

3. The method of claim 2, wherein, The determination process of the cable force F(S, D) of each monitored cable comprises: obtaining acceleration data of the cable to be replaced and adjacent cables by using an accelerometer; obtaining the time variation results of the frequencies of the adjacent cables by using a short-time Fourier transform method on the acceleration data; improving the time and frequency resolutions of the cable frequencies by using a synchronous compression transform on the time variation results of the frequencies of the adjacent cables; determining the multiple relationship between the cable frequencies at each time based on the improved time and frequency resolutions of the cable frequencies, identifying the frequency order based on the multiple relationship between the cable frequencies, and converting the frequency order to obtain the cable force of each monitored cable.

4. The cable group safety warning method based on space-time prediction in the cable replacement process according to claim 2, wherein The time covariance function employs an exponential kernel function is expressed as: ； wherein is a kernel function standard deviation parameter, is a natural exponential function, is a scale parameter; the spatial covariance function adopts a square exponential kernel function.

5. The method of claim 2, wherein, The calculation formula of the total cable force is: ； wherein is the total load bearing capacity of the cable; is the coefficient of the Daniell effect, p is the number of steel wires, is the tensile strength of the steel wire, is the standard cross-sectional area of a single steel wire.

6. The method of claim 2, wherein, The calculation formula of the cable group risk state probability is: ； wherein P 1 is the cable group risk state probability, P [ a , b ) represents the total carrying capacity of the cable body of each cable in the cable group N r and the predicted cable force of each cable F N r / F is the probability value belonging to the interval [a, b), a and b are preset limit ranges, a < b, n is the number of cables in the cable group, i ∈ [1, n].​ 7. The method of claim 6, wherein, The calculation formula of the cable group dangerous state probability is: ； wherein, P 2 is the probability of the cable system being in a dangerous state, P [ 0 , a ) represents the total cable carrying capacity of each cable in the cable system N r and the predicted cable force of each cable F N r / F is the probability value belonging to the interval [0, a), and max represents taking the maximum value.​ 8. The method of claim 7, wherein, The determination of the safety risk state according to the cable group risk state probability and the cable group dangerous state probability comprises: When And then the group is secure in the corresponding process; When And then the index population as a whole is in a high risk state in the corresponding process; When there is a high risk of dislocation in the corresponding joint. P0 is the upper limit of the probability value.

9. A computer-readable storage medium storing one or more programs, the one or more programs comprising instructions for: The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods of claims 1-8.

10. A computer device, comprising: The method comprises: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for performing any of the methods of claims 1-8.

Images