A bridge condition assessment method and system based on multi-source sensor data

Abstract

The present invention provides a bridge state assessment method and system based on multi-source sensor data, which relates to the technical field of bridge state assessment. First, historical monitoring data is collected at the starting point, center and end point of the bridge, pre-processed, and the average is taken for subsequent assessment. The partial state value of the bridge is updated by Kalman filtering to obtain the acceleration and displacement data after the state update. Subsequently, the data is evaluated for state by an expert scoring method, and a random forest model is constructed for state prediction. The current monitoring data is evaluated by the model. If the current state predicted by the model is lower than the state threshold, the abnormal index of the bridge is calculated and compared with the preset abnormal threshold. If the abnormal index exceeds the standard, an early warning is issued. This method effectively integrates multi-source data, improves the accuracy and real-time performance of bridge monitoring, and provides a scientific basis for the safe operation of the bridge.

Description

Technical Field

[0001] The present invention relates to the technical field of bridge status assessment, and in particular to a bridge status assessment method and system based on multi-source sensor data. Background Art

[0002] As an important infrastructure, bridges bear the heavy responsibility of transportation, and ensuring their safety and reliability is crucial. With the increase in traffic flow and the influence of environmental factors, bridges may suffer varying degrees of damage and aging during use. Bridge condition assessment can timely detect potential structural problems, reduce traffic accidents and casualties caused by bridge failures, and protect the safety of public life and property. Through scientific and reasonable assessment methods, data support can be provided for the maintenance and repair of bridges, helping management departments optimize resource allocation, extend the service life of bridges, and reduce maintenance costs. In addition, with the development of intelligent transportation systems, the real-time and intelligent level of bridge condition assessment has received increasing attention, which not only helps to improve the overall efficiency of the transportation system, but also helps to promote sustainable transportation development.

[0003] Traditional bridge inspection methods mainly rely on manual inspection and regular maintenance, which is not only time-consuming and labor-intensive, but also difficult to fully reflect the real-time status of the bridge. In addition, the existing technology lacks effective comprehensive analysis methods in the assessment of bridge health status, resulting in bridge safety accidents that often occur when some hidden dangers are not discovered in time.

[0004] The prior art has the following deficiencies:

[0005] Traditional bridge condition assessment often relies on manual regular inspection and assessment, which is not only time-consuming and labor-intensive, but also easily affected by the experience and subjective judgment of inspectors, resulting in insufficient accuracy and consistency of assessment results. Most traditional technologies cannot achieve real-time monitoring of bridge conditions, resulting in the inability to timely detect and respond to dynamic changes and sudden damage to bridges, which may cause safety hazards at critical moments. In data collection, there may be a single monitoring point that cannot fully reflect the overall condition of the bridge. In addition, traditional assessment methods often process different types of data in isolation and lack comprehensive analysis of multi-source information, resulting in the inability to fully assess the health status and operating performance of bridges. They lack intelligence and automation, and cannot effectively use advanced technologies such as big data and machine learning to improve the accuracy and efficiency of assessments.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0007] The purpose of the present invention is to provide a bridge status assessment method and system based on multi-source sensor data to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A bridge condition assessment method based on multi-source sensor data, the specific steps include:

[0010] Step 1: Collect historical monitoring data at monitoring points on the bridge and pre-process the historical monitoring data, wherein the monitoring data includes strain, acceleration, displacement, temperature and humidity of the bridge;

[0011] Step 2: Analyze the preprocessed acceleration to obtain the speed data of the bridge, combine the displacement, speed and acceleration data of the bridge as the state vector, use Kalman filtering to update the state value, and determine the historical monitoring data of the bridge;

[0012] Step 3: Use the expert scoring method to evaluate the data collected at each moment in the historical monitoring data and assign the corresponding condition assessment index. Take the strain, acceleration and displacement data in the historical monitoring data of the bridge as input and the corresponding condition assessment index as label to build a condition assessment model based on the random forest algorithm.

[0013] Step 4: Input the strain, acceleration and displacement data of the bridge at the current moment into the trained state assessment model, obtain the predicted state assessment index, and compare it with the preset state threshold to determine whether to execute the next step;

[0014] Step 5: Calculate the abnormal index of the bridge at the current moment and compare it with the preset abnormal threshold, and issue an early warning for data with abnormal status.

[0015] Furthermore, the specific logic for collecting historical monitoring data of bridges and preprocessing them is as follows:

[0016] The starting point, center and end point of the bridge main beam are set as monitoring points, and strain sensors, acceleration sensors, displacement sensors, temperature sensors and humidity sensors are installed at the monitoring points to collect strain, acceleration, displacement, temperature and humidity data at each monitoring point; the preprocessing step includes filtering the collected historical monitoring data with mean filtering and removing abnormal values, calculating the Z-score of each data, and when the absolute value of the Z-score of the data is When the data is missing, it is considered as an outlier and removed, and the missing values ​​are filled by using the mean imputation method.

[0017] Furthermore, the specific logic of Kalman filtering is as follows:

[0018] The formula for obtaining the speed at the bridge monitoring point is:

[0019] ;

[0020] in, for The speed of time, for The speed of time, for The acceleration of time, is the time interval;

[0021] The speed data and the displacement and acceleration data at the corresponding moment in the historical monitoring data are combined to obtain the state vector. The state at the current moment is predicted by Kalman filtering combined with the state at the previous time point. The predicted state is compared with the actual data, and the Kalman gain is calculated. The estimated value is updated by combining the predicted and observed information. The displacement and acceleration data after Kalman filtering replace the original historical monitoring data at the corresponding moment, and the mean of each data in the historical monitoring data of the three monitoring points of the bridge is used as the historical monitoring data of the bridge.

[0022] Furthermore, the specific logic for constructing the status assessment model is as follows:

[0023] The expert scoring method was used to evaluate the historical monitoring data of the bridge. The condition assessment index was a scoring system of 0-100. The historical monitoring data was divided into a training set and a test set at a ratio of 7:3. The strain, acceleration and displacement data in the historical monitoring data of the bridge in the training set were used as input, and the condition assessment index was used as the label. The random forest model in machine learning was selected to construct the condition assessment model, and the strain, acceleration and displacement data in the test set were input to evaluate the performance of the model. The evaluation indicators included accuracy, precision and recall. If all changes in the evaluation indicators were less than 0.01 in five iterations, the model training was considered complete.

[0024] Furthermore, the specific logic for evaluating the current state of the bridge is as follows:

[0025] Collect monitoring data from three monitoring points of the bridge at the current moment, and execute the above steps to obtain the strain, acceleration and displacement data of the bridge at the current moment, input them into the state assessment model, obtain the state assessment index predicted by the model, and compare it with the preset state threshold. If the state assessment index predicted by the model at the current moment is higher than the state threshold, continue to monitor the bridge; if the state assessment index predicted by the model at the current moment is lower than the state threshold, execute the next step to further judge the state of the bridge.

[0026] Furthermore, the specific logic for calculating the abnormal index of the bridge at the current moment is:

[0027] The state index of each monitoring data is obtained, and the formula for calculating the strain state index is as follows:

[0028] ;

[0029] in, represents the strain state index, represents the strain of the bridge at the current moment, Indicates the strain reference value;

[0030] Similarly, the acceleration state index, displacement state index, temperature state index, and humidity state index are respectively ;

[0031] The formula for obtaining the abnormal index of the bridge at the current moment is:

[0032] ;

[0033] in, Indicates the abnormal index of the bridge at the current moment, are the weights of their respective items, ,and ;

[0034] The abnormal index of the bridge at the current moment is compared with the preset abnormal threshold. If the abnormal index at the current moment is lower than the preset abnormal threshold, the bridge will continue to be monitored. If the abnormal index at the current moment is higher than the preset abnormal threshold, an abnormal warning will be issued.

[0035] The present invention further provides a bridge state assessment system based on multi-source sensor data, wherein the bridge state assessment system based on multi-source sensor data is used to implement the above-mentioned bridge state assessment method based on multi-source sensor data, and comprises:

[0036] A data acquisition and preprocessing module is used to acquire historical monitoring data at monitoring points on the bridge and preprocess the historical monitoring data, wherein the monitoring data includes strain, acceleration, displacement, temperature and humidity of the bridge;

[0037] The state value update module is used to analyze the pre-processed acceleration to obtain the speed data of the bridge, combine the displacement, speed and acceleration data of the bridge as the state vector, use Kalman filtering to update the state value, and determine the historical monitoring data of the bridge;

[0038] The condition assessment model building module is used to use the expert scoring method to conduct condition assessment on the data collected at each moment in the historical monitoring data and assign the corresponding condition assessment index. The strain, acceleration and displacement data in the historical monitoring data of the bridge are used as input, and the corresponding condition assessment index is used as a label to build a condition assessment model based on the random forest algorithm.

[0039] The state prediction and judgment module is used to input the strain, acceleration and displacement data of the bridge at the current moment into the trained state assessment model, obtain the predicted state assessment index, and compare it with the preset state threshold to determine whether to execute the next step;

[0040] The abnormal monitoring and early warning module is used to calculate the abnormal index of the bridge at the current moment, compare it with the preset abnormal threshold, and issue an early warning for data with abnormal status.

[0041] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0042] This scheme's bridge condition assessment method based on multi-source sensor data can efficiently and accurately assess the health status of the bridge through real-time monitoring and intelligent analysis, significantly improving the scientificity and timeliness of monitoring. This method combines multiple sensor data and uses Kalman filtering and machine learning algorithms to reduce the limitations of traditional manual inspections, achieve early warning of abnormal conditions, and reduce accident risks. By constructing an abnormal index calculation formula based on multi-parameter monitoring data, the health status of the bridge can be effectively and comprehensively assessed. This scheme not only improves the accuracy of bridge maintenance management, but also provides a reliable basis for decision-making, helps to extend the service life of the bridge, optimize resource allocation, and thus ensure traffic safety and public interests. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0044] Figure 2 Schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0045] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0046] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0047] Example:

[0048] See also Figure 1 , the present invention provides a technical solution:

[0049] A bridge condition assessment method based on multi-source sensor data, the specific steps include:

[0050] Step 1: Collect historical monitoring data at monitoring points on the bridge and pre-process the historical monitoring data, wherein the monitoring data includes strain, acceleration, displacement, temperature and humidity of the bridge;

[0051] In this embodiment, the specific logic for collecting historical monitoring data of the bridge and performing preprocessing is as follows:

[0052] The starting point, center and end point of the bridge main beam are set as monitoring points, and strain sensors, acceleration sensors, displacement sensors, temperature sensors and humidity sensors are installed at the monitoring points to collect strain, acceleration, displacement, temperature and humidity data at each monitoring point; the preprocessing step includes filtering the collected historical monitoring data with mean filtering and removing abnormal values, calculating the Z-score of each data, and when the absolute value of the Z-score of the data is When the data is missing, it is considered as an outlier and removed, and the missing values ​​are filled by using the mean imputation method.

[0053] By installing a variety of sensors (strain, acceleration, displacement, temperature and humidity) at the key monitoring points of the bridge (the starting point, center and end point of the main beam), the multi-dimensional status information of the bridge can be fully collected. This comprehensive data collection helps to more accurately reflect the health status of the bridge and provide sufficient basic data for subsequent evaluation. The preprocessed data can improve the adaptability and accuracy of the subsequent status assessment model and lay a solid foundation for the application of machine learning or other data analysis technologies. Good data quality is the key to building an effective assessment model.

[0054] Step 2: Analyze the preprocessed acceleration to obtain the speed data of the bridge, combine the displacement, speed and acceleration data of the bridge as the state vector, use Kalman filtering to update the state value, and determine the historical monitoring data of the bridge;

[0055] In this embodiment, the specific logic for performing Kalman filtering is:

[0056] The formula for obtaining the speed at the bridge monitoring point is:

[0057] ;

[0058] in, for The speed of time, for The speed of time, for The acceleration of time, is the time interval;

[0059] Combining displacement, velocity and acceleration to establish a state vector is expressed as:

[0060] ;

[0061] in, represents the state vector, represents displacement, Indicates speed, represents acceleration;

[0062] The state transfer equation is determined as:

[0063] ;

[0064] in, It is expressed as:

[0065] ;

[0066] in, is the state transfer matrix, express The state vector at time, express The state vector at time, is the process noise, which is set to zero-mean Gaussian white noise;

[0067] The observation model maps the state to the observation space. The model is described as:

[0068] ;

[0069] Among them, the observation matrix for:

[0070] ;

[0071] in, is the observation vector, is the current state vector, is the observation noise, which is set to zero-mean Gaussian white noise;

[0072] The initial state estimate and initial error covariance are set as:

[0073] ;

[0074] in, is the initial state estimate, is the initial error covariance, is the set initial error, is the identity matrix;

[0075] State prediction through state transition model:

[0076] ;

[0077] in, for The predicted state vector at time , based only on The state vector at the moment does not take into account the current observation vector;

[0078] Update state error covariance:

[0079] ;

[0080] in, It is expressed as:

[0081] ;

[0082] in, for The error covariance of the moment prediction is based only on The state vector at the moment does not take into account the current observation vector. for The actual error covariance at time , is the transpose of the state transfer matrix, is the process noise covariance matrix, are the process noise variances of displacement, velocity and acceleration respectively;

[0083] Calculate the Kalman gain:

[0084] ;

[0085] in, It is expressed as:

[0086] ;

[0087] in, is the transpose of the observation matrix, is the observation noise covariance matrix, are the observation noise variances of displacement, velocity and acceleration respectively;

[0088] Get status update formula:

[0089] ;

[0090] in, for The estimated value of the state at time , conditioned on the current observation has been taken into consideration, is the Kalman gain, for The observation vector at time instant;

[0091] The displacement and acceleration data after Kalman filtering replace the historical monitoring data at the original corresponding time, and the mean of each data in the historical monitoring data of the three monitoring points of the bridge is used as the historical monitoring data of the bridge.

[0092] Kalman filtering can be used to fuse data from different sources to obtain a more accurate state estimate, which can help to more comprehensively evaluate the overall state of the bridge. Kalman filtering can effectively reduce the noise and error between different sensor data through weighted updates, providing more accurate state information, which is especially important in dynamic environments. Kalman filtering is a recursive algorithm that can quickly calculate a new state estimate each time the data is updated. This real-time performance is very helpful for timely detection of abnormal bridge conditions. By combining prediction and observation, Kalman filtering provides a powerful framework that allows the system to not only make state estimates based on current observations, but also use predicted information to improve estimates.

[0093] Step 3: Use the expert scoring method to evaluate the data collected at each moment in the historical monitoring data and assign the corresponding condition assessment index. Take the strain, acceleration and displacement data in the historical monitoring data of the bridge as input and the corresponding condition assessment index as label to build a condition assessment model based on the random forest algorithm.

[0094] In this embodiment, the specific logic for constructing the state assessment model is:

[0095] The expert scoring method was used to evaluate the historical monitoring data of the bridge. The condition assessment index was a scoring system of 0-100. The historical monitoring data was divided into a training set and a test set at a ratio of 7:3. The strain, acceleration and displacement data in the historical monitoring data of the bridge in the training set were used as input, and the condition assessment index was used as the label. The random forest model in machine learning was selected to construct the condition assessment model, and the strain, acceleration and displacement data in the test set were input to evaluate the performance of the model. The evaluation indicators included accuracy, precision and recall. If all changes in the evaluation indicators were less than 0.01 in five iterations, the model training was considered complete.

[0096] The expert scoring method is used to evaluate the condition of historical monitoring data. With the help of the knowledge and experience of industry experts, an accurate evaluation index can be given to the bridge condition. This method makes the evaluation results more authoritative and credible, ensuring that the model's labeled data reflects the true health of the bridge. Random forest is selected as the condition assessment model to make full use of its powerful classification and regression capabilities. Random forest can handle high-dimensional data and improve the overall prediction effect through ensemble learning. It has strong robustness and accuracy. Multiple indicators such as accuracy, precision and recall rate are used to evaluate the model performance, making the evaluation results more comprehensive and able to fully reflect the performance of the model in condition assessment. This multi-dimensional evaluation method helps to optimize the parameters and structure of the model.

[0097] Step 4: Input the strain, acceleration and displacement data of the bridge at the current moment into the trained state assessment model, obtain the predicted state assessment index, and compare it with the preset state threshold to determine whether to execute the next step;

[0098] In this embodiment, the specific logic for evaluating the state of the bridge at the current moment is:

[0099] Collect monitoring data from three monitoring points of the bridge at the current moment, and execute the above steps to obtain the strain, acceleration and displacement data of the bridge at the current moment, input them into the state assessment model, obtain the state assessment index predicted by the model, and compare it with the preset state threshold. If the state assessment index predicted by the model at the current moment is higher than the state threshold, continue to monitor the bridge; if the state assessment index predicted by the model at the current moment is lower than the state threshold, execute the next step to further judge the state of the bridge.

[0100] Step 5: Calculate the abnormal index of the bridge at the current moment, compare it with the preset abnormal threshold, and issue an early warning for data with abnormal status;

[0101] In this embodiment, the specific logic for calculating the abnormal index of the bridge at the current moment is:

[0102] The state index of each monitoring data is obtained, and the formula for calculating the strain state index is as follows:

[0103] ;

[0104] in, represents the strain state index, represents the strain of the bridge at the current moment, Indicates the strain reference value;

[0105] Similarly, the acceleration state index, displacement state index, temperature state index, and humidity state index are respectively ; Set the strain reference value to 0.01 and the acceleration reference value to 1.0, and the unit is , the displacement reference value is 3.0, in meters, the temperature reference value is 25, in degrees Celsius, and the humidity reference value is 50%;

[0106] The formula for obtaining the abnormal index of the bridge at the current moment is:

[0107] ;

[0108] in, Indicates the abnormal index of the bridge at the current moment, are the weights of their respective items, ,and ;

[0109] The abnormal index of the bridge at the current moment is compared with the preset abnormal threshold. If the abnormal index at the current moment is lower than the preset abnormal threshold, the bridge will continue to be monitored. If the abnormal index at the current moment is higher than the preset abnormal threshold, an abnormal warning will be issued.

[0110] This formula comprehensively considers the five key monitoring indicators of the bridge: strain, acceleration, displacement, temperature and humidity, reflecting the complexity of the bridge status. Different parameters can comprehensively evaluate the health status of the bridge and avoid evaluation errors caused by relying on only a single indicator. ) increases, the abnormal index The strain state index (SSI) will also increase accordingly, indicating that the health of the bridge is deteriorating and there may be potential safety hazards. ) and acceleration state index ( ) are usually closely related to the structural health of the bridge. Structural health problems may cause a sharp increase in these indicators. Using exponential form can effectively amplify the impact of these indicators, especially when the strain or acceleration reaches a certain threshold, the threat to the safety of the bridge will become more significant. Temperature ( ) and humidity ( ) and other environmental factors on bridges are usually relatively stable, and their changes may not be linear. Using logarithmic form can reduce the sensitivity of these parameters to the influence of abnormal index, especially when the data range is large, which can effectively reduce the impact of extreme values.

[0111] Strain is a direct reflection of the load borne by the structure. Excessive strain may cause material damage or fatigue, which is directly related to the safety of the bridge. Therefore, a higher weight should be given when monitoring strain. Acceleration reflects the response of the bridge under dynamic loads (such as traffic, wind, etc.). Excessive acceleration may indicate that the bridge has structural problems or safety hazards. Although it is not as direct as strain, it is also important, so it can be given a higher weight. Displacement is used to detect the deformation of the bridge under load. Although the change in displacement can reflect the structural condition to a certain extent, its importance is relatively reduced, especially in short-term monitoring. Therefore, it can be given a medium weight. Temperature affects material properties and structural behavior (such as thermal expansion and contraction), but under normal circumstances, temperature changes will not directly cause serious safety problems in the bridge. Although it needs to be monitored, its impact on abnormal conditions is relatively small, so it can be given a lower weight. Humidity mainly affects the durability and long-term performance of materials, such as the durability of concrete. Although excessive humidity may cause problems such as corrosion, it has less impact on short-term safety than other factors, so it is given the lowest weight. This constraint ensures that the anomaly index is a normalized indicator, ensuring that the total contribution of the weighted combination of the five factors is 100%.

[0112] See also Figure 2 The present invention further provides a bridge state assessment system based on multi-source sensor data, wherein the bridge state assessment system based on multi-source sensor data is used to implement the above-mentioned bridge state assessment method based on multi-source sensor data, and comprises:

[0113] A data acquisition and preprocessing module is used to acquire historical monitoring data at monitoring points on the bridge and preprocess the historical monitoring data, wherein the monitoring data includes strain, acceleration, displacement, temperature and humidity of the bridge;

[0114] The state value update module is used to analyze the pre-processed acceleration to obtain the speed data of the bridge, combine the displacement, speed and acceleration data of the bridge as the state vector, use Kalman filtering to update the state value, and determine the historical monitoring data of the bridge;

[0115] The condition assessment model building module is used to use the expert scoring method to conduct condition assessment on the data collected at each moment in the historical monitoring data and assign the corresponding condition assessment index. The strain, acceleration and displacement data in the historical monitoring data of the bridge are used as input, and the corresponding condition assessment index is used as a label to build a condition assessment model based on the random forest algorithm.

[0116] The state prediction and judgment module is used to input the strain, acceleration and displacement data of the bridge at the current moment into the trained state assessment model, obtain the predicted state assessment index, and compare it with the preset state threshold to determine whether to execute the next step;

[0117] The abnormal monitoring and early warning module is used to calculate the abnormal index of the bridge at the current moment, compare it with the preset abnormal threshold, and issue an early warning for data with abnormal status.

[0118] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0119] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0120] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0121] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A bridge condition assessment method based on multi-source sensor data, characterized in that: The specific steps include: Step 1: Collect historical monitoring data at monitoring points on the bridge and pre-process the historical monitoring data, wherein the monitoring data includes strain, acceleration, displacement, temperature and humidity of the bridge; Step 2: Analyze the preprocessed acceleration to obtain the speed data of the bridge, combine the displacement, speed and acceleration data of the bridge as the state vector, use Kalman filtering to update the state value, and determine the historical monitoring data of the bridge; Step 3: Use the expert scoring method to evaluate the data collected at each moment in the historical monitoring data and assign the corresponding condition assessment index. Take the strain, acceleration and displacement data in the historical monitoring data of the bridge as input and the corresponding condition assessment index as label to build a condition assessment model based on the random forest algorithm. Step 4: Input the strain, acceleration and displacement data of the bridge at the current moment into the trained state assessment model, obtain the predicted state assessment index, and compare it with the preset state threshold to determine whether to execute the next step; Step 5: Calculate the abnormal index of the bridge at the current moment, compare it with the preset abnormal threshold, and issue an early warning for data with abnormal status; The state index of each monitoring data is obtained, and the formula for calculating the strain state index is as follows: in, represents the strain state index, represents the strain of the bridge at the current moment, Indicates the strain reference value; Similarly, the acceleration state index, displacement state index, temperature state index, and humidity state index are respectively ; The formula for obtaining the abnormal index of the bridge at the current moment is: in, Indicates the abnormal index of the bridge at the current moment, are the weights of their respective items, ,and ; The abnormal index of the bridge at the current moment is compared with the preset abnormal threshold. If the abnormal index at the current moment is lower than the preset abnormal threshold, the bridge will continue to be monitored. If the abnormal index at the current moment is higher than the preset abnormal threshold, an abnormal warning will be issued.

2. A bridge condition assessment method based on multi-source sensor data according to claim 1, characterized in that: The specific logic for collecting historical monitoring data of bridges and preprocessing them is as follows: The starting point, center and end point of the bridge main beam are set as monitoring points, and strain sensors, acceleration sensors, displacement sensors, temperature sensors and humidity sensors are installed at the monitoring points to collect strain, acceleration, displacement, temperature and humidity data at each monitoring point; the preprocessing step includes filtering the collected historical monitoring data with mean filtering and removing abnormal values, calculating the Z-score of each data, and when the absolute value of the Z-score of the data is When the data is missing, it is considered as an outlier and removed, and the missing values ​​are filled by using the mean imputation method.

3. The bridge condition assessment method based on multi-source sensor data according to claim 2 is characterized in that: The specific logic of Kalman filtering is: The formula for obtaining the speed at the bridge monitoring point is: in, for The speed of time, for The speed of time, for The acceleration of time, is the time interval; The speed data and the displacement and acceleration data at the corresponding moment in the historical monitoring data are combined to obtain the state vector. The state at the current moment is predicted by Kalman filtering combined with the state at the previous time point. The predicted state is compared with the actual data, and the Kalman gain is calculated. The estimated value is updated by combining the predicted and observed information. The displacement and acceleration data after Kalman filtering replace the original historical monitoring data at the corresponding moment, and the mean of each data in the historical monitoring data of the three monitoring points of the bridge is used as the historical monitoring data of the bridge.

4. The bridge condition assessment method based on multi-source sensor data according to claim 3 is characterized in that: The specific logic for building the state assessment model is: The expert scoring method was used to evaluate the historical monitoring data of the bridge. The condition assessment index was a scoring system of 0-100. The historical monitoring data was divided into a training set and a test set at a ratio of 7:

3. The strain, acceleration and displacement data in the historical monitoring data of the bridge in the training set were used as input, and the condition assessment index was used as the label. The random forest model in machine learning was selected to construct the condition assessment model, and the strain, acceleration and displacement data in the test set were input to evaluate the performance of the model. The evaluation indicators included accuracy, precision and recall. If all changes in the evaluation indicators were less than 0.01 in five iterations, the model training was considered complete.

5. The bridge condition assessment method based on multi-source sensor data according to claim 4 is characterized in that: The specific logic for evaluating the current state of the bridge is as follows: Collect monitoring data from three monitoring points of the bridge at the current moment, and execute the above steps to obtain the strain, acceleration and displacement data of the bridge at the current moment, input them into the state assessment model, obtain the state assessment index predicted by the model, and compare it with the preset state threshold. If the state assessment index predicted by the model at the current moment is higher than the state threshold, continue to monitor the bridge; if the state assessment index predicted by the model at the current moment is lower than the state threshold, execute the next step to further judge the state of the bridge.

6. A bridge condition assessment system based on multi-source sensor data, characterized in that: The bridge condition assessment system based on multi-source sensor data is used to implement the bridge condition assessment method based on multi-source sensor data according to any one of claims 1 to 5, comprising: A data acquisition and preprocessing module is used to acquire historical monitoring data at monitoring points on the bridge and preprocess the historical monitoring data, wherein the monitoring data includes strain, acceleration, displacement, temperature and humidity of the bridge; The state value update module is used to analyze the pre-processed acceleration to obtain the speed data of the bridge, combine the displacement, speed and acceleration data of the bridge as the state vector, use Kalman filtering to update the state value, and determine the historical monitoring data of the bridge; The condition assessment model building module is used to use the expert scoring method to conduct condition assessment on the data collected at each moment in the historical monitoring data and assign the corresponding condition assessment index. The strain, acceleration and displacement data in the historical monitoring data of the bridge are used as input, and the corresponding condition assessment index is used as a label to build a condition assessment model based on the random forest algorithm. The state prediction and judgment module is used to input the strain, acceleration and displacement data of the bridge at the current moment into the trained state assessment model, obtain the predicted state assessment index, and compare it with the preset state threshold to determine whether to execute the next step; The abnormal monitoring and early warning module is used to calculate the abnormal index of the bridge at the current moment, compare it with the preset abnormal threshold, and issue an early warning for data with abnormal status.

Images