CN109902344B - Device and system for predicting structural performance of middle-small span bridge group - Google Patents
Device and system for predicting structural performance of middle-small span bridge group Download PDFInfo
- Publication number
- CN109902344B CN109902344B CN201910061475.4A CN201910061475A CN109902344B CN 109902344 B CN109902344 B CN 109902344B CN 201910061475 A CN201910061475 A CN 201910061475A CN 109902344 B CN109902344 B CN 109902344B
- Authority
- CN
- China
- Prior art keywords
- bridge
- neural network
- group
- predicting
- network model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The invention provides a system for predicting structural performance of a bridge group with a medium and small span, which is used for predicting structural performance of a bridge group consisting of a plurality of bridges with medium and small spans, and is characterized by comprising the following steps: a detection report storage unit for storing a year-over detection report for each bridge; a database construction unit for constructing a relational database of the bridge group according to the past year detection report; the model building part is used for building a neural network model according to the relational database, training and checking, wherein the neural network model is used for predicting the structural performance degradation of the bridge group; and a performance prediction unit for predicting the performance trend of the overall structure and the local members of the bridge group by acquiring the performance parameters of each bridge according to the prediction neural network model.
Description
Technical Field
The invention belongs to the field of bridge safety, and particularly relates to a device and a system for predicting structural performance of a group of bridges with medium and small spans.
Background
China is the country with the largest number of highway bridges in the world. According to statistical information of the transportation department, 80.53 ten thousand highway bridges are present in China by the end of 2016, and the accumulated length is 4916.97 ten thousand linear meters. With the increase of service life of bridges, a large number of newly built bridges gradually enter an aging stage, and various structural degradations are inevitable. Therefore, the management maintenance work for pushing various in-service bridges is unprecedented. However, for a specified traffic road network, although a large amount of precious materials containing precious structural information are accumulated in annual structural inspection, corresponding means are always lacking to fully utilize the precious materials, and data disasters are caused. On the other hand, the existing bridge management and maintenance method is only implemented at the level of the single bridge, but not at the road network level, and various commonalities of bridge structures in the same area are ignored, so that the management and maintenance work efficiency is greatly reduced.
Structural performance assessment of small and medium span bridge clusters has been plagued with a number of problems in engineering practice. For example, building an evaluation model of a small and medium span bridge group requires a huge amount of data as support. Therefore, the existing historical bridge detection data and the traffic flow observation records of all road sections are required to be subjected to data mining, interesting and valuable information is extracted, and a relational database is obtained through data integration, data cleaning and data conversion. At the same time, how to simulate the complex nonlinear and logical relations between the degradation trend of the bridge performance and the basic parameters based on the database is a difficult task. The machine learning method based on the neural network provides a practical and effective way, and a reasonable model can serve management and maintenance of all bridges in the traffic road network.
Disclosure of Invention
In order to solve the problems, the future development trend of the existing bridge can be effectively predicted based on the neural network model, and suggestions about maintenance schemes are given, and the invention adopts the following technical scheme:
the invention provides a device for predicting structural performance of a bridge group with a medium and small span, which is used for predicting structural performance of a bridge group consisting of a plurality of bridges with medium and small spans, and is characterized by comprising the following steps: a detection report storage unit for storing a year-over detection report for each bridge; a database construction unit for constructing a relational database of the bridge group according to the past year detection report; the model storage part is used for storing a neural network model which is used for forecasting the structural performance degradation of the bridge group and is subjected to training and inspection according to the relational database; a scheme storage section storing a plurality of predetermined maintenance schemes for maintaining the bridge; and a performance prediction unit that predicts the trend of performance changes of the entire structure and the local members of the bridge group by performing calculation using the neural network model.
The invention provides a device for predicting the structural performance of a small and medium span bridge group, which can also have the characteristics that a database construction part comprises: the data extraction unit is used for extracting original evaluation data of each bridge from the annual detection report, wherein the original evaluation data comprise technical condition scores, bridge ages, structure types, traffic volumes and maintenance behaviors of each bridge; a cleaning rule storage unit storing data cleaning rules for cleaning the original evaluation data; the data cleaning unit cleans the original evaluation data according to the data cleaning rule so as to obtain predicted evaluation data; and the data processing unit is used for processing the prediction evaluation data and constructing a relational database which is based on the prediction evaluation data and is an attribute set.
The invention provides a device for predicting structural performance of a group of bridges with medium and small spans, which can also have the characteristics that an attribute set comprises maintenance behavior attributes, structure type attributes, bridge age attributes, traffic volume attributes and technical condition scoring attributes of the bridges, and a data processing unit comprises: a maintenance behavior processing subunit, configured to perform binary conversion on the maintenance behavior attribute, and if the bridge is maintained, set a value of the corresponding maintenance behavior attribute to 1, and otherwise set the value to 0; a structure type processing subunit, configured to perform vectorization processing on the structure type attribute;
the bridge age processing subunit performs normalization transformation on the bridge age attribute:
a' is bridge age after normalization transformation, a is bridge age before normalization transformation, a max Is the maximum value of bridge age;
traffic processing subunit, which performs normalization transformation on traffic attributes:
beta' is the traffic after normalization and beta is normalizationTraffic before conversion, beta max Is the maximum value of traffic volume;
the technical condition scoring processing subunit performs normalization transformation on the technical condition scoring attribute:
c' is the technical condition score after normalization transformation and c is the technical condition score before normalization transformation.
The invention provides a device for predicting structural performance of a middle-small span bridge group, which can also have the characteristics that an original neural network model is a multi-hidden-layer feedforward neural network model and comprises an input layer and an output layer, wherein the input layer is provided with 6 input layer neurons, 3 of the input neurons are corresponding to structural types, the other 3 of the input neurons are respectively corresponding to bridge age, traffic volume and maintenance behavior, the output layer is provided with 1 output layer neuron, the output neuron is corresponding to annual technical condition scores, the input layer neurons and the output layer neurons are all connected between layers of the original neural network model in a defining way, and meanwhile, no connection exists in the layers of the original neural network model.
The invention provides a device for predicting the structural performance of a middle-small span bridge group, which can also have the characteristic that the network learning rate of an original neural network model is initialized to 0.1.
The invention provides a device for predicting the structural performance of a middle-small span bridge group, which can also have the characteristics that an original neural network model also comprises hidden layers, and 20 hidden layer neurons are arranged in the hidden layers.
The invention provides a device for predicting the structural performance of a middle-small span bridge group, which can also have the characteristic that the loss function of an original neural network model is defined as the mean square error between a predicted value and a true value output by the model.
The invention provides a device for predicting the structural performance of a small and medium span bridge group, which can also have the characteristics that the performance predicting part comprises: the bridge record generating unit generates corresponding bridge records for different bridge ages from the last year to be predicted according to the increase of the bridge ages; the maintenance behavior value unit is used for acquiring the value of the maintenance behavior in the bridge record according to a preset maintenance scheme; the bridge record filling unit is used for filling the structure type and the traffic volume into each bridge record respectively; and the technical condition score output unit takes the bridge record as the input of the neural network model so as to output the predicted technical condition scores of each year under the preset maintenance scheme.
The invention provides a device for predicting the structural performance of a small and medium span bridge group, which can also have the characteristic that the traffic volume is the annual average daily traffic volume.
The invention provides a system for predicting structural performance of a group of small and medium span bridges, which is characterized by comprising the following steps: the performance prediction device is used for predicting the structural performance of a bridge group consisting of a plurality of bridges with medium and small spans; the storage server stores all the annual detection reports of the bridges; and the detection report acquisition device is used for acquiring the calendar year detection reports of all bridges in the bridge group from the storage server and sending the calendar year detection reports to the performance prediction device, wherein the performance prediction device is the performance prediction device for the middle-small span bridge group structure.
The actions and effects of the invention
According to the device for predicting the structural performance of the middle and small span bridge group, the database construction part is provided, so that information extraction, data integration and normalization can be carried out on the regional bridge detection report accumulated throughout the year, and structural parameters and performance degradation trend thereof can be converted into a relational database. The model storage part is provided, so that the relationship between each parameter of the bridge age, the type, the traffic volume and the maintenance behavior and the structural state can be simulated by training an artificial neural network model. The performance prediction part is provided, so that the performance change and degradation trend of the bridge in the future area can be further predicted based on a mature neural network model, and the obtained structure performance prediction result of the middle and small span bridge group has high practicability through practical calculation verification, and provides effective decision support for management and maintenance of the middle and small span bridge group.
In conclusion, the device for predicting the structural performance of the small and medium span bridge group integrates the data mining technology, effectively and fully utilizes mass detection data accumulated in the long-term bridge inspection work, establishes a neural network model, converts the extracted data into valuable knowledge in the field of bridge management and maintenance, and realizes the bridge structural performance evaluation prediction and management and maintenance guidance of the small and medium span bridge group.
Drawings
FIG. 1 is a block diagram of a device for predicting structural performance of a group of small and medium span bridges;
FIG. 2 is a block diagram of the data construction unit of the present invention;
FIG. 3 is a schematic diagram of information interaction between a relational database and a neural network model of the present invention;
FIG. 4 is a schematic diagram of the structure of the neural network of the present invention;
FIG. 5 is a comparative schematic of predicted results for different maintenance schemes for the bridge of the present invention;
FIG. 6 is a workflow diagram of a device for predicting structural performance of a group of small and medium span bridges according to the present invention;
fig. 7 is a block diagram of a system for predicting structural performance of a group of small and medium span bridges of the present invention.
Detailed Description
In order to make the technical means, creation characteristics, achievement purposes and effects of the present invention easy to understand, the beam bridge safety monitoring and evaluating device of the present invention is specifically described below with reference to the accompanying drawings.
Example 1 ]
Fig. 1 is a block diagram of a device for predicting structural performance of a group of bridges with medium and small spans.
As shown in fig. 1, the device 100 for predicting structural performance of a bridge group with a small and medium span according to the present invention is configured to predict structural performance of a bridge group consisting of a plurality of bridges with a small and medium span in a target traffic network, and includes a detection report storage unit 10, a database construction unit 20, a model storage unit 30, a scenario storage unit 40, a performance prediction unit 50, a management and maintenance advice acquisition unit 60, a screen storage unit 70, and an input display unit 80.
The detection report storage unit 10 stores an annual detection report for each bridge.
Fig. 2 is a block diagram of the data construction unit according to the present invention.
As shown in fig. 2, the database construction unit 20 is configured to construct a relational database of bridge groups from the past year detection report, and includes a data extraction unit 201, a cleansing rule storage unit 202, a data cleansing unit 203, and a data processing unit 204.
The data extraction unit 201 is configured to extract, from the historical detection report, raw evaluation data of each bridge, where the raw evaluation data includes a technical condition score, a bridge age, a structure type, a traffic volume, and a maintenance behavior of each bridge. And selecting the characteristics, screening bridge age, structure type, average annual traffic volume, maintenance behavior and annual technical condition scoring fields of the bridge as an attribute set of a relational database, and extracting and integrating data of each report.
The cleansing rule storage unit 202 stores data cleansing rules for cleansing the original evaluation data.
The data cleansing unit 203 is configured to cleansing the original evaluation data according to a data cleansing rule to obtain predicted evaluation data. Namely, deleting the data with the missing or error phenomenon. Specifically, according to the data cleaning rule, if any attribute value of a record in the database has the phenomena of deletion and error, the record is deleted, so that the validity and usability of the data are ensured. For example, if the bridge age attribute value of a record is-1 or blank, then the entire record is deleted.
The data processing unit 204 processes the predictive evaluation data and builds a relational database based on the predictive evaluation data as a set of attributes.
The attribute set includes a maintenance behavior attribute, a structure type attribute, a bridge age attribute, a traffic volume attribute, and a technical condition scoring attribute of the bridge,
the data processing unit 204 includes: maintenance action processing subunit 204a, structure type processing subunit 204b, bridge age processing subunit 204c, traffic processing subunit 204d, and technical condition scoring processing subunit 204e.
The maintenance behavior processing subunit 204a is configured to perform binary conversion on the maintenance behavior attribute, and if the bridge is maintained, the corresponding value of the maintenance behavior attribute is set to 1, and otherwise, is set to 0.
The structure type processing subunit 204b is configured to perform vectorization processing on the structure type attribute, and divide all bridges in the relational database into three types, i.e. a girder bridge, a box girder bridge, and others, and convert them into (1, 0), (0, 1, 0), (0, 1), respectively.
The bridge age processing subunit 204c is configured to perform normalization transformation on the bridge age attribute, and the calculation formula is as follows:
wherein a' is bridge age after normalization transformation, a is bridge age before normalization transformation, a max Is the maximum value of bridge age.
The traffic processing subunit 204d is configured to perform normalization transformation on the traffic attribute, and the calculation formula is as follows:
wherein, beta' is the traffic after normalization, beta is the traffic before normalization, beta max Is the maximum value of traffic.
The technical condition score processing subunit 204e is configured to perform normalization transformation on the technical condition score attribute, and the calculation formula is as follows:
wherein c' is the technical condition score after normalization transformation and c is the technical condition score before normalization transformation.
Fig. 3 is a schematic diagram of information interaction between the relational database and the neural network model according to the present invention, and fig. 4 is a schematic diagram of the structure of the neural network according to the present invention.
As shown in fig. 3 to 4, the model storage unit 30 stores a neural network model for which training test is completed based on the relational database.
The neural network model is used for predicting structural performance degradation of the bridge group and is a multi-hidden-layer feedforward neural network model and comprises an input layer, an output layer and a hidden layer.
The input layer has 6 input layer neurons, 3 of which correspond to structural types (respectively plate girder bridge, box girder bridge and others), and the remaining 3 correspond to bridge age, traffic volume and repair actions, respectively.
The output layer has 1 output layer neuron, which corresponds to the annual state of the art score.
The input layer neuron and the output layer neuron are connected between layers of the neural network model in a definition mode, the corresponding weight coefficients are randomly sampled according to the interval of 0-1 to be initialized, and meanwhile no connection exists in the layers of the neural network model.
The number of hidden layers is 1 layer, and 20 hidden layer neurons are set.
In this embodiment, the network learning rate of the neural network model is initialized to 0.1, and the training method of the neural network model specifically includes:
defining a loss function of the neural network model as a predicted value y of the model output i ' and true value y i The mean square error between the two is outputted as normalized technical condition score by the model, and if n records in the database are imported into the model as a data set, the calculation formula of the loss is as follows
And performing iterative training on the network by adopting a BP algorithm until the error is lower than a preset value.
The scenario storage portion 40 stores a plurality of predetermined maintenance scenarios for the bridge, including the cost of each maintenance scenario.
FIG. 5 is a comparative schematic of predicted results for different maintenance scenarios for the bridge of the present invention.
As shown in fig. 5, the performance predicting unit 50 predicts the performance trend of the entire structure and the partial members of the bridge group by using the technical score status of the corresponding bridge using the neural network model with the attribute-concentrated performance parameter as input data, and includes a bridge record generating unit 501, a maintenance behavior evaluating unit 502, a bridge record filling unit 503, and a technical score outputting unit 504.
The bridge record generating unit 501 is configured to generate corresponding bridge records for different bridge ages from the last year to be predicted according to the increase of the bridge ages.
The maintenance action value unit 502 is configured to obtain a value of a maintenance action in the bridge record according to a predetermined maintenance scheme.
The bridge record filling unit 503 is configured to fill the structure type and the traffic volume into each bridge record, that is, the bridge record includes bridge age, repair behavior, structure type and traffic volume of the bridge under the predetermined repair scheme.
The technical condition score output unit 504 is configured to record the bridge as input data of the neural network model, so as to output the technical condition score of each year under the predetermined maintenance scheme through model calculation.
The management advice acquisition unit 60 is configured to acquire management advice for each bridge in the corresponding bridge group according to the technical status score of each year. For example, in combination with the cost of each predetermined maintenance scenario in the scenario storage section 40, the technical condition scores that each maintenance scenario can improve at the same cost are ranked, and the higher the technical condition score that can improve, the better the corresponding maintenance scenario is, that is, the optimal maintenance proposal.
The screen storage unit 70 stores a display screen of the original evaluation data, a display screen of the predicted evaluation data, a display screen of the performance prediction, a display screen of the management advice, a setting screen of the neural network model, a setting screen of the data cleaning rule, and a setting screen of the preset maintenance schedule.
The input display unit 80 is a liquid crystal display having a touch screen function, and is configured to display a display screen of original evaluation data, a display screen of predicted evaluation data, a display screen of performance prediction, and a display screen of management advice, and the input display unit 80 is also configured to display a setting screen of a neural network model for allowing a user to set model parameters, a setting screen for setting data cleansing rules, and a setting screen for setting a preset maintenance schedule.
FIG. 6 is a workflow diagram of the device for predicting structural performance of a group of bridges with medium and small spans.
Before predicting the structural performance of the bridge group, the user determines the target traffic network area to be predicted first, selects the bridges to be predicted in the area as the bridge group, collects and summarizes the historical detection reports of all bridges in the bridge group, and stores the detection reports in the detection report storage unit 10.
The user sets the model parameters of the neural network model, sets the data cleaning rule, sets the maintenance schedule, and sets the preset maintenance schedule by inputting the setting screen of the neural network model of the display unit 80, and stores the schedule in the schedule storage unit 40.
The working flow of the device 100 for predicting the structural performance of a group of bridges with small and medium spans according to the present embodiment is described below with reference to the accompanying drawings, and specifically includes the following steps:
step S1, the database construction unit 20 extracts bridge age, type, average annual traffic volume, maintenance behavior and technical condition score of each year of each bridge in the past year detection report, and constructs a relational database;
step S2, a neural network model is built based on the data in the relational database, training and inspection are performed, and the neural network model after the training inspection is stored in the model storage part 30;
step S3, the performance predicting part 50 takes the attribute concentrated performance parameters as input data, namely the structural type, annual traffic volume, bridge age and maintenance behavior of the bridge to be predicted, and outputs the technical scoring condition of the corresponding bridge by utilizing the neural network model so as to predict the performance change trend of the whole structure and the local components of the bridge group;
and S4, according to the technical scoring condition of the bridges output by the model, proposing a management and maintenance suggestion of each bridge in the corresponding bridge group and displaying the management and maintenance suggestion on a display picture of the management and maintenance suggestion.
Example 2 ]
Fig. 7 is a block diagram of a system for predicting structural performance of a group of small and medium span bridges of the present invention.
As shown in fig. 7, the system 500 for predicting the performance of the middle and small span bridge group structure in embodiment 2 includes the device 100 for predicting the performance of the middle and small span bridge group structure in embodiment 1, the detection report acquiring device 300, and the storage server 400. The device 100 for predicting the structural performance of the middle and small span bridge group, the device 300 for acquiring the detection report and the storage server 400 are all connected through wireless communication.
The storage server 400 stores all bridge design drawings, past year inspection reports, and repair records. Such as a database of inspection reports and maintenance records for a road and government agency, a transportation investment group, a municipal maintenance company, a database of original design drawings for a design house, etc.
The detection report obtaining device 300 is used for collecting the annual detection reports of all bridges in the bridge group to be detected from the storage server 400 and sending the annual detection reports to the middle and small span bridge group structure performance prediction device 100.
Before predicting the structural performance of the bridge group, the user determines the target traffic network area to be predicted and selects the bridge to be predicted in the area as the bridge group, and the detection report acquisition device 300 collects and summarizes the historical detection reports of all bridges in the bridge group from the storage server 400, and then sends the detection reports to the medium-and-small span bridge group structural performance prediction device 100 and stores the detection reports in the detection report storage unit 10.
Example operation and Effect
According to the device for predicting the structural performance of the middle and small span bridge group, the database construction part is provided, so that information extraction, data integration and normalization can be carried out on the regional bridge detection report accumulated throughout the year, and structural parameters and performance degradation trend thereof can be converted into a relational database. The model storage part is provided, so that the relationship between each parameter of the bridge age, the type, the traffic volume and the maintenance behavior and the structural state can be simulated by training an artificial neural network model. The performance prediction part is provided, so that the performance change and degradation trend of the bridge in the future area can be further predicted based on a mature neural network model, and the obtained structure performance prediction result of the middle and small span bridge group has high practicability through practical calculation verification, and provides effective decision support for management and maintenance of the middle and small span bridge group.
In conclusion, the device for predicting the structural performance of the small and medium span bridge group integrates the data mining technology, effectively and fully utilizes mass detection data accumulated in the long-term bridge inspection work, establishes a neural network model, converts the extracted data into valuable knowledge in the field of bridge management and maintenance, and realizes the bridge structural performance evaluation prediction and management and maintenance guidance of the small and medium span bridge group.
The database construction part comprises a data extraction unit, a cleaning rule storage unit, a data cleaning unit and a data processing unit, so that data with missing and error phenomena in the original data can be removed, and the calculation result of the neural network model is more accurate.
The data processing unit comprises a maintenance behavior processing subunit, a structure type processing subunit, a bridge age processing subunit, a traffic processing subunit and a technical condition scoring processing subunit, so that each attribute in the attribute set can be processed and calculated respectively, the data in the relational database is more accurate, and the calculation efficiency of the neural network model is higher.
The multi-hidden-layer feedforward neural network model is adopted, so that the model is a mature and effective calculation model comprising an input layer, an output layer and a hidden layer, and the technical condition scoring of the bridge obtained by outputting the model is more accurate, and the structural performance prediction of the bridge group is also more accurate.
The system for predicting the structural performance of the middle and small span bridge group also comprises the detection report acquisition device and the storage server, and detection personnel can acquire the calendar year detection reports of all bridges in the bridge group to be detected from the storage server through wireless communication and send the calendar year detection reports to the performance prediction device, so that the detection personnel can acquire calendar year detection reports of the bridges and excavate the data more accurately and comprehensively, and the accuracy and the working efficiency of bridge group prediction results are further improved.
The foregoing describes in detail preferred embodiments of the present invention. It should be understood that numerous modifications and variations can be made in accordance with the concepts of the invention without requiring creative effort by one of ordinary skill in the art. Therefore, all technical solutions which can be obtained by logic analysis, reasoning or limited experiments based on the prior art by the person skilled in the art according to the inventive concept shall be within the scope of protection defined by the claims.
For example, in the scheme storage part in the embodiment, a predetermined maintenance scheme is stored, the technical condition scores of the bridges under different maintenance schemes are obtained through calculation by using a neural network model, so that the maintenance scheme with the highest technical condition score can be improved under the same cost as the optimal maintenance suggestion. The device for predicting the structural performance of the middle-small span bridge group is also provided with a ranking part, the ranking part is provided with a comprehensive scoring calculation unit and a ranking unit, the comprehensive scoring calculation unit is used for carrying out weighted calculation on the maintenance scheme according to manpower and material resources consumed by the maintenance scheme and time factors, so that the score of the maintenance scheme is obtained, and the ranking unit is used for ranking the maintenance scheme according to the ranking order. Further, the screen storage unit stores a setting screen for setting the weight value, and the input display unit displays the setting screen for setting the weight value so that the user can set the weight value.
Claims (8)
1. A small and medium span bridge group structural performance prediction apparatus for predicting structural performance of a bridge group consisting of a plurality of small and medium span bridges, comprising:
a detection report storage unit that stores a past year detection report for each bridge;
a database construction unit for constructing a relational database of the bridge group according to the past year detection report;
a model storage unit which stores a neural network model for predicting structural performance degradation of the bridge group, the neural network model being subjected to training and inspection based on the relational database;
a scheme storage section storing a plurality of predetermined maintenance schemes for maintaining the bridge; and
a performance prediction unit for predicting the performance trend of the entire structure and the local members of the bridge group by calculating using the neural network model,
wherein the database construction section includes:
the data extraction unit is used for extracting original evaluation data of each bridge from the past year detection report, wherein the original evaluation data comprise technical condition scores, bridge ages, structure types, traffic volumes and maintenance behaviors of each bridge;
a cleaning rule storage unit storing a data cleaning rule for cleaning the original evaluation data;
the data cleaning unit cleans the original evaluation data according to the data cleaning rule so as to obtain predicted evaluation data;
a data processing unit for processing the predicted evaluation data and constructing a relational database based on the predicted evaluation data as an attribute set including maintenance behavior attribute, structure type attribute, bridge age attribute, traffic volume attribute and technical condition scoring attribute of the bridge,
the performance prediction unit includes:
the bridge record generating unit generates corresponding bridge records for different bridge ages from the last year to be predicted according to the increase of the bridge ages;
the maintenance behavior value unit is used for acquiring the value of the maintenance behavior in the bridge record according to the preset maintenance scheme;
the bridge record filling unit is used for filling the structure type and the traffic volume into each bridge record respectively;
and the technical condition score output unit is used for taking the bridge record as the input of the neural network model so as to output the predicted technical condition score of each year under the preset maintenance scheme.
2. The device for predicting structural performance of a group of small and medium span bridges according to claim 1, wherein:
wherein the data processing unit comprises:
a maintenance behavior processing subunit, configured to perform binary conversion on the maintenance behavior attribute, and if the bridge is maintained, set a corresponding value of the maintenance behavior attribute to 1, and otherwise set the value to 0;
a structure type processing subunit, configured to perform vectorization processing on the structure type attribute;
and the bridge age processing subunit performs normalization transformation on the bridge age attribute, and the calculation formula is as follows:
a' is bridge age after normalization transformation, a is bridge age before normalization transformation, a max Is the maximum value of bridge age;
and the traffic processing subunit performs normalization transformation on the traffic attributes, and the calculation formula is as follows:
beta' is the traffic after normalization, beta is the traffic before normalization, beta max Is the maximum value of traffic volume;
the technical condition scoring processing subunit performs normalization transformation on the technical condition scoring attribute, and the calculation formula is as follows:
c' is the technical condition score after normalization transformation and c is the technical condition score before normalization transformation.
3. The device for predicting structural performance of a group of small and medium span bridges according to claim 1, wherein:
wherein the neural network model is a multi-hidden layer feedforward neural network model and comprises an input layer and an output layer,
the input layer has 6 input layer neurons, 3 of which correspond to the structural type, the remaining 3 correspond to the bridge age, the traffic volume and the repair action, respectively,
the output layer having 1 output layer neuron, the output neuron corresponding to the state of the art score per year,
the input layer neurons and the output layer neurons both define connections between layers of the neural network model, while there are no connections within layers of the neural network model.
4. The device for predicting structural performance of a group of small and medium span bridges according to claim 3, wherein:
wherein, the network learning rate of the neural network model is initialized to 0.1.
5. The device for predicting structural performance of a group of small and medium span bridges according to claim 3, wherein:
the neural network model further comprises hidden layers, and 20 hidden layer neurons are set in the hidden layers.
6. The device for predicting structural performance of a group of small and medium span bridges according to claim 3, wherein:
the loss function of the neural network model is defined as the mean square error between the predicted value and the true value of the model output.
7. The device for predicting structural performance of a group of small and medium span bridges according to claim 1, wherein:
wherein the traffic volume is an average daily traffic volume.
8. The utility model provides a medium and small span bridge crowd structural performance prediction system which characterized in that includes:
the performance prediction device is used for predicting the structural performance of a bridge group consisting of a plurality of bridges with medium and small spans;
the storage server stores all the past year detection reports of the bridge; and
a detection report acquisition device for acquiring the calendar year detection reports of all the bridges in the bridge group from the storage server and sending the calendar year detection reports to the performance prediction device,
wherein the performance prediction device is a device for predicting the structural performance of a small and medium span bridge group according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910061475.4A CN109902344B (en) | 2019-01-23 | 2019-01-23 | Device and system for predicting structural performance of middle-small span bridge group |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910061475.4A CN109902344B (en) | 2019-01-23 | 2019-01-23 | Device and system for predicting structural performance of middle-small span bridge group |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109902344A CN109902344A (en) | 2019-06-18 |
| CN109902344B true CN109902344B (en) | 2023-06-27 |
Family
ID=66944011
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910061475.4A Active CN109902344B (en) | 2019-01-23 | 2019-01-23 | Device and system for predicting structural performance of middle-small span bridge group |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109902344B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111242347B (en) * | 2019-12-28 | 2021-01-01 | 浙江大学 | Bridge management and maintenance aid decision-making system based on historical weight updating |
| CN111738996B (en) * | 2020-06-09 | 2023-04-07 | 交通运输部公路科学研究所 | Bridge health monitoring and early warning system based on machine learning |
| CN113469374B (en) * | 2021-09-02 | 2021-12-24 | 北京易真学思教育科技有限公司 | Data prediction method, device, equipment and medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106017556A (en) * | 2016-05-20 | 2016-10-12 | 清华大学合肥公共安全研究院 | Bridge safety operation monitoring and management system |
| CN108629144A (en) * | 2018-06-11 | 2018-10-09 | 湖北交投智能检测股份有限公司 | A kind of bridge health appraisal procedure |
-
2019
- 2019-01-23 CN CN201910061475.4A patent/CN109902344B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106017556A (en) * | 2016-05-20 | 2016-10-12 | 清华大学合肥公共安全研究院 | Bridge safety operation monitoring and management system |
| CN108629144A (en) * | 2018-06-11 | 2018-10-09 | 湖北交投智能检测股份有限公司 | A kind of bridge health appraisal procedure |
Non-Patent Citations (5)
| Title |
|---|
| 中小跨径混凝土梁桥退化预测和维护策略研究;时笑鹏;《中国优秀硕士学位论文全文数据库 工程科技II辑》;20190115;正文第1-97页 * |
| 公路桥梁养护决策与管理系统研究;赵进;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20090815;正文第1-74页 * |
| 基于BP神经网络的桥梁技术状态评估;向木生等;《交通科技》;20111031;正文第41-43页 * |
| 寒区桥梁养护管理系统研究;李晓辉;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20111215;正文第1-72页 * |
| 李晓辉.寒区桥梁养护管理系统研究.《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》.2011,正文第1-72页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109902344A (en) | 2019-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109285346B (en) | Urban road network traffic state prediction method based on key road sections | |
| CN109902344B (en) | Device and system for predicting structural performance of middle-small span bridge group | |
| CN109614669B (en) | Network-level bridge structure performance evaluation and prediction method | |
| CN108199795A (en) | The monitoring method and device of a kind of equipment state | |
| CN111738462B (en) | Fault first-aid repair active service early warning method for electric power metering device | |
| CN105574593B (en) | Track state static detection and control system and method based on cloud computing and big data | |
| RU2670937C1 (en) | Forecasting maintenance operations to be applied to an engine | |
| CN104778622A (en) | Method and system for predicting TPS transaction event threshold value | |
| CN101853290A (en) | Meteorological service performance evaluation method based on geographical information system (GIS) | |
| CN116681187B (en) | Enterprise carbon quota prediction method based on enterprise operation data | |
| CN111178585A (en) | Fault reporting amount prediction method based on multi-algorithm model fusion | |
| Chen et al. | Vulnerability modeling, assessment, and improvement in urban metro systems: A probabilistic system dynamics approach | |
| CN111950895A (en) | BIM-based assembly type building all-stage design and evaluation system | |
| CN117172556B (en) | Construction risk early warning method and system for bridge engineering | |
| CN108537336B (en) | Air quality prediction method based on deep neural network | |
| CN110555551A (en) | air quality big data management method and system for smart city | |
| CN111199298A (en) | Flood forecasting method and system based on neural network | |
| CN103942251A (en) | Method and system for inputting high altitude meteorological data into database based on multiple quality control methods | |
| e Matos et al. | An overview of COST Action TU 1406, Quality Specifications for Roadway Bridges (BridgeSpec) | |
| Whipple | Predictive storm damage modeling and optimizing crew response to improve storm response operations | |
| CN116245399A (en) | Model training method and device, nonvolatile storage medium and electronic equipment | |
| CN116149895A (en) | Big data cluster performance prediction method and device and computer equipment | |
| Kosgodagan et al. | Expert judgment in life-cycle degradation and maintenance modelling for steel bridges | |
| CN114861973A (en) | Wind power generation system risk assessment method considering multi-source factors | |
| CN115018110A (en) | XGboost-based road blocking loss prediction method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |