Concrete data processing method and system

Abstract

The invention provides a concrete data processing method and system, and the method comprises the steps: obtaining change data corresponding to a plurality of measurement points in a concrete dam, and obtaining a plurality of monitoring clusters comprising a plurality of clustering measurement points based on the change data; obtaining a final prediction model; selecting one of the monitoring clusters as a to-be-analyzed cluster, selecting a cluster measurement point in the to-be-analyzed cluster as a to-be-analyzed measurement point, obtaining a prediction analysis monitoring value of the to-be-analyzed measurement point through the final prediction model, and obtaining an abnormal threshold value of the to-be-analyzed cluster based on the prediction analysis monitoring value; and performing anomaly judgment on the real monitoring value of the to-be-analyzed monitoring point based on the anomaly threshold. Through the above mode, on the basis of considering the relevance between the data of the single measuring point, the abnormal value is identified by combining the relevance between different measuring points in terms of time and space, so that the identification precision of the abnormal value is effectively improved, and the accuracy of subsequent deformation monitoring analysis is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly relates to a method and system for processing concrete data. Background Art

[0002] A concrete dam refers to a dam made of cast concrete or assembled with precast concrete blocks. A concrete dam can discharge water or take water through the dam body, eliminating the need for dedicated water discharge and water intake structures. During the construction process, construction diversion and flood control during construction are relatively easy. Compared with an earth-rock dam, its hub layout is more compact, facilitating operation and management.

[0003] Among the numerous monitoring items of a concrete dam, deformation monitoring is one of the more important monitoring items. By obtaining the deformation conditions of the concrete dam through deformation monitoring, once an anomaly is detected, it can be analyzed, studied, measures can be taken, and processed in a timely manner to prevent accidents and ensure the safe operation of the concrete dam. During the deformation monitoring process, corresponding concrete data needs to be collected. However, due to the influence of instrument failures, environmental factors, etc., there are abnormal values in the collected concrete data. The existence of abnormal values will reduce the accuracy of deformation monitoring.

[0004] Currently, generally based on the 3σ principle, abnormal values are identified in the concrete data obtained from a single measurement point. However, this method only considers the correlation between the data of a single measurement point, which is prone to misjudgment or missed judgment of abnormal values, thereby affecting the accuracy of deformation monitoring. Summary of the Invention

[0005] Aiming at the deficiencies of the prior art, the purpose of the present invention is to provide a method and system for processing concrete data, aiming to solve the technical problem that in the prior art, when identifying abnormal values in the concrete data obtained from a single measurement point based on the 3σ principle, it is prone to misjudgment or missed judgment of abnormal values, thereby affecting the accuracy of deformation monitoring.

[0006] To achieve the above purpose, in the first aspect, an embodiment of the present application provides a method for processing concrete data, including the following steps:

[0007] Obtain change data corresponding to a plurality of measurement points in a concrete dam, and based on the change data, divide the plurality of measurement points into a plurality of monitoring clusters, where each monitoring cluster includes a plurality of clustered measurement points;

[0008] Train an initial prediction model based on the monitoring data corresponding to the clustered measurement points to obtain a final prediction model;

[0009] Select one of the monitoring clusters as the cluster to be analyzed, select the cluster measurement points in the cluster to be analyzed as the measurement points to be analyzed, obtain the predicted analysis monitoring value of the measurement points to be analyzed through the final prediction model, and obtain the anomaly threshold of the cluster to be analyzed based on the predicted analysis monitoring value;

[0010] Based on the anomaly threshold, perform anomaly determination on the true monitoring value of the measurement points to be analyzed to complete the identification of outliers.

[0011] Compared with the prior art, the beneficial effects of the present invention are as follows: By aggregating several of the measurement points into the monitoring clusters based on the change data and performing overall analysis of subsequent outliers on the monitoring clusters, the spatial correlation between each measurement point is considered; By constructing the final prediction model and obtaining the predicted analysis monitoring value of the measurement points to be analyzed based on the final prediction model, and then determining the anomaly threshold according to the difference between the predicted analysis detection value and the true monitoring value, the time correlation between each measurement point is considered; That is, on the basis of considering the correlation between the data of a single measurement point, the identification of outliers is completed by combining the correlations between different measurement points in terms of time and space, effectively improving the identification accuracy of outliers, and further improving the accuracy of subsequent deformation monitoring analysis.

[0012] Further, the change data includes the magnitude of the deformation value, the change range of the deformation value, and the change speed of the deformation value. The step of separating several of the measurement points into several monitoring clusters based on the change data, where the monitoring cluster includes several cluster measurement points, includes:

[0013] Select one of the measurement points as the reference measurement point, and select the remaining measurement points as candidate measurement points. Obtain the first distance between the candidate measurement point and the reference measurement point based on the magnitude of the deformation value, obtain the second distance between the candidate measurement point and the reference measurement point based on the change range of the deformation value, and obtain the third distance between the candidate measurement point and the reference measurement point based on the change speed of the deformation value;

[0014] Determine the fitting distance between the candidate measurement point and the reference measurement point through the first distance, the second distance, and the third distance;

[0015] Based on the fitting distance, select several similar measurement points from several candidate measurement points, select the reference measurement point and several similar measurement points as cluster measurement points, and aggregate several of the cluster measurement points into a monitoring cluster.

[0016] Furthermore, the calculation formula of the fitting distance is:

[0017]

[0018] Among them, Dis ij represents the fitting distance between the i-th reference measurement point and the j-th candidate measurement point, and q1, q2, and q3 all represent distance weights. represents the first distance between the i-th reference measurement point and the j-th candidate measurement point. represents the second distance between the i-th reference measurement point and the j-th candidate measurement point. represents the third distance between the i-th reference measurement point and the j-th candidate measurement point.

[0019] Furthermore, the step of selecting a plurality of similar measurement points from a plurality of the candidate measurement points based on the fitting distance includes:

[0020] comparing the fitting distances between the plurality of candidate measurement points and the reference measurement point with a distance threshold respectively;

[0021] selecting the candidate measurement points with the fitting distance greater than the distance threshold as the similar measurement points.

[0022] Furthermore, the step of training an initial prediction model based on the monitoring data corresponding to the clustered measurement points to obtain a final prediction model includes:

[0023] Based on the time series, splitting the monitoring data corresponding to the clustered measurement points into pre-order data and post-order data;

[0024] Constructing an initial prediction model, using the pre-order data as the input value of the initial prediction model, and using the post-order data as the output value of the initial prediction model to train the initial prediction model to obtain a final prediction model.

[0025] Furthermore, the step of obtaining the anomaly threshold of the cluster to be analyzed based on the predicted analysis monitoring value includes:

[0026] Obtaining the true monitoring value of the measurement point to be analyzed, and obtaining the monitoring difference of the measurement point to be analyzed based on the true monitoring value and the predicted analysis monitoring value;

[0027] Obtaining the difference ratio of the measurement point to be analyzed based on the monitoring difference, so as to obtain the anomaly threshold of the cluster to be analyzed through the difference ratio.

[0028] Furthermore, the calculation formula of the difference ratio is:

[0029]

[0030] Among them, z i represents the difference ratio of the i-th measurement point to be analyzed, c i represents the monitoring difference of the i-th measurement point to be analyzed, and i = 1, 2,..., n;

[0031] The calculation formula of the abnormal threshold is as follows:

[0032] yc j = c1*z1 + c2*z2... + Cn *z n ,

[0033] where yc j represents the abnormal threshold of the jth clustering to be analyzed.

[0034] Furthermore, the step of performing abnormal determination on the true monitoring value of the measuring point to be analyzed based on the abnormal threshold to complete the identification of abnormal values includes:

[0035] Selecting a value to be analyzed from the true monitoring values corresponding to the measuring point to be analyzed, obtaining a predicted evaluation value corresponding to the value to be analyzed based on the final prediction model, and obtaining the actual difference between the predicted evaluation value and the value to be analyzed;

[0036] Comparing the actual difference with the abnormal threshold. If the actual difference is greater than three times the abnormal threshold, the value to be analyzed is determined as an abnormal value.

[0037] Further, after the step of performing abnormal determination on the true monitoring value of the measuring point to be analyzed based on the abnormal threshold to complete the identification of abnormal values, it further includes:

[0038] Obtaining a first replacement value based on the true monitoring values adjacent to the abnormal value, selecting the predicted evaluation value corresponding to the abnormal value as the second replacement value, obtaining a final replacement value through the first replacement value and the second replacement value, and replacing the abnormal value with the final replacement value.

[0039] In a second aspect, an embodiment of the present application provides a concrete data processing system, which is applied to the concrete data processing method as described in the first aspect above. The system includes:

[0040] A classification module, configured to obtain change data corresponding to a plurality of measuring points in a concrete dam, and separate the plurality of measuring points into a plurality of monitoring clusters based on the change data. The monitoring clusters include a plurality of clustered measuring points;

[0041] A construction module, configured to train an initial prediction model based on the monitoring data corresponding to the clustered measuring points to obtain a final prediction model;

[0042] An evaluation module is configured to select one of the monitoring clusters as a cluster to be analyzed, select the cluster measurement points in the cluster to be analyzed as measurement points to be analyzed, obtain the predicted analysis monitoring values of the measurement points to be analyzed through the final prediction model, and obtain the anomaly threshold of the cluster to be analyzed based on the predicted analysis monitoring values;

[0043] An analysis module is configured to perform anomaly determination on the true monitoring values of the measurement points to be analyzed based on the anomaly threshold to complete the identification of outliers.

[0044] In a third aspect, an embodiment of the present application provides a computer, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the concrete data processing method described in the first aspect above is implemented.

[0045] In a fourth aspect, an embodiment of the present application provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the concrete data processing method described in the first aspect above is implemented. Description of the Drawings

[0046] Figure 1 is a flowchart of the concrete data processing method in the first embodiment of the present invention;

[0047] Figure 2 is a structural block diagram of the concrete data processing system in the second embodiment of the present invention;

[0048] The following specific embodiments will further illustrate the present invention in conjunction with the above-mentioned drawings. Specific Embodiments

[0049] To facilitate the understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided to make the disclosure of the present invention more thorough and comprehensive.

[0050] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly on the other element or there can also be a middle element. When an element is considered to be "connected" to another element, it can be directly connected to the other element or there may be a middle element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0051] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs. The terms used in the specification of this invention are only for the purpose of describing specific embodiments and are not intended to limit the invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0052] Please refer to Figure 1 , the concrete data processing method provided by the first embodiment of the present invention includes the following steps:

[0053] S10: Obtain the change data corresponding to several measuring points in the concrete dam, and based on the change data, separate several of the measuring points into several monitoring clusters, where the monitoring cluster includes several clustered measuring points;

[0054] The change data includes the magnitude of the deformation value, the change range of the deformation value, and the change speed of the deformation value.

[0055] The step S10 includes:

[0056] S110: Select one of the measuring points as the reference measuring point, and select the remaining measuring points as candidate measuring points. Based on the magnitude of the deformation value, obtain the first distance between the candidate measuring point and the reference measuring point. Based on the change range of the deformation value, obtain the second distance between the candidate measuring point and the reference measuring point. Based on the change speed of the deformation value, obtain the third distance between the candidate measuring point and the reference measuring point;

[0057] In this embodiment, the first distance, the second distance, and the third distance are respectively the difference in the magnitude of the deformation value, the difference in the change range of the deformation value, and the difference in the change speed of the deformation value between the candidate measuring point and the reference measuring point.

[0058] S120: Determine the fitting distance between the candidate measuring point and the reference measuring point through the first distance, the second distance, and the third distance;

[0059] The calculation formula for the fitting distance is:

[0060]

[0061] where Dis ij represents the fitting distance between the i-th reference measuring point and the j-th candidate measuring point, and q1, q2, and q3 all represent distance weights, represents the first distance between the i-th reference measuring point and the j-th candidate measuring point, denote the second distance between the i-th reference measurement point and the j-th candidate measurement point, denote the third distance between the i-th reference measurement point and the j-th candidate measurement point.

[0062] S130: Based on the fitting distance, select several similar measurement points from several candidate measurement points, select the reference measurement point and several similar measurement points as clustering measurement points, and aggregate several clustering measurement points into a monitoring cluster;

[0063] Compare the fitting distances between several candidate measurement points and the reference measurement point with a distance threshold respectively; select the candidate measurement points with the fitting distance greater than the distance threshold as similar measurement points.

[0064] After obtaining one of the monitoring clusters, repeat steps S110 to S130 for the remaining measurement points to complete the acquisition of several monitoring clusters.

[0065] S20: Train an initial prediction model based on the monitoring data corresponding to the clustering measurement points to obtain a final prediction model;

[0066] The step S20 includes:

[0067] S210: Based on the time series, split the monitoring data corresponding to the clustering measurement points into pre-sequence data and post-sequence data;

[0068] S220: Construct an initial prediction model, use the pre-sequence data as the input value of the initial prediction model, and use the post-sequence data as the output value of the initial prediction model to train the initial prediction model to obtain a final prediction model;

[0069] It can be understood that several clustering measurement points can all form the pre-sequence data and the post-sequence data. Furthermore, by repeating the training multiple times, the prediction accuracy of the final prediction model can be improved. In this embodiment, the final prediction model is a bidirectional long short-term memory neural network model, which has been widely used and will not be elaborated here.

[0070] S30: Select one of the monitoring clusters as the cluster to be analyzed, select the clustering measurement points in the cluster to be analyzed as the measurement points to be analyzed, obtain the predicted analysis monitoring value of the measurement points to be analyzed through the final prediction model, and obtain the anomaly threshold of the cluster to be analyzed based on the predicted analysis monitoring value;

[0071] The step S30 includes:

[0072] S310: Obtain the true monitoring value of the measurement points to be analyzed, and obtain the monitoring difference of the measurement points to be analyzed based on the true monitoring value and the predicted analysis monitoring value;

[0073] The monitored difference is the difference between the true monitored value and the predicted analysis monitored value.

[0074] S320: Obtain the ratio of the difference of the measured point to be analyzed based on the monitored difference, so as to obtain the anomaly threshold of the cluster to be analyzed through the ratio of the difference;

[0075] The calculation formula of the ratio of the difference is:

[0076]

[0077] where z i represents the ratio of the difference of the i-th measured point to be analyzed, and c i represents the monitored difference of the i-th measured point to be analyzed, and i = 1, 2,..., n;

[0078] The calculation formula of the anomaly threshold is:

[0079] y cj = c1*z1 + c2*z2... + c n *Z n ,

[0080] where yc j represents the anomaly threshold of the j-th cluster to be analyzed.

[0081] S40: Make an anomaly determination on the true monitored value of the measured point to be analyzed based on the anomaly threshold, so as to complete the identification of the anomaly value;

[0082] The step S40 includes:

[0083] S410: Select the value to be analyzed from the true monitored values corresponding to the measured point to be analyzed, obtain the predicted evaluation value corresponding to the value to be analyzed based on the final prediction model, and obtain the actual difference between the predicted evaluation value and the value to be analyzed;

[0084] The value to be analyzed has the same essential meaning as the true monitored value, and the predicted evaluation value has the same essential meaning as the predicted analysis monitored value. The different expressions here are only for distinction.

[0085] S420: Compare the actual difference with the anomaly threshold. If the actual difference is greater than 3 times the anomaly threshold, then determine the value to be analyzed as an anomaly value;

[0086] It can be understood that if the actual difference is less than 3 times the anomaly threshold, then determine the value to be analyzed as a normal value and retain the value to be analyzed.

[0087] By aggregating several of the measurement points into the monitoring clusters based on the change data and performing an overall analysis of subsequent outliers for the monitoring clusters, the spatial correlation between each measurement point is considered; by constructing the final prediction model and obtaining the predicted analysis monitoring value of the measurement point to be analyzed based on the final prediction model, and then determining the anomaly threshold according to the difference between the predicted analysis detection value and the true monitoring value, the temporal correlation between each measurement point is considered; that is, on the basis of considering the correlation between the data of a single measurement point, the identification of outliers is completed by combining the temporal and spatial correlations between different measurement points, effectively improving the identification accuracy of outliers, and further improving the accuracy of subsequent deformation monitoring analysis.

[0088] Preferably, the method further includes:

[0089] S50: Obtain a first replacement value based on the true monitoring value adjacent to the outlier, select the predicted evaluation value corresponding to the outlier as the second replacement value, obtain a final replacement value through the first replacement value and the second replacement value, and replace the outlier with the final replacement value;

[0090] If the outlier is directly removed, it will cause the data of this measurement point to be missing, and further cause certain errors in the analysis data of deformation monitoring obtained based on this measurement point, affecting the analysis effect of deformation monitoring. Therefore, after identifying the outlier, by obtaining the final replacement value, the accuracy of subsequent deformation monitoring analysis can be guaranteed to a certain extent.

[0091] Please refer to Figure 2 , the second embodiment of the present invention provides a concrete data processing system, which is applied to the concrete data processing method in the above embodiment, and the parts that have been described will not be repeated. As used hereinafter, terms such as "module", "unit", "sub-unit", etc. can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0092] The system includes:

[0093] A classification module 10, configured to obtain change data corresponding to several measurement points in a concrete dam, and separate several of the measurement points into several monitoring clusters based on the change data, where the monitoring clusters include several clustered measurement points;

[0094] The classification module 10 includes:

[0095] The first unit is configured to select one of the measurement points as a reference measurement point, and select the remaining measurement points as candidate measurement points. Based on the magnitude of the deformation value, obtain a first distance between the candidate measurement point and the reference measurement point. Based on the change range of the deformation value, obtain a second distance between the candidate measurement point and the reference measurement point. Based on the change speed of the deformation value, obtain a third distance between the candidate measurement point and the reference measurement point;

[0096] The second unit is configured to determine a fitting distance between the candidate measurement point and the reference measurement point through the first distance, the second distance, and the third distance;

[0097] The third unit is configured to select a number of similar measurement points from a number of the candidate measurement points based on the fitting distance, select the reference measurement point and the number of similar measurement points as clustering measurement points, and aggregate the number of clustering measurement points into a monitoring cluster;

[0098] Specifically, the third unit is configured to compare the fitting distances between a number of the candidate measurement points and the reference measurement point with a distance threshold respectively; select the candidate measurement points with the fitting distance greater than the distance threshold as similar measurement points;

[0099] The construction module 20 is configured to train an initial prediction model based on the monitoring data corresponding to the clustering measurement points to obtain a final prediction model;

[0100] The construction module 20 includes:

[0101] The fourth unit is configured to, based on a time series, divide the monitoring data corresponding to the clustering measurement points into pre-sequence data and post-sequence data;

[0102] The fifth unit is configured to construct an initial prediction model, use the pre-sequence data as input values of the initial prediction model, and use the post-sequence data as output values of the initial prediction model to train the initial prediction model to obtain a final prediction model;

[0103] The evaluation module 30 is configured to select one of the monitoring clusters as a cluster to be analyzed, select the clustering measurement points in the cluster to be analyzed as measurement points to be analyzed, obtain a predicted analysis monitoring value of the measurement points to be analyzed through the final prediction model, and obtain an anomaly threshold of the cluster to be analyzed based on the predicted analysis monitoring value;

[0104] The evaluation module 30 includes:

[0105] The sixth unit is configured to obtain a true monitoring value of the measurement points to be analyzed, and obtain a monitoring difference of the measurement points to be analyzed based on the true monitoring value and the predicted analysis monitoring value;

[0106] A seventh unit, configured to obtain a ratio of the difference of the measurement point to be analyzed based on the monitoring difference, so as to obtain an abnormal threshold of the cluster to be analyzed through the ratio of the difference;

[0107] An analysis module 40, configured to perform an abnormality determination on the true monitoring value of the measurement point to be analyzed based on the abnormal threshold, so as to complete the identification of abnormal values;

[0108] The analysis module 40 includes:

[0109] An eighth unit, configured to select a value to be analyzed from the true monitoring values corresponding to the measurement point to be analyzed, obtain a predicted evaluation value corresponding to the value to be analyzed based on the final prediction model, and obtain an actual difference between the predicted evaluation value and the value to be analyzed;

[0110] A ninth unit, configured to compare the actual difference with the abnormal threshold. If the actual difference is greater than three times the abnormal threshold, the value to be analyzed is determined as an abnormal value;

[0111] Preferably, the system further includes:

[0112] An execution module 50, configured to obtain a first replacement value based on the true monitoring values adjacent to the abnormal value, select the predicted evaluation value corresponding to the abnormal value as a second replacement value, obtain a final replacement value through the first replacement value and the second replacement value, and replace the abnormal value with the final replacement value.

[0113] The present invention also provides a computer, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the concrete data processing method described in the above technical solution is implemented.

[0114] The present invention also provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the concrete data processing method described in the above technical solution is implemented.

[0115] In the description of this specification, the descriptions with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0116] The above-described embodiments merely represent several implementation manners of the present invention. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent for the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several variations and improvements can still be made, and these all fall within the protection scope of the present invention. Therefore, the protection scope of the patent for the present invention shall be subject to the appended claims.

Claims

1. A method for processing concrete data, characterized in that, Including the following steps: Obtain change data corresponding to a number of measuring points in the concrete dam, and based on the change data, separate the number of measuring points into a number of monitoring clusters, where the monitoring cluster includes a number of clustered measuring points; Train an initial prediction model based on the monitoring data corresponding to the clustered measuring points to obtain a final prediction model; Select one of the monitoring clusters as the cluster to be analyzed, select the clustered measuring points in the cluster to be analyzed as the measuring points to be analyzed, obtain the predicted analysis monitoring values of the measuring points to be analyzed through the final prediction model, and obtain the anomaly threshold of the cluster to be analyzed based on the predicted analysis monitoring values; Perform anomaly determination on the true monitoring values of the measuring points to be analyzed based on the anomaly threshold to complete the identification of anomaly values.

2. The concrete data processing method according to claim 1, characterized in that, The change data includes the magnitude of the deformation value, the change amplitude of the deformation value, and the change speed of the deformation value. The step of separating the number of measuring points into a number of monitoring clusters based on the change data, where the monitoring cluster includes a number of clustered measuring points, includes: Select one of the measuring points as a reference measuring point, and select the remaining measuring points as candidate measuring points. Obtain the first distance between the candidate measuring point and the reference measuring point based on the magnitude of the deformation value, obtain the second distance between the candidate measuring point and the reference measuring point based on the change amplitude of the deformation value, and obtain the third distance between the candidate measuring point and the reference measuring point based on the change speed of the deformation value; Determine the fitting distance between the candidate measuring point and the reference measuring point through the first distance, the second distance, and the third distance; Select a number of similar measuring points from the number of candidate measuring points based on the fitting distance, select the reference measuring point and the number of similar measuring points as clustered measuring points, and aggregate the number of clustered measuring points into a monitoring cluster.

3. The concrete data processing method according to claim 2, wherein The calculation formula of the fitting distance is: Among them, Dis ij represents the fitting distance between the i-th reference measurement point and the j-th candidate measurement point, and q1, q2, and q3 all represent distance weights. represents the first distance between the i-th reference measurement point and the j-th candidate measurement point. represents the second distance between the i-th reference measurement point and the j-th candidate measurement point. represents the third distance between the i-th reference measurement point and the j-th candidate measurement point.

4. The concrete data processing method according to claim 2, characterized in that, The step of selecting a number of similar measuring points from the number of candidate measuring points based on the fitting distance includes: Compare the fitting distances between the number of candidate measuring points and the reference measuring point with a distance threshold respectively; Select the candidate measuring points with the fitting distance greater than the distance threshold as similar measuring points.

5. The concrete data processing method according to claim 1, characterized in that The step of training an initial prediction model based on the monitoring data corresponding to the clustered measuring points to obtain a final prediction model includes: Based on the time series, cut the monitoring data corresponding to the clustered measuring points into pre-order data and post-order data; Construct an initial prediction model, use the pre-order data as the input value of the initial prediction model, and use the post-order data as the output value of the initial prediction model to train the initial prediction model to obtain a final prediction model.

6. The concrete data processing method according to claim 1, characterized in that The step of obtaining the anomaly threshold of the cluster to be analyzed based on the predicted analysis monitoring values includes: Obtain the true monitoring values of the measuring points to be analyzed, and obtain the monitoring difference of the measuring points to be analyzed based on the true monitoring values and the predicted analysis monitoring values; Obtain the difference ratio of the measuring points to be analyzed based on the monitoring difference, so as to obtain the anomaly threshold of the cluster to be analyzed through the difference ratio.

7. The concrete data processing method according to claim 6, wherein The calculation formula of the difference ratio is: Among them, z i represents the ratio of the difference of the i-th measurement point to be analyzed, c i represents the monitoring difference of the i-th measurement point to be analyzed, and i = 1, 2,..., n; The calculation formula of the anomaly threshold is: yc j = c1*z1 + c2*z2… + c n *z n , Among them, yc j represents the anomaly threshold of the j-th clustering to be analyzed.

8. The concrete data processing method according to claim 1, characterized in that The step of performing anomaly determination on the true monitoring value of the measurement point to be analyzed based on the anomaly threshold to complete the identification of anomaly values includes: Select a value to be analyzed from the true monitoring values corresponding to the measurement point to be analyzed, obtain a predicted evaluation value corresponding to the value to be analyzed based on the final prediction model, and obtain the actual difference between the predicted evaluation value and the value to be analyzed; Compare the actual difference with the anomaly threshold. If the actual difference is greater than three times the anomaly threshold, determine the value to be analyzed as an anomaly value.

9. The concrete data processing method according to claim 8, characterized in that, After the step of performing anomaly determination on the true monitoring value of the measurement point to be analyzed based on the anomaly threshold to complete the identification of anomaly values, it further includes: Obtain a first replacement value based on the true monitoring values adjacent to the anomaly value, select the predicted evaluation value corresponding to the anomaly value as the second replacement value, obtain a final replacement value through the first replacement value and the second replacement value, and replace the anomaly value with the final replacement value.

10. A concrete data processing system, which is applied to the concrete data processing method according to any one of claims 1 to 9, characterized in that, The system includes: A classification module for obtaining change data corresponding to a number of measurement points in a concrete dam, and separating the number of measurement points into a number of monitoring clusters based on the change data, where the monitoring clusters include a number of clustered measurement points; A construction module for training an initial prediction model based on the monitoring data corresponding to the clustered measurement points to obtain a final prediction model; An evaluation module for selecting one of the monitoring clusters as a cluster to be analyzed, selecting the clustered measurement points in the cluster to be analyzed as measurement points to be analyzed, obtaining the predicted analysis monitoring values of the measurement points to be analyzed through the final prediction model, and obtaining the anomaly threshold of the cluster to be analyzed based on the predicted analysis monitoring values; An analysis module for performing anomaly determination on the true monitoring values of the measurement points to be analyzed based on the anomaly threshold to complete the identification of anomaly values.

Images