A mobile road surface water accumulation monitoring data analysis and processing system

Abstract

The present invention relates to the technical field of data processing, and in particular to a mobile road surface water accumulation monitoring data analysis and processing system. The system includes a data acquisition module for obtaining a road surface water depth data sequence within a preset time period; a data analysis module for obtaining the interference degree of each data in the road surface water depth data sequence; a filtering processing module for, when performing moving average filtering on the road surface water depth data sequence by using a sliding window, for any sliding window, obtaining an adaptive weight for the corresponding data according to the interference degree of each data in any sliding window, and performing mean filtering on all data within any sliding window to obtain a weighted mean; and a data monitoring module for forming a target filtering data sequence from the weighted means of all sliding windows, thereby enhancing the denoising effect of filtering the road surface water depth data and improving the accuracy of subsequent water accumulation warning based on the target filtering data sequence.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a mobile road surface water accumulation monitoring data analysis and processing system. Background Art

[0002] In recent years, during summer (especially the flood season), heavy rainfall occurs frequently. Coupled with low terrain, backflow caused by rising river water levels, and imperfect urban drainage systems, urban waterlogging occurs frequently, especially in low-lying areas, tunnels, overpasses, underground railways, underground garages, underground shopping malls and other areas of the city. In order to avoid urban waterlogging disasters, reduce casualties and economic losses, it is necessary to comprehensively understand the urban waterlogging situation, monitor the river water level and the water accumulation levels of underpass bridges and low-lying sections in real time, and establish an urban waterlogging early warning system.

[0003] In the prior art, when performing mobile real-time monitoring of road surface water accumulation, the depth of road surface water accumulation is measured by mounting sensors on a moving vehicle, and the monitored road surface water accumulation depth data is transmitted to a cloud server or a local data processing center in real time. Then, the collected road surface water accumulation depth data is analyzed, and according to the data analysis results, a water accumulation warning is generated. When the water level or rainfall exceeds a preset value, an alarm is automatically issued to notify relevant departments to take countermeasures in advance.

[0004] Since the monitored road surface water accumulation depth data is easily affected by the factors of the sensor itself (such as the sensor being affected by electromagnetic interference, etc.) or external environmental factors (such as high wind speed will cause the sensor to shake to varying degrees, which may interfere with the measurement of the sensor, and floating objects in the water accumulation will change the spectral reflection characteristics of the water body, resulting in changes in the performance of the water body in the remote sensing image, etc.), the collected road surface water accumulation depth data is inaccurate. Therefore, a moving average filtering process is performed on the collected road surface water accumulation depth data. Specifically, the data mean in each moving window is calculated as the denoised data after filtering for the corresponding moving window. However, if the data mean of the moving window is directly used as the denoised data after the corresponding filtering process, the denoising effect on the abnormal data in the moving window will be reduced, thus affecting the subsequent analysis results of the denoised data and the accuracy of the water accumulation warning.

[0005] Therefore, how to enhance the denoising effect of filtering the road surface water accumulation depth data to improve the accuracy of subsequent water accumulation warnings has become an urgent problem to be solved. Summary of the Invention

[0006] In view of this, an embodiment of the present invention provides a mobile road surface water accumulation monitoring data analysis and processing system to solve the problem of how to enhance the denoising effect of filtering the road surface water accumulation depth data to improve the accuracy of subsequent water accumulation warnings.

[0007] In an embodiment of the present invention, a mobile road surface water accumulation monitoring data analysis and processing system is provided. The system includes the following modules:

[0008] A data acquisition module, configured to obtain a road surface water accumulation depth data sequence within a preset time period;

[0009] A data analysis module, configured to respectively obtain the interference degree of each data in the road surface water accumulation depth data sequence according to the position relationship of each data in the road surface water accumulation depth data sequence and the data difference between each data and its adjacent data;

[0010] A filtering processing module, configured to, in the process of performing moving average filtering on the road surface water accumulation depth data sequence by using a sliding window of a preset size, for any sliding window, respectively obtain the adaptive weight of the corresponding data according to the interference degree of each data in the any sliding window, and perform mean filtering on all data in the any sliding window according to the adaptive weight of each data in the any sliding window to obtain the corresponding weighted mean value;

[0011] A data monitoring module, configured to form a target filtering data sequence from the weighted mean values corresponding to all sliding windows, and perform real-time monitoring and early warning on road surface water accumulation according to the target filtering data sequence.

[0012] Preferably, in the data analysis module, respectively obtaining the interference degree of each data in the road surface water accumulation depth data sequence according to the position relationship of each data in the road surface water accumulation depth data sequence and the data difference between each data and its adjacent data includes:

[0013] Respectively mark each data between the nth data and the (N - n + 1)th data in the road surface water accumulation depth data sequence as the target data in the road surface water accumulation depth data sequence, where N is the number of elements in the road surface water accumulation depth data sequence;

[0014] For any target data, in the road surface water accumulation depth data sequence, respectively obtain a preset number of elements on both sides adjacent to the any target data, and form a target data sequence with the any target data, and obtain the dissimilarity degree of the any target data according to the element difference in the target data sequence;

[0015] Obtain the local mutation degree of the any target data according to the adjacent data of the target data sequence in the road surface water accumulation depth data sequence;

[0016] Perform weighted summation on the dissimilarity degree and the local mutation degree to obtain the interference degree of the any target data;

[0017] According to the interference degree of each of the target data, the interference degree of each non-target data in the road surface water depth data sequence is obtained respectively.

[0018] Preferably, in the data analysis module, according to the element differences in the target data sequence, the dissimilarity degree of any one of the target data is obtained, including:

[0019] Mark each element in the target data sequence except the any one of the target data as the remaining data, obtain the absolute value of the difference between the any one of the target data and each of the remaining data, get the average absolute difference value, use the average absolute difference value as the variable of the hyperbolic tangent function, and obtain the corresponding first hyperbolic tangent value;

[0020] Obtain the sum of the elements in the target data sequence, denoted as the first data sum, calculate the proportion of each element in the target data sequence to the first data sum respectively, and form a first ratio sequence;

[0021] Obtain the mean value of all the remaining data in the target data sequence, use the mean value as the reference data of the any one of the target data, form a reference data sequence with each of the remaining data and the reference data, obtain the sum of the elements in the reference data sequence, denoted as the second data sum, calculate the proportion of each element in the reference data sequence to the second data sum respectively, and form a second ratio sequence;

[0022] Obtain the cross entropy between the first ratio sequence and the second ratio sequence, use the cross entropy as the variable of the hyperbolic tangent function, and obtain the corresponding second hyperbolic tangent value;

[0023] Use the mean value of the first hyperbolic tangent value and the second hyperbolic tangent value as the dissimilarity degree of the any one of the target data.

[0024] Preferably, in the data analysis module, according to the adjacent data of the target data sequence in the road surface water depth data sequence, the local mutation degree of any one of the target data is obtained, including:

[0025] In the road surface water depth data sequence, obtain a first data sequence composed of a preset first number of elements before the target data sequence. In a two-dimensional plane, construct a first data change curve of the first data sequence, where the horizontal axis of the two-dimensional plane is time and the vertical axis is the road surface water depth. Obtain a second data sequence composed of the preset first number of elements after the target data sequence. In the two-dimensional plane, construct a second data change curve of the second data sequence. Obtain the Spearman correlation coefficient between the first data change curve and the second data change curve. Obtain the addition result of the Spearman correlation coefficient and the constant 1, and perform normalization processing on the reciprocal of the addition result to obtain a first normalization result;

[0026] Obtain the mean value of the elements in the road surface water depth data sequence except the elements in the target data sequence, denoted as the first mean value. Obtain the mean value of the elements in the target data sequence, denoted as the second mean value. Obtain the absolute value of the first difference between the first mean value and the second mean value, and perform normalization processing on the absolute value of the first difference to obtain a second normalization result;

[0027] Take the mean value of the first normalization result and the second normalization result as the local mutation degree of any target data.

[0028] Preferably, in the data analysis module, according to the interference degree of each target data, obtain the interference degree of each non-target data in the road surface water depth data sequence respectively, including:

[0029] According to the interference degree of a preset target number of consecutive target data including the first target data in the road surface water depth data sequence, obtain the minimum interference degree, and take the minimum interference degree as the interference degree of each data before the nth data in the road surface water depth data sequence;

[0030] According to the interference degree of the preset target number of consecutive target data including the last target data in the road surface water depth data sequence, obtain the minimum interference degree, and take the minimum interference degree as the interference degree of each data after the (N - n + 1)th data in the road surface water depth data sequence.

[0031] Preferably, in the filtering processing module, according to the interference degree of each data in any sliding window, obtain the adaptive weight of the corresponding data respectively, including:

[0032] According to the interference degree of each data in any sliding window, obtain the total interference degree. Obtain the ratio between the total interference degree and the interference degree of each data respectively to obtain the total ratio. Respectively obtain the proportion between each ratio and the total ratio as the adaptive weight of the corresponding data.

[0033] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0034] The present invention provides a mobile road surface water accumulation monitoring data analysis and processing system, including a data acquisition module for obtaining a road surface water depth data sequence within a preset time period; a data analysis module for obtaining the interference degree of each data in the road surface water depth data sequence according to the positional relationship of each data in the road surface water depth data sequence and the data difference between each data and its adjacent data; a filtering processing module for, in the process of performing moving average filtering on the road surface water depth data sequence by using a sliding window of a preset size, for any sliding window, obtaining the adaptive weight of the corresponding data according to the interference degree of each data in the any sliding window, and performing mean filtering on all data in the any sliding window according to the adaptive weight of each data in the any sliding window to obtain the corresponding weighted mean value; a data monitoring module for forming a target filtering data sequence by using the weighted mean values corresponding to all sliding windows, and performing real-time monitoring and early warning on the road surface water accumulation according to the target filtering data sequence. Among them, when performing moving average filtering processing on the road surface water depth data sequence, according to the interference degree of each data in each sliding window, the adaptive weight of the corresponding data in each sliding window is obtained. The greater the interference degree, the smaller the adaptive weight of the corresponding data. Then, mean filtering is performed on all data in each sliding window to obtain the weighted mean value corresponding to each sliding window, and a target filtering data sequence is formed, thereby eliminating the reduction of the denoising effect of the data in the sliding window caused by using the data mean value in the sliding window, enhancing the denoising effect of filtering the road surface water depth data, and improving the accuracy of subsequent water accumulation early warning based on the target filtering data sequence. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0036] Figure 1 It is a structural block diagram of a mobile road surface water accumulation monitoring data analysis and processing system provided in Embodiment 1 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0037] Embodiments of the present disclosure will be described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present disclosure, and should not be construed as a limitation on the present disclosure.

[0038] It should be noted that the terms "first", "second", etc. in the specification of the present disclosure and the above-mentioned accompanying drawings are used to distinguish similar objects and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present disclosure.

[0039] In order to illustrate the technical solution of the present invention, specific embodiments will be used for illustration below.

[0040] See Figure 1 , which is a structural block diagram of a mobile road surface water accumulation monitoring data analysis and processing system provided in the first embodiment of the present invention. As Figure 1 shown, the system may include:

[0041] A data acquisition module 11, configured to obtain a sequence of road surface water accumulation depth data within a preset time period.

[0042] By installing sensors (such as a mobile road surface condition monitor) on a mobile monitoring vehicle, during the driving process of the mobile monitoring vehicle, the depth of water accumulation on the road surface is quickly and real-time monitored by using the mobile road surface condition monitor, and a sequence of road surface water accumulation depth data within 2 hours of monitoring is collected.

[0043] It should be noted that the monitoring time period set in the embodiment of the present invention is 2 hours, which is not limited thereto and can be set according to specific real-time scenarios.

[0044] A data analysis module 12, configured to obtain the interference degree of each data in the road surface water accumulation depth data sequence respectively according to the position relationship of each data in the road surface water accumulation depth data sequence and the data difference between each data and its adjacent data.

[0045] Considering that the road surface water depth data sequence is subject to varying degrees of interference during the acquisition process, including environmental factors, sensor self-factors, etc., in the prior art, the sliding mean filtering technique is usually used to filter the acquired road surface water depth data sequence to obtain the denoised road surface water depth data sequence. However, due to the influence of sensor self-device factors or external environmental factors, the degree of interference of each road surface water depth data collected is different. If the data mean in the sliding window is directly used as the denoised data after filtering, the denoising effect on the abnormal data in the sliding window is reduced, thereby affecting the denoising effect of the road surface water depth data sequence.

[0046] Therefore, to solve the above problems, before performing sliding mean filtering on the road surface water depth data sequence, the embodiments of the present invention first obtain the interference degree of each data in the road surface water depth data sequence, which is used to assign an adaptive weight to each data according to the interference degree during subsequent sliding mean filtering, so as to improve the filtering effect of the sliding mean filtering.

[0047] For any data in the road surface water depth data sequence, when it is interfered by the sensor and environmental factors, it is manifested as: the similarity degree between this data and adjacent data is low or the local mutability of the neighborhood where this data is located is relatively high. Among them, the size of the similarity degree between this data and adjacent data reflects the correlation degree between this data and adjacent data. The higher the similarity degree, the higher the correlation degree between this data and adjacent data, and the less interference it is subject to; also, because the similarity between this data and adjacent data does not necessarily represent whether there is mutability in the neighborhood where this data is located, it is possible that the similarity degree between this data and adjacent data is high, but the local mutability of the neighborhood where this data is located is large, which is also a situation where the water accumulation data is interfered. Therefore, it is necessary to comprehensively analyze the noise interference degree of each data in the road surface water depth data sequence.

[0048] Among them, the method for obtaining the noise interference degree of each data in the road surface water depth data sequence is as follows:

[0049] (1) Mark each data between the nth data and the (N - n + 1)th data in the road surface water depth data sequence as the target data in the road surface water depth data sequence, where N is the number of elements in the road surface water depth data sequence.

[0050] In an embodiment, set n = 6, then mark the 6th data to the (N - 5)th data in the road surface water depth data sequence as the target data in this road surface water depth data sequence.

[0051] (2)For any target data, in the road surface water depth data sequence, respectively obtain a preset number of elements on both sides adjacent to the any target data, and form a target data sequence with the any target data. According to the element differences in the target data sequence, obtain the dissimilarity degree of the any target data.

[0052] Specifically, mark each element in the target data sequence except the any target data as the remaining data, obtain the absolute value of the difference between the any target data and each of the remaining data, get the average absolute difference value, use the average absolute difference value as the variable of the hyperbolic tangent function, and obtain the corresponding first hyperbolic tangent value;

[0053] Obtain the sum of the elements in the target data sequence, denoted as the first data sum, calculate the proportion of each element in the target data sequence to the first data sum respectively, and form a first ratio sequence;

[0054] Obtain the mean value of all the remaining data in the target data sequence, use the mean value as the reference data of the any target data, form a reference data sequence with each of the remaining data and the reference data, obtain the sum of the elements in the reference data sequence, denoted as the second data sum, calculate the proportion of each element in the reference data sequence to the second data sum respectively, and form a second ratio sequence;

[0055] Obtain the cross entropy between the first ratio sequence and the second ratio sequence, use the cross entropy as the variable of the hyperbolic tangent function, and obtain the corresponding second hyperbolic tangent value;

[0056] Take the mean value of the first hyperbolic tangent value and the second hyperbolic tangent value as the dissimilarity degree of the any target data.

[0057] In an embodiment, considering that the higher the correlation degree of adjacent data in the local domain, that is, the greater the similarity degree, if the similarity degree between a certain data and its adjacent data is lower, that is, the correlation degree between the data and its adjacent data is lower, it indicates that the possibility of this data being interfered by noise is greater. Therefore, by obtaining the dissimilarity degree of the data, it represents the possibility of this data being interfered by noise.

[0058] Therefore, taking the i-th target data as an example, set the preset number to 2 (this is not limited). In the road surface water depth data sequence, respectively obtain 2 consecutive elements adjacent to the left of the i-th target data and 2 consecutive elements adjacent to the right of the i-th target data, and form a target data sequence with the i-th target data. Mark each element in the target data sequence except the i-th target data as the remaining data, and obtain the absolute value of the difference between the i-th target data and each of the remaining data, to get the average absolute difference value.

[0059] Obtain the proportion of each element in the target data sequence among all elements respectively to form the first ratio sequence. Obtain the mean value of all the remaining data, use this mean value as the reference data for the \(i\)-th target data, and form a reference data sequence with the reference data and all the remaining data. Obtain the proportion of each element in the reference data sequence among all elements respectively to form the second ratio sequence, and calculate the cross-entropy between the first ratio sequence and the second ratio sequence.

[0060] According to the mean absolute difference and the cross-entropy, obtain the dissimilarity degree of the \(i\)-th target data, and its expression is:

[0061]

[0062] where, represents the dissimilarity degree of the \(i\)-th target data, represents the \(i\)-th target data, represents the \(h\)-th remaining data in the target data sequence, \(H\) represents the total number of the remaining data in the target data sequence, \(p(d)\) represents the \(d\)-th proportion in the first ratio sequence, \(q(d)\) represents the \(d\)-th proportion in the second ratio sequence, \(\lg()\) represents the logarithmic function with base 10, \(\tanh()\) represents the hyperbolic tangent function, and \(|\ |\) represents the absolute value symbol.

[0063] It should be noted that, the larger the value of, the greater the difference between the \(i\)-th target data and its adjacent elements, that is, the lower the correlation degree between the \(i\)-th target data and its adjacent elements, indicating that the possibility of the \(i\)-th target data being an abnormal data is greater, that is, the possibility of being affected by noise interference is greater, corresponding to a greater dissimilarity degree of the \(i\)-th target data; the larger the value of, the greater the difference degree between the target data sequence and the reference data sequence, that is, the greater the difference between the \(i\)-th target data and its adjacent elements, indicating that the possibility of the \(i\)-th target data being an abnormal data is greater, that is, the possibility of being affected by noise interference is greater, corresponding to a greater dissimilarity degree of the \(i\)-th target data.

[0064] (3) According to the adjacent data of the target data sequence in the road surface water depth data sequence, obtain the local mutation degree of any target data.

[0065] Specifically, in the road surface water depth data sequence, obtain the first data sequence composed of the preset first number of elements before the target data sequence. In a two-dimensional plane, construct the first data change curve of the first data sequence, where the horizontal axis of the two-dimensional plane is time and the vertical axis is the road surface water depth. Obtain the second data sequence composed of the preset first number of elements after the target data sequence. In the two-dimensional plane, construct the second data change curve of the second data sequence. Obtain the Spearman correlation coefficient between the first data change curve and the second data change curve. Obtain the addition result of the Spearman correlation coefficient and the constant 1, and perform normalization processing on the reciprocal of the addition result to obtain the first normalization result;

[0066] Obtain the mean value of the elements in the road surface water depth data sequence except the elements in the target data sequence, denoted as the first mean value. Obtain the mean value of the elements in the target data sequence, denoted as the second mean value. Obtain the absolute value of the first difference between the first mean value and the second mean value, and perform normalization processing on the absolute value of the first difference to obtain the second normalization result;

[0067] Take the mean value of the first normalization result and the second normalization result as the local mutation degree of any target data.

[0068] In an embodiment, after obtaining the dissimilarity degree, considering that in the local neighborhood where the current data belongs, if the correlation degree between the current data and its adjacent data is greater, it does not necessarily mean that the current data is less affected by noise. It may also be that all the data in the local neighborhood where the current data belongs are affected by noise and mutate, resulting in a relatively large correlation degree between the current data and its adjacent data, that is, the data difference between the current data and its adjacent data is relatively small. To further prove whether the current data is affected by noise, take the local neighborhood where the current data is located as a whole. According to the difference between the data on the adjacent two sides of the local neighborhood where the current data is located, if the difference between the data on the adjacent two sides of the local neighborhood is greater, and there is an obvious difference between the overall data in the local area and the rest of the data, it indicates that the whole of the local neighborhood where the current data is located is more likely to be affected by noise, and further indicates that the current data is more likely to be noise data. Therefore, by obtaining the local mutation degree of the local neighborhood where the current data is located as the local mutation degree of the current data, it is further proved whether the current data is affected by noise.

[0069] Therefore, taking the i-th target data as an example, in the road surface water depth data sequence, obtain the target data sequence corresponding to the i-th target data. Based on the above value of n and the value of the preset quantity, set the preset first quantity to 3, where the value of the preset first quantity is greater than 2 and less than or equal to 3. Obtain the first 3 elements before the target data sequence to form the first data sequence. In the two-dimensional plane, construct the first data change curve of the first data sequence, where the horizontal axis of this two-dimensional plane represents time and the vertical axis represents the road surface water depth. Obtain the 3 elements after the target data sequence to form the second data sequence. In the two-dimensional plane, construct the second data change curve of the second data sequence. Obtain the Spearman correlation coefficient between the first fitting curve and the second fitting curve. The Spearman correlation coefficient belongs to the prior art and will not be elaborated here.

[0070] Obtain the mean value of the elements in the road surface water depth data sequence except for the elements in the target data sequence, denoted as the first mean value. Obtain the mean value of the elements in the target data sequence, denoted as the second mean value. According to the Spearman correlation coefficient, the first mean value and the second mean value, obtain the local mutation degree of the i-th target data, and its expression is:

[0071]

[0072] Where represents the local mutation degree of the i-th target data, represents the first data change curve, represents the second data change curve, represents the Spearman correlation coefficient, D represents the target data sequence, represents the j-th data in the target data sequence, represents the k-th data in the road surface water depth data sequence except for the target data sequence, N represents the number of elements in the road surface water depth data sequence, R represents the number of elements in the target data sequence, | | represents the absolute value symbol, norm() represents the normalization function, and 1 represents a constant.

[0073] It should be noted that The smaller the value of , the lower the correlation degree between the first data change curve and the second data change curve, that is, the more inconsistent the change degree of the data on both sides adjacent to the target data sequence, indicating that the target data sequence corresponding to the i-th target data is more likely to be affected by noise interference, that is, the i-th target data is more likely to be an abnormal data, and the local mutation degree corresponding to the i-th target data is greater; represents the mean value of the target data sequence, The larger the value, the greater the difference between the mean value within the local neighborhood where the i-th target data is located and the mean value of the remaining data, indicating that the data within the local neighborhood where the i-th target data is located is more dispersed compared to the remaining data, that is, the overall fluctuation degree of the local neighborhood where the i-th target data is located is greater, and the possibility of being affected by noise interference is greater, which also means that the possibility of the i-th target data being an abnormal data is greater, corresponding to a greater local mutation degree of the i-th target data.

[0074] (4) Perform weighted summation on the dissimilarity degree and the local mutation degree to obtain the interference degree of any target data.

[0075] In one embodiment, after obtaining the dissimilarity degree and the local mutation degree of the target data, considering that the higher the dissimilarity degree and the greater the local mutation degree of the target data, the greater the possibility that the target data is affected by noise interference, corresponding to a greater interference degree. Conversely, the lower the dissimilarity degree and the smaller the local mutation degree, the smaller the possibility that the target data is affected by noise interference, corresponding to a smaller interference degree.

[0076] Therefore, taking the i-th target data as an example, according to the dissimilarity degree and the local mutation degree of the i-th target data, the interference degree of the i-th target data is obtained, and its expression is:

[0077]

[0078] Where, represents the interference degree of the i-th target data, represents the dissimilarity degree of the i-th target data, represents the local mutation degree of the i-th target data, represents the weight of the dissimilarity degree, represents the weight of the local mutation degree.

[0079] Preferably, the weights of the dissimilarity degree and the local mutation degree set in the embodiments of the present invention are empirical values, that is , and there is no limitation on this, and it can be set according to specific implementation scenarios.

[0080] It should be noted that the larger the value, the smaller the correlation degree between the i-th target data and its adjacent data, indicating that the possibility of the i-th target data being an abnormal data is greater, that is, the possibility of being affected by noise interference is greater, corresponding to a greater interference degree of the i-th target data; The larger the value, the greater the overall fluctuation degree of the local neighborhood where the $i$-th target data is located, indicating that the local neighborhood where the $i$-th target data is located is more likely to be affected by noise interference, that is, the $i$-th target data is more likely to be abnormal data, and the greater the interference degree corresponding to the $i$-th target data.

[0081] According to the above method for obtaining the interference degree of the $i$-th target data, the interference degree of each target data is obtained respectively.

[0082] (5) According to the interference degree of each of the target data, the interference degree of each non-target data in the road surface water depth data sequence is obtained respectively.

[0083] Specifically, according to the interference degrees of a preset number of consecutive target data including the first target data in the road surface water depth data sequence, the minimum interference degree is obtained, and the minimum interference degree is used as the interference degree of each data before the $n$-th data in the road surface water depth data sequence;

[0084] According to the interference degrees of the preset number of consecutive target data including the last target data in the road surface water depth data sequence, the minimum interference degree is obtained, and the minimum interference degree is used as the interference degree of each data after the $(N - n + 1)$-th data in the road surface water depth data sequence.

[0085] In an embodiment, considering that the number of actually monitored road surface water depth data is large, the 5 data before the 6th data in the road surface water depth data sequence and the 5 data after the $(N - 5)$-th data have little impact on the overall data, and the correlation degree between the 5 data before the 6th data and the adjacent local data on their right is relatively high, and the correlation degree between the 5 data after the $(N - 5)$-th data and the adjacent local data on their left is relatively high.

[0086] Therefore, the preset number of targets is set to 5, which is not limited herein. The interference degrees of the $n$-th data to the $(n + 4)$-th data in the road surface water depth data sequence (that is, the interference degrees of the 6th data to the 10th data in the road surface water depth data sequence) are obtained respectively, and the minimum interference degree is obtained. The minimum interference degree is used as the interference degree of the 1st data to the 5th data in the road surface water depth data sequence respectively.

[0087] The interference degrees of the $(N - n + 1)$-th data to the $(N - n - 3)$-th data in the road surface water depth data sequence (that is, the interference degrees of the $(N - 5)$-th data to the $(N - 9)$-th data in the road surface water depth data sequence) are obtained respectively, and the minimum interference degree is obtained. The minimum interference degree is used as the interference degree of the $(N - 4)$-th data to the $N$-th data in the road surface water depth data sequence respectively.

[0088] So far, the interference degree of each data in the road surface water depth data sequence is obtained.

[0089] The filtering processing module 13 is used to, in the process of performing moving average filtering on the road surface water depth data sequence by using a sliding window with a preset size, for any sliding window, respectively obtain the adaptive weights of the corresponding data according to the interference degree of each data in any sliding window, and perform mean filtering on all the data in any sliding window according to the adaptive weights of each data in any sliding window to obtain the corresponding weighted mean value.

[0090] After obtaining the interference degree of each data in the road surface water depth data sequence, based on the interference degree of each data in the road surface water depth data sequence, the moving average filtering process of the road surface water depth data sequence is optimized to enhance the overall denoising effect of performing moving average filtering on the road surface water depth data sequence. Therefore, in the process of performing moving average filtering on the road surface water depth data sequence by using a sliding window with a preset size, first set the preset size to 10, and there is no limitation on this, then, for any sliding window, respectively obtain the adaptive weights of the corresponding data according to the interference degree of each data in any sliding window.

[0091] Considering that the interference degrees of each data in the sliding window are different, if a certain data in the sliding window is more affected by noise interference, it indicates that the data is more affected by abnormal interference. If the same weight is given to this data and the normal data in the sliding window, the denoising effect on this data will decrease. Therefore, in the embodiments of the present invention, a smaller weight is given to the data with a large interference degree, so as to improve the filtering effect of the sliding window, and further weaken the influence of the noise interference when performing moving average filtering on the road surface water depth data sequence.

[0092] Among them, respectively obtaining the adaptive weights of the corresponding data according to the interference degree of each data in any sliding window includes:

[0093] According to the interference degree of each data in the any sliding window, obtain the total interference degree, obtain the ratio between the total interference degree and the interference degree of each data, obtain the total ratio, and respectively obtain the proportion between each ratio and the total ratio as the adaptive weight of the corresponding data.

[0094] In an embodiment, taking the i-th data in the k-th sliding window as an example, according to the interference degree of the i-th data in the k-th sliding window, obtain the adaptive weight corresponding to the i-th data in the k-th sliding window, and its expression is:

[0095]

[0096]

[0097] Among them, represents the adaptive weight corresponding to the i-th data in the k-th sliding window, represents the degree of interference of the i-th data in the k-th sliding window, represents the ratio between the sum of the degrees of interference of all data in the k-th sliding window and the degree of interference of the i-th data.

[0098] According to the above method for obtaining the adaptive weight corresponding to the i-th data in the k-th sliding window, the adaptive weight corresponding to each data in the k-th sliding window is obtained respectively. Based on the adaptive weight corresponding to each data in the k-th sliding window, after obtaining the adaptive weight corresponding to each data in the k-th sliding window, weighted mean filtering is performed on the k-th sliding window based on the adaptive weight to obtain the corresponding average value, denoted as the weighted mean. Among them, the calculation expression of the weighted mean is:

[0099]

[0100] Among them, represents the weighted mean corresponding to the k-th sliding window, M represents the number of elements in the k-th sliding window, represents the adaptive weight corresponding to the j-th data in the k-th sliding window, represents the j-th data in the k-th sliding window.

[0101] The data monitoring module 14 is used to form a target filtered data sequence from the weighted means corresponding to all sliding windows, and based on the target filtered data sequence, real-time monitoring and early warning of road surface water accumulation are performed.

[0102] According to the above method for obtaining the weighted mean corresponding to the k-th sliding window, the weighted mean corresponding to each sliding window in the process of sliding mean filtering of the road surface water depth data sequence is obtained respectively. All weighted means are formed into a target filtered data sequence. At this time, the sliding mean filtering process of the road surface water depth data sequence is completed. After that, the target filtered data sequence is analyzed, and a visualization report is generated based on the analysis result, high-risk areas are identified, and water accumulation early warning information is generated, etc., so that subsequent water accumulation early warning is more accurate and efficient. For example, the water accumulation anomaly threshold is set to 15 cm, which is not limited here. For any weighted mean in the target filtered data sequence, if the weighted mean exceeds the water accumulation anomaly threshold, the water accumulation early warning is automatically triggered to remind the staff to take corresponding measures.

[0103] It should be noted that the focus of the present invention lies in how to optimize the moving average filtering process of the road surface water depth data sequence. As for real-time monitoring and early warning of road surface water accumulation based on the filtered road surface water depth data sequence (i.e., the target filtered data sequence), it belongs to the prior art and will not be elaborated here in detail.

[0104] In summary, the present invention provides a mobile road surface water accumulation monitoring data analysis and processing system, including a data acquisition module for obtaining a road surface water depth data sequence within a preset time period; a data analysis module for obtaining the interference degree of each data in the road surface water depth data sequence according to the position relationship of each data in the road surface water depth data sequence and the data difference between each data and its adjacent data; a filtering processing module for, during the process of performing moving average filtering on the road surface water depth data sequence using a sliding window of a preset size, for any sliding window, obtaining the adaptive weight of the corresponding data according to the interference degree of each data in the any sliding window, and performing mean filtering on all data in the any sliding window according to the adaptive weight of each data in the any sliding window to obtain the corresponding weighted mean; a data monitoring module for forming a target filtered data sequence from the weighted means corresponding to all sliding windows and performing real-time monitoring and early warning of road surface water accumulation according to the target filtered data sequence. Among them, when performing moving average filtering on the road surface water depth data sequence, according to the interference degree of each data in each sliding window, the adaptive weight of the corresponding data in each sliding window is obtained. The greater the interference degree, the smaller the adaptive weight of the corresponding data. Then, mean filtering is performed on all data in each sliding window to obtain the weighted mean corresponding to each sliding window, and a target filtered data sequence is formed, thereby eliminating the reduction in the denoising effect of the data in the sliding window caused by using the data mean in the sliding window, enhancing the denoising effect of filtering the road surface water depth data, and improving the accuracy of subsequent water accumulation early warning based on the target filtered data sequence.

[0105] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A mobile road surface waterlogging monitoring data analysis and processing system, characterized in that, The system includes: A data acquisition module, configured to obtain a sequence of road surface water depth data within a preset time period; A data analysis module, configured to obtain the interference degree of each data in the road surface water depth data sequence respectively according to the positional relationship of each data in the road surface water depth data sequence and the data difference between each data and its adjacent data; A filtering processing module, configured to, in the process of performing moving average filtering on the road surface water depth data sequence by using a sliding window of a preset size, for any sliding window, obtain the adaptive weight of the corresponding data respectively according to the interference degree of each data in the any sliding window, and perform mean filtering on all data in the any sliding window according to the adaptive weight of each data in the any sliding window to obtain the corresponding weighted mean value; A data monitoring module, configured to form a target filtering data sequence with the weighted mean values corresponding to all sliding windows, and perform real-time monitoring and early warning on road surface water accumulation according to the target filtering data sequence; In the data analysis module, obtaining the interference degree of each data in the road surface water depth data sequence respectively according to the positional relationship of each data in the road surface water depth data sequence and the data difference between each data and its adjacent data includes: Respectively mark each data between the nth data and the (N - n + 1)th data in the road surface water depth data sequence as the target data in the road surface water depth data sequence, where N is the number of elements in the road surface water depth data sequence; For any target data, in the road surface water depth data sequence, respectively obtain a preset number of elements on both sides adjacent to the any target data, and form a target data sequence with the any target data, and obtain the dissimilarity degree of the any target data according to the element difference in the target data sequence; Obtain the local mutation degree of the any target data according to the adjacent data of the target data sequence in the road surface water depth data sequence; Perform weighted summation on the dissimilarity degree and the local mutation degree to obtain the interference degree of the any target data; Obtain the interference degree of each non-target data in the road surface water depth data sequence respectively according to the interference degree of each target data.

2. The mobile road surface water accumulation monitoring data analysis and processing system according to claim 1, characterized in that, In the data analysis module, obtaining the dissimilarity degree of the any target data according to the element difference in the target data sequence includes: Mark each element except the any target data in the target data sequence as the remaining data, obtain the absolute value of the difference between the any target data and each of the remaining data respectively, obtain the average absolute difference value, and use the average absolute difference value as the variable of the hyperbolic tangent function to obtain the corresponding first hyperbolic tangent value; Obtain the sum of elements in the target data sequence, denoted as the first data sum, and calculate the proportion of each element in the target data sequence to the first data sum respectively to form a first ratio sequence; Obtain the mean of all the remaining data in the target data sequence, use the mean as the reference data for any target data, form a reference data sequence with each of the remaining data and the reference data, obtain the sum of the elements of the reference data sequence, denoted as the second data sum, and calculate the proportion of each element in the reference data sequence to the second data sum respectively to form a second ratio sequence; Obtain the cross-entropy between the first ratio sequence and the second ratio sequence, use the cross-entropy as the variable of the hyperbolic tangent function to obtain the corresponding second hyperbolic tangent value; Use the mean of the first hyperbolic tangent value and the second hyperbolic tangent value as the dissimilarity degree of any target data.

3. A mobile road surface water accumulation monitoring data analysis and processing system according to claim 1, characterized in that, In the data analysis module, obtain the local mutation degree of any target data according to the adjacent data of the target data sequence in the road surface water depth data sequence, including: In the road surface water depth data sequence, obtain the first data sequence composed of the preset first number of elements before the target data sequence. In a two-dimensional plane, construct the first data change curve of the first data sequence, where the horizontal axis of the two-dimensional plane is time and the vertical axis is the depth of the road surface water. Obtain the second data sequence composed of the preset first number of elements after the target data sequence. In a two-dimensional plane, construct the second data change curve of the second data sequence. Obtain the Spearman correlation coefficient between the first data change curve and the second data change curve, obtain the sum of the Spearman correlation coefficient and the constant 1, and perform normalization processing on the reciprocal of the sum result to obtain the first normalization result; Obtain the mean of the elements in the road surface water depth data sequence except the elements in the target data sequence, denoted as the first mean, obtain the mean of the elements in the target data sequence, denoted as the second mean, obtain the absolute value of the first difference between the first mean and the second mean, and perform normalization processing on the absolute value of the first difference to obtain the second normalization result; Use the mean of the first normalization result and the second normalization result as the local mutation degree of any target data.

4. A mobile road surface water accumulation monitoring data analysis and processing system according to claim 1, characterized in that, In the data analysis module, obtain the interference degree of each non-target data in the road surface water depth data sequence according to the interference degree of each target data, including: According to the interference degree of the preset target number of consecutive target data including the first target data in the road surface water depth data sequence, obtain the minimum interference degree, and use the minimum interference degree as the interference degree of each data before the nth data in the road surface water depth data sequence; According to the interference degree of the preset target number of consecutive target data including the last target data in the road surface water depth data sequence, obtain the minimum interference degree, and use the minimum interference degree as the interference degree of each data after the (N - n + 1)th data in the road surface water depth data sequence.

5. The mobile road surface water accumulation monitoring data analysis and processing system according to claim 1, wherein, In the filtering processing module, obtain the adaptive weight of the corresponding data according to the interference degree of each data in any sliding window, including: According to the interference degree of each data in any of the sliding windows, obtain the total interference degree, obtain the ratio between the total interference degree and the interference degree of each data, obtain the total ratio, and respectively obtain the proportion between each ratio and the total ratio as the adaptive weight of the corresponding data.

Images