Medium and high frequency monitoring data timing method, system and storage medium
Through the medium and high frequency monitoring data timing method, time weight sequence and uniformly distributed random number reflection mapping are used to solve the problem of time loss and noise of monitoring data, improve the accuracy of data analysis and the stability of the system, and are suitable for many types of medium and high frequency monitoring data.
Patent Information
- Application Number
- CN202411049724.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-01
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-08-01
AI Technical Summary
The existing monitoring data timing technology has problems such as time loss, noise and outliers, and time out of synchronization, which affects the accuracy of data analysis.
Using the medium and high frequency monitoring data timing method, by dividing time intervals and monitoring numerical intervals according to the numerical size, calculating and generating time weight sequences, standardizing time intervals, building a time weight array, and using uniformly distributed random number reflection to obtain the corresponding time of the monitoring data , ensure the accuracy and completeness of data time.
It effectively avoids the problem of time missing, improves the accuracy of data analysis and the quality of monitoring data, simplifies the calculation amount, and is suitable for various types of medium and high-frequency data such as bridge health monitoring, and improves the stability and accuracy of the system.
Smart Images

Figure CN118568506B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to a data timing method, and specifically to a medium- and high-frequency monitoring data timing method, system and storage medium. Technical Background
[0002] Structural monitoring data timing technology mainly focuses on ensuring the accuracy of the timestamp of monitoring data. In structural monitoring, the accuracy and real-time nature of data are crucial for the assessment and early warning of structural status. Through timing technology, it can ensure that data from different sources and at different time points can be effectively compared and analyzed, so as to more accurately assess the health status of the structure. With the continuous development of intelligent and automated technologies, structural monitoring data timing technology will be more and more widely used in the monitoring of various types of building structures. This technology is combined with advanced big data technologies such as the Internet of Things and artificial intelligence to achieve a more intelligent and automated structural monitoring and early warning system.
[0003] The existing monitoring data timing technology mainly involves four important steps: data collection, timestamp addition, data storage, and time series analysis, as follows:
[0004] Data collection: It is necessary to collect data of the monitored objects in real time or regularly, such as environmental indicators (such as air quality, water quality), transaction data of the financial market, etc.; Timestamp addition: While collecting data, add an accurate timestamp to each piece of data to record the time when the data was generated; Data storage: Store the data with timestamps in the database for subsequent analysis and processing; Time series analysis: Use time series analysis methods, such as ARIMA model, LSTM neural network, etc., to analyze the data and extract time-related features and information.
[0005] The literature "Han Fang, Xia Yangchun, Chen Tianchi, et al. Experimental study on damage monitoring of axially compressed timber columns based on time inversion method [J]. Experimental Technology and Management. 2023.40(11).52-57" proposes a method for structural damage monitoring based on piezoelectric ceramic sensors. This method applies an excitation signal to the driver PZT-1, and then performs time inversion on the signal received by the sensor PZT-2 in the time domain to obtain an inverted excitation signal. The inverted excitation signal is then transmitted back through the same sensor PZT-2, and the signal extracted at the driver PZT-1 is the time inversion reconstruction signal. However, the time inversion accuracy of this method is highly dependent on the measurement accuracy and usage status of the sensor. When the sensor fails or there is a network problem, data in some time periods will be missing, affecting the accuracy of the analysis. Secondly, the noise or outliers contained in the monitoring data will interfere with the results of the time series analysis. In addition, there may be time synchronization errors between different monitoring units, resulting in inaccurate timestamps of the data.
[0006] In summary, the existing monitoring data timing technology has typical problems such as time loss, noise and outliers, and time asynchrony. Summary of the invention
[0007] In response to the problems existing in the prior art, the present application proposes a method, system and storage medium for assigning timing to medium and high frequency monitoring data, which can accurately and quickly restore the time corresponding to the monitoring data and ensure the quality of the monitoring data.
[0008] In a first aspect, the present application provides a method for timing medium and high frequency monitoring data, including the following data training steps:
[0009] S1, divide several time intervals and several monitoring value intervals E according to the value size m , the time length of the time interval is T n , adjacent time intervals are continuous, n=1,…,N, m=1,…,M;
[0010] S2, the same monitoring value interval E m Within, the monitoring data corresponds to the time length T n The internal distribution is regarded as approximately uniform, and the monitoring value interval E is calculated. m The monitoring data appears in the time length T n The interval probability β mn ;
[0011] S3, standardize the time interval, and set the time length T n Compressed to a standardized time length T' n , T' n =β mn ;
[0012] S4, generate monitoring value interval E m The time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ];
[0013] S5, construct time weight array B M×N , the monitoring value interval E m The time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ] is the time weight array B M×N The m-th row element in ;
[0014] And so on, until B is generated M×N All elements:
[0015] B M × N = [ β 11 .... β 1 n .... β 1 N ……………… β m 1 ... β mn ... β mN ……………… β M 1 ... β Mn ... β MN ] , β mn ∈[0,1].
[0016] In a possible implementation manner of the first aspect, the monitoring value interval E m The monitoring data appears in the time length T n The probability of being within β mn for:
[0017] ;
[0018] C m The monitoring value interval E m Internal data volume, C mn The monitoring value interval E m In the time length T n The amount of data inside.
[0019] In a possible implementation manner of the first aspect, before generating the time weight sequence, a self-checking step is further included, which specifically includes the following contents:
[0020] For the same monitoring value interval E m , first check Is it true? If true, generate the time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ].
[0021] Furthermore, the timing method for medium and high frequency monitoring data includes the following time inversion steps:
[0022] S6, index monitoring data a corresponding to the monitoring value interval E m ;
[0023] S7, generate a uniformly distributed random number x between [0,1] corresponding to a;
[0024] S8, for the same monitoring value interval E m , determine the r standardized time lengths T' that are completely contained by the length of the interval [0,x] i , i=1,…,r,r≤N;
[0025] S9, reverse mapping the uniformly distributed random number x to obtain the corresponding time t of the monitoring data a:
[0026] ;
[0027] And so on, traverse all monitoring data until all corresponding times are generated.
[0028] In a possible implementation manner of the first aspect, step S8 specifically includes the following content:
[0029] The monitoring value interval E m The time weight sequence of m1 ,β m2 ,…,β mn ,…,β mN ], determine the maximum value of r for which the following inequality holds:
[0030] ;
[0031] The time weight sequence corresponding to the maximum r value [β m1 ,β m2 ,…,β mn ,…,β mN ] are the required r standardized time lengths.
[0032] In a second aspect, the present application provides a system for executing the medium and high frequency monitoring data timing method, comprising:
[0033] The interval division unit divides the time length T according to the value size n and several monitoring value intervals E m ;
[0034] Interval probability calculation unit, calculates the monitoring value interval E m The monitoring data appears at each time length T n The probability of being within β mn ;
[0035] Time standardization unit, standardize time interval, and convert time length T n Compressed to a standardized time length T' n , T' n =β mn ;
[0036] Time weight sequence generation unit, generates a single monitoring value interval E m In N time lengths T n The time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ];
[0037] Time weight array generation unit, constructing M monitoring value intervals E m Array B of time weights in the standardized time interval M×N .
[0038] In a possible implementation manner of the second aspect, the system further includes:
[0039] Self-check unit, check the same monitoring value interval E m Inside Is it true? If true, the time weight sequence generation unit generates the time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ].
[0040] Furthermore, the system also includes:
[0041] Data index unit, indexes the monitoring value interval E corresponding to the monitoring data a m ;
[0042] A random number generation unit generates a uniformly distributed random number x between [0,1] corresponding to a;
[0043] Standardized time search unit, for the same monitoring value interval E m , determine the r standardized time lengths T' that are completely contained by the length of the interval [0,x] i , i=1,…,r,r≤N;
[0044] The time inversion unit inversely maps the uniformly distributed random number x to obtain the corresponding time t of the monitoring data a, and traverses all the monitoring data until all the corresponding times are generated.
[0045] In a possible implementation manner of the second aspect, the standardized time search unit of the system converts the time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ] are accumulated in sequence, and the maximum r value whose sum of elements is less than or equal to the uniformly distributed random number x is found.
[0046] In a third aspect, the present application provides a computer-readable storage medium having computer program instructions stored thereon, wherein the computer program instructions implement the timing method when executed by a processor.
[0047] Compared with the prior art, the present application calculates the time weight according to the measured data to determine the corresponding time of the monitoring data, and uses uniform distribution sampling to obtain the time of all monitoring data, thereby forming time series monitoring data. The time weight array can more completely reflect the two-dimensional relationship between time and parameters. The method used to form time series data can be used to analyze the changes in monitoring data within a certain period of time, avoiding the possible problems of time loss, and providing support for the maintenance and development of bridge health monitoring data systems.
[0048] Specifically, the data timing steps of the present application are concise and clear, with low calculation amount, and the operation speed is greatly improved compared with the traditional machine learning model; the accuracy can be effectively controlled by adjusting the number of intervals, the timing quality can be guaranteed, and human intervention is convenient; the timing is directly based on the value of the monitoring data, and the intrinsic relationship between the monitoring data and time will not be lost in the process; the time corresponding to the monitoring data is inverted through the inverse mapping algorithm, which is suitable for various types of medium and high frequency monitoring data such as vehicle weight, vehicle distance, strain value, etc., and has a wide range of adaptability. BRIEF DESCRIPTION OF THE DRAWINGS
[0049] Figure 1 This is a data training flow chart of a method for timing medium and high frequency monitoring data according to an embodiment of the present application;
[0050] Figure 2 This is a time inversion flow chart of a method for timing medium and high frequency monitoring data according to an embodiment of the present application;
[0051] Figure 3 This is a flow chart of a method for timing medium and high frequency monitoring data according to an embodiment of the present application;
[0052] Figure 4 A measured vehicle weight-time probability diagram of an embodiment of the present application;
[0053] Figure 5 A time weight array diagram of an embodiment of the present application;
[0054] Figure 6 A simulated vehicle weight-time series diagram according to an embodiment of the present application. DETAILED DESCRIPTION
[0055] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings.
[0056] It should be clear that the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.
[0057] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms "a", "said" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.
[0058] In the first aspect, a method for timing medium and high frequency monitoring data provided by an embodiment of the present application includes two stages: data training and time inversion:
[0059] like Figure 1 As shown, the process of data training is as follows:
[0060] S1, divide N time intervals and M monitoring value intervals E according to the value size m , the amount of data in each interval is nearly equal; the length of the time interval is T n , adjacent time intervals are continuous, n=1,…,N, m=1,…,M;
[0061] S2, the same monitoring value interval E m Within, the monitoring data corresponds to the time length T n The internal distribution is regarded as approximately uniform, and the monitoring value interval E is calculated. m The monitoring data appears in the time length T n The interval probability β mn ;
[0062] S3, standardize the time interval, and set the time length T n Compressed to a standardized time length T' n , T' n =β mn ;
[0063] S4, generate monitoring value interval E m The time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ], this step can compress the data distribution information into a standardized numerical range;
[0064] S5, construct time weight array B M×N , the monitoring value interval E m The time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ] is the time weight array B M×N The m-th row element in ;
[0065] And so on, until B is generated M×N All elements:
[0066] B M × N = [ β 11 .... β 1 n .... β 1 N ……………… β m 1 ... β mn ... β mN ……………… β M 1 ... β Mn ... β MN ] , β mn ∈[0,1].
[0067] Specifically, in step S2, the monitoring value interval E m The monitoring data appears at each time length T n The probability of being within β mn for:
[0068] ;
[0069] C m The monitoring value interval E m Internal data volume, C mn The monitoring value interval E m In the time length T n The amount of data inside.
[0070] Preferably, before generating the time weight sequence in S4, a self-checking step is also included, which specifically includes the following contents:
[0071] For the same monitoring value interval E m , first check Is it true? If true, generate the time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ].
[0072] like Figure 2 As shown, time inversion is performed based on the above data training, and the steps are as follows:
[0073] S6, index monitoring data a corresponding to the monitoring value interval E m ;
[0074] S7, generate a uniformly distributed random number x between [0,1] corresponding to a;
[0075] S8, for the same monitoring value interval E m , determine the r standardized time lengths T' that are completely contained by the length of the interval [0,x] i , i=1,…,r,r≤N; this step uses random number x to reversely map out the data distribution;
[0076] S9, reverse mapping the uniformly distributed random number x to obtain the corresponding time t of the monitoring data a:
[0077] ;
[0078] in, is the remaining term, which indicates the length of time contained in the random number x in the r+1th time interval;
[0079] And so on, traverse all monitoring data until all corresponding times are generated.
[0080] In the above step S9, it is assumed that a1, a2, ..., a pThere are p monitoring data. According to the value of the monitoring data, the m (m=1,…,M) corresponding to the qth (q=0,1,2…,p) parameter is indexed in sequence to determine the interval E where the monitoring data is located. m , and then according to the uniform distribution of random numbers x between [0,1] q Inverse Mapping q The corresponding time interval range, where T'1~T' r Completely x q Including, add up the time lengths of the first r time intervals, and determine the proportion of the r+1th time interval by calculating the remaining term, take the corresponding time lengths, and add them up to get the total continuous time t (r=1,2...,N, m has been determined).
[0081] Specifically, the process of step S8 is as follows:
[0082] Monitoring value interval E m The time weight sequence of m1 ,β m2 ,…,β mn ,…,β mN ], determine the maximum value of r for which the following inequality holds:
[0083] ;
[0084] The time weight sequence corresponding to the maximum r value [β m1 ,β m2 ,…,β mn ,…,β mN ] are the required r standardized time lengths.
[0085] In a second aspect, the present application provides a system for a high-frequency monitoring data timing method, comprising:
[0086] The interval division unit divides the time length T according to the value size n and several monitoring value intervals E m ;
[0087] Interval probability calculation unit, calculates the monitoring value interval E m The monitoring data appears at each time length T n The probability of being within β mn ;
[0088] Time standardization unit, standardize time interval, and convert time length T n Compressed to a standardized time length T' n , T' n =β mn ;
[0089] Time weight sequence generation unit, generates a single monitoring value interval E m The time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ];
[0090] Time weight array generation unit, constructing M monitoring value intervals E m Array B of time weights in the standardized time interval M×N .
[0091] Preferably, the system also includes a self-checking unit for checking the same monitoring value interval E m Inside Is it true? If true, the time weight sequence generation unit generates a time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ].
[0092] Furthermore, the system also includes:
[0093] Data index unit, indexes the monitoring value interval E corresponding to the monitoring data a m ;
[0094] A random number generation unit generates a uniformly distributed random number x between [0,1] corresponding to a;
[0095] Standardized time search unit, for the same monitoring value interval E m , determine the r standardized time lengths T' that are completely contained by the length of the interval [0,x] i , i=1,…,r,r≤N;
[0096] The time inversion unit inversely maps the uniformly distributed random number x to obtain the corresponding time t of the monitoring data a, and traverses all the monitoring data until all the corresponding times are generated.
[0097] Among them, the standardized time search unit converts the time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ] are accumulated in sequence, and the maximum r value whose sum of elements is less than or equal to the uniformly distributed random number x is found.
[0098] It should be understood that the division of the processing units in the above system is only a division of logical functions, and in actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated. In addition, the processing units in the system can be implemented in the form of a processor calling software; for example, the system includes a processor, the processor is connected to a memory, and instructions are stored in the memory. The processor calls the instructions stored in the memory to implement any of the above methods or implement the functions of the processing units of the system, wherein the processor is a general-purpose processor, such as a central processing unit or a microprocessor, and the memory is a memory within the system or a memory outside the system.
[0099] In a third aspect, the present application further provides a computer-readable storage medium having computer program instructions stored thereon, wherein the computer program instructions, when executed by a processor, implement the timing method in the above embodiment;
[0100] For example, the above can be achieved Figure 1 and / or Figure 2 The steps of the timing method shown in .
[0101] Computer readable storage media can be tangible devices that can hold and store instructions used by instruction execution devices. Computer readable storage media can be, for example, but not limited to, electronic storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the above.
[0102] The computer program instructions used to perform the operations of the present application may be source code or object code written in any combination of one or more programming languages, and the computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server.
[0103] The following is an example of a bridge in Nanjing. Figure 3-Figure 6 , the timing scheme of this application is described in detail, and the feasibility and beneficial effects of this application are verified.
[0104] Bridge overview and data source: A bridge in Nanjing is a double-tower, three-span steel box girder suspension bridge designed for highway Class I load. The weighing accuracy of the weighing system is better than 10%, and the accuracy of traffic flow statistics is greater than 95%. When the first axle of the vehicle passes through the front coil of the piezoelectric film weighing sensor, the sensor is triggered, and the upper gantry performs license plate recognition and speed measurement. When the last axle of the vehicle passes through the rear coil of the sensor, the WIM system stops working. The data is processed and calculated by the central controller in the roadside chassis to form complete vehicle weighing information.
[0105] The larger the data volume of the timing method of this application, the better the timing effect. In order to ensure the data volume, it is advisable to use medium and high frequency monitoring data, and it is advisable to use monitoring data types with good correlation between numerical values and time.
[0106] like Figure 3 The following is a flow chart of the timing method for this example. The steps are as follows:
[0107] (1) First, divide the monitoring data interval and time interval. In this example, the monitoring data type is vehicle weight. The total number of vehicle weight parameter intervals is M = 50, and the total number of time intervals is N = 288. The length of a single vehicle weight interval is 1 kN, and the length of a single time interval is 5 minutes. Figure 4 As shown, there are several similar strip-shaped distributions, whose length and width correspond to a single monitoring data interval and time interval respectively.
[0108] (2) Calculate the probability β of each vehicle weight parameter interval appearing in different time intervals mn , determine if Figure 4 The daily vehicle weight-time probability β is shown mn It can be seen that the 14~18kN vehicle weight is almost evenly distributed throughout the day except for 720~960min, which is because most of the vehicles in this vehicle weight range are two-axle sedans that are more common throughout the day; the 37~40kN vehicle weight is approximately evenly distributed and densely distributed in 1000~1400min; it is approximately evenly distributed and sparse in 0~700min, and most of the vehicles in this vehicle weight range are small trucks with scattered travel times and less load; the 14~18kN vehicle weight is sparser in 720~960min than in other periods, and the 37~40kN vehicle weight is almost non-distributed in 700~1000min because of the limited flow measures on the bridge during this period. The number of vehicles in the remaining vehicle weight ranges in the measured data is very small, and is concentrated in a few time points, which is in line with the characteristics of relatively single types of vehicles passing through holidays.
[0109] (3) Compress the time interval 0~1400min to obtain the standardized time interval [0,1]. The length of the nth standardized time interval is T' n , T' n =β mn ; This step divides the time intervals of equal length and unequal weight according to β mn Map into time intervals of unequal length and equal weight, and then merge the time intervals with probability 0 so that the horizontal length of the interval is proportional to its weight.
[0110] (4) Generate a time weight array after self-checking; check ∑T' n =∑β mn =1 is true, if true, immediately generate the time weight sequence [β m1 ,β m2 ,…,β mn,…,β mN ], construct M monitoring value intervals E m Array B of time weights in normalized time M×N ,like Figure 5 Shown is a schematic diagram of a normalized time array.
[0111] (5) After completing the above data training, the monitoring data is timed. First, the monitoring value interval corresponding to the monitoring data is indexed, and a uniformly distributed random number between [0, 1] is generated. Then, the continuous time is inverted and the monitoring data points are traversed until all the monitoring data points are timed.
[0112] Figure 6 The vehicle weight-time series obtained by simulation is shown in Figure 2. Figure 4 and Figure 6 Features, vehicle weight-time probability β mn The simulated vehicle weight-time series has a similar strip distribution at vehicle weights of 10-20 kN and 35-40 kN. There are fewer vehicles at 720-960 min, i.e., 12-17 o'clock, and there are slightly fewer vehicles at 0-300 min, i.e., 0-5 o'clock. Trucks of 35-40 kN mostly appear at 18-12 o'clock. Therefore, it can be considered that the simulated vehicle weight-time series generation effect is better. The smaller strip distribution may be related to the measurement method and accuracy of the WIM system. We can try to improve the measurement and recording equipment. The fit between the original data and the timed data R 2 =0.9641, indicating that the timing effect is good.
[0113] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.
Claims
1. A method for timing medium and high frequency monitoring data, wherein the type of monitoring data is vehicle weight, characterized in that: The data training steps include the following: S1, divide several time intervals and several monitoring value intervals E according to the value size m , the time length of the time interval is T n , adjacent time intervals are continuous, n=1,…,N, m=1,…,M, where M represents the total number of monitoring value intervals of vehicle weight, and N represents the total number of time intervals; S2, the same monitoring value interval E m Within, the monitoring data corresponds to the time length T n The internal distribution is regarded as approximately uniform, and the monitoring value interval E is calculated. m The monitoring data appears in the time length T n The interval probability β mn ; S3, standardize the time interval, and set the time length T n Compressed to a standardized time length T' n , T' n =β mn ; S4, generate monitoring value interval E m The time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ]; S5, construct time weight array B M×N , the monitoring value interval E m The time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ] is the time weight array B M×N The m-th row element in ; And so on, until B is generated M×N All elements: ,β mn ∈[0,1]; S6, the monitoring value interval E corresponding to the index monitoring data m ; S7, generating a uniformly distributed random number x between [0,1] corresponding to the monitoring data; S8, for the same monitoring value interval E m , determine the r standardized time lengths T' that are completely contained by the length of the interval [0,x] i , i=1,…,r,r≤N; specifically includes the following contents: The monitoring value interval E m The time weight sequence of m1 ,β m2 ,…,β mn ,…,β mN ], determine the maximum value of r for which the following inequality holds: ; The time weight sequence corresponding to the maximum r value [β m1 ,β m2 ,…,β mn ,…,β mN ] are the required r standardized time lengths; S9, reverse mapping the uniformly distributed random number x to obtain the corresponding time t of the monitoring data: ; And so on, traverse all monitoring data until all corresponding times are generated.
2. The method for timing medium and high frequency monitoring data according to claim 1, characterized in that: The monitoring value interval E m The monitoring data appears in the time length T n The probability of being within β mn for: ; C m The monitoring value interval E m Internal data volume, C mn The monitoring value interval E m In the time length T n The amount of data inside.
3. The method for timing medium and high frequency monitoring data according to claim 1, characterized in that: Before generating the time weight sequence, a self-checking step is also included, which specifically includes the following contents: For the same monitoring value interval E m , first check Is it true? If true, generate the time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ].
4. A system for executing the medium and high frequency monitoring data timing method as described in any one of claims 1 to 3, characterized in that: include: The interval division unit divides the time length T according to the value size n and several monitoring value intervals E m ; Interval probability calculation unit, calculates the monitoring value interval E m The monitoring data appears at each time length T n The probability of being within β mn ; Time standardization unit, standardize time interval, and convert time length T n Compressed to a standardized time length T' n , T' n =β mn ; Time weight sequence generation unit, generates a single monitoring value interval E m In N time lengths T n The time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ]; Time weight array generation unit, constructing M monitoring value intervals E m Array B of time weights in the standardized time interval M×N ; Data index unit, indexes the monitoring data corresponding to the monitoring value interval E m ; A random number generation unit generates a uniformly distributed random number x between [0,1] corresponding to the monitoring data; Standardized time search unit, for the same monitoring value interval E m , determine the r standardized time lengths T' that are completely contained by the length of the interval [0,x] i , i=1,…,r,r≤N; Specifically, the time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ] are accumulated in sequence, and the maximum r value whose sum of elements is less than or equal to the uniformly distributed random number x is found; The time inversion unit inversely maps the uniformly distributed random number x to obtain the corresponding time t of the monitoring data, and traverses all the monitoring data until all the corresponding times are generated.
5. The system according to claim 4, characterized in that Also includes: Self-check unit, check the same monitoring value interval E m Inside Is it true? If true, the time weight sequence generation unit generates the time weight sequence [β m1 ,β m2 ,…,β mn ,…,β mN ].
6. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the timing method described in any one of claims 1 to 3 is implemented.
Citation Information
Patent Citations
Spatial interpolation method for long-span bridge monitoring data
CN104778331A
Time axis unification system and method
CN112363383A
Bridge health early warning method based on time sequence and multi-sensor fusion
CN114925518A
Monitoring data compression method and device and storage medium
CN115987291A
Space-time normalized graph convolutional neural network traffic flow prediction method and system, and medium
CN117392830A
