Structural uniform temperature rolling prediction method, device and system, and storage medium

Abstract

The invention discloses a structure uniform temperature rolling prediction method, device and system, and a storage medium. The method comprises the following steps: S1, obtaining historical structure temperature data; s2, obtaining daily average temperature and fluctuation temperature according to the historical structure temperature data; s3, modeling the daily average temperature by using a trigonometric function and an autoregressive moving average model to obtain a daily average temperature predicted value; s4, according to the fluctuation temperature, performing statistics to obtain a daily fluctuation interval of the fluctuation temperature; and S5, according to the daily average temperature prediction value and the daily fluctuation interval of the fluctuation temperature, predicting the maximum and minimum value of the uniform temperature in the next day. By adopting the technical scheme of the invention, the advanced prediction of the daily average temperature and the uniform temperature daily extreme value is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of structural health monitoring, and in particular relates to a method and device, a system, and a storage medium for predicting uniform temperature of a structure. Background Art

[0002] Temperature load is one of the most important factors in structural design, which will have a significant impact on the safety, stability and performance of the structure. Since the cross-sectional dimensions of different structures are different, the area weighted average method is usually used to obtain the uniform temperature of the structure, which is used as an indicator to measure the overall temperature load of the structure and eliminate the influence of the cross-sectional dimensions of the structure.

[0003] The uniform temperature of the structure directly affects the overall expansion and contraction of the structure. In recent years, with the acceleration of urbanization and the rapid development of infrastructure construction, large-span and super-high structures have developed rapidly. Compared with conventional structures, the uniform temperature effect of large-span and super-high structures has increased significantly. For example, the expansion and contraction deformation of the beam end caused by the uniform temperature change of the large-span bridge determines the selection of the expansion joint. In addition, the beam-rail effect caused by the uniform temperature deformation of the large-span railway bridge will also affect the internal force of the rail on the bridge and control the setting of the rail expansion and contraction adjuster. The uniform temperature deformation of the high pier of the bridge affects the smoothness of the track on the bridge, threatening the high-speed and safe operation of the high-speed railway.

[0004] The numerical simulation results have non-negligible errors and the operation is time-consuming, and it does not have advantages in terms of efficiency and accuracy.

[0005] The structural temperature monitoring system can monitor the temperature changes of the structure in real time and accurately, providing engineers with key temperature data. These data play an important role in evaluating the thermal response characteristics of the structure, estimating the deformation caused by temperature, optimizing the structural design and construction, etc. However, the real-time monitoring system cannot identify the temperature load and deformation in the future in advance. For high-speed railway bridges, when the temperature-induced deformation exceeds the standard, the high-speed train may be about to reach the bridge or already on the bridge. Therefore, it is necessary to predict the structural temperature load, realize the advance warning of the temperature-induced deformation of the structure, and improve the service safety of the engineering structure. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a method and device, system and storage medium for rolling prediction of structural uniform temperature, so as to realize advance prediction of daily average temperature and daily extreme value of uniform temperature.

[0007] To achieve the above object, the present invention adopts the following technical solution:

[0008] A structural uniform temperature rolling prediction method, comprising:

[0009] Step S1, obtaining historical structural temperature data;

[0010] Step S2, obtaining the daily average temperature and fluctuating temperature according to the historical structural temperature data;

[0011] Step S3, using a trigonometric function and an autoregressive moving average model to model the daily average temperature, and obtaining a predicted value of the daily average temperature;

[0012] Step S4, obtaining the daily fluctuation range of the fluctuating temperature according to the fluctuating temperature;

[0013] Step S5: predict the maximum value of the average temperature in the next day based on the predicted value of the daily average temperature and the daily fluctuation range of the fluctuating temperature.

[0014] Preferably, in step S1, the temperature data of each measuring point of the structure at the same time is integrated to calculate the uniform temperature of the cross section of the structure at that time.

[0015]

[0016] Among them, T ij is the uniform temperature of the structural section at measuring point i at time j, A i is the area represented by measuring point i.

[0017] The uniform temperature of the structural section at different times is calculated cyclically to obtain the historical structural temperature data.

[0018] Preferably, in step S3, the average temperature at all times of the day is averaged to obtain the daily average temperature.

[0019]

[0020] Among them, T m is the daily average temperature, T m_j is the average temperature of the cross section at time j, and n is the number of times the temperature is collected on that day;

[0021] Subtract the daily average temperature from the average temperature at each time of the day to get the fluctuating temperature.

[0022] T f =T m_j -T m

[0023] Among them, T f The temperature is fluctuating.

[0024] The present invention also provides a structural uniform temperature rolling prediction device, comprising:

[0025] An acquisition module is used to obtain historical structural temperature data;

[0026] Decomposition module, used to obtain daily average temperature and fluctuating temperature based on historical structural temperature data;

[0027] The first processing module is used to model the daily average temperature using a quasi-trigonometric function and an autoregressive moving average model to obtain a predicted value of the daily average temperature;

[0028] The second processing module is used to obtain the daily fluctuation range of the fluctuating temperature according to the fluctuating temperature;

[0029] The prediction module is used to predict the maximum value of the average temperature in the next day based on the predicted value of the daily average temperature and the daily fluctuation range of the fluctuating temperature.

[0030] As a preference, the acquisition module fuses the temperature data of each measuring point of the structure at the same moment, and calculates the uniform temperature of the structure cross section at that moment.

[0031]

[0032] Among them, T ij is the uniform temperature of the structural section at measuring point i at time j, A i is the area represented by measuring point i.

[0033] The uniform temperature of the structural section at different times is calculated cyclically to obtain the historical structural temperature data.

[0034] Preferably, the decomposition module averages the average temperature at all times of the day to obtain the daily average temperature.

[0035]

[0036] Among them, T m is the daily average temperature, T m_j is the average temperature of the cross section at time j, and n is the number of times the temperature is collected on that day;

[0037] Subtract the daily average temperature from the average temperature at each time of the day to get the fluctuating temperature.

[0038] T f =T m_j -T m

[0039] Among them, T f The temperature is fluctuating.

[0040] The present invention also provides a structural uniform temperature rolling prediction system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a structural uniform temperature rolling prediction method when executed by the processor.

[0041] The present invention also provides a storage medium, on which a computer program is stored, and the computer program executes a structural uniform temperature rolling prediction method when running.

[0042] The present invention can realize the advance prediction of the future daily average temperature extreme value only by relying on the historical monitoring data of the structure itself, and has the advantages of high efficiency and accuracy, convenience and speed, and high degree of automation. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 It is a flow chart of a method for rolling prediction of uniform temperature of a structure according to an embodiment of the present invention;

[0045] Figure 2 Schematic diagram of steel box girder measurement point layout

[0046] Figure 3 Schematic diagram comparing the prediction results of the method of the embodiment of the present invention and the deep learning method; wherein (a) is the daily average temperature change predicted by the deep learning method, (b) is the daily extreme value of the uniform temperature of the structure predicted by the deep learning method, (c) is the daily average temperature change predicted by the method of the embodiment of the present invention, and (d) is the daily extreme value of the uniform temperature of the structure predicted by the method of the embodiment of the present invention;

[0047] Figure 4 Schematic diagram of the comparison of prediction errors between the method according to the embodiment of the present invention and the deep learning method. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] Embodiment 1:

[0051] like Figure 1 As shown, an embodiment of the present invention provides a method for predicting uniform temperature of a structure, including:

[0052] Step S1, obtaining historical structural temperature data;

[0053] Step S2, obtaining the daily average temperature and fluctuating temperature according to the historical structural temperature data;

[0054] Step S3, using a trigonometric function and an autoregressive moving average model to model the daily average temperature, and obtaining a predicted value of the daily average temperature;

[0055] Step S4, obtaining the daily fluctuation range of the fluctuating temperature according to the fluctuating temperature;

[0056] Step S5: predict the maximum value of the average temperature in the next day based on the predicted value of the daily average temperature and the daily fluctuation range of the fluctuating temperature.

[0057] As an implementation method of the present invention, in step S1, a typical cross section of the structure is selected to arrange a temperature field monitoring system. The monitoring area of ​​a single measuring point is reasonably and evenly divided according to the size and shape characteristics of the cross section, and a temperature sensor is arranged in the center of each divided monitoring area to monitor the real-time changes in the temperature field of the cross section of the structure; the temperature data acquisition instrument collects and stores the temperature sensor data of each measuring point in the structure at a certain frequency.

[0058] The temperature data of each measuring point of the structure at the same time are integrated, and the uniform temperature of the structure section at that time is calculated according to the following formula:

[0059]

[0060] Among them, T ij is the uniform temperature of the structural section at measuring point i at time j, A i is the area represented by measuring point i.

[0061] By cyclically calculating the uniform temperature of the structural section at different times according to the above formula, the historical structural temperature data for a period of time can be finally obtained.

[0062] As an implementation method of an embodiment of the present invention, in step S2, the historical structural temperature data is decomposed to obtain the daily average temperature and the fluctuating temperature; the daily average temperature is the annual fluctuation item, which reflects the annual variation law of the uniform temperature, and the fluctuating temperature is the random fluctuation item, which reflects the random variation law of the uniform temperature affected by the time-varying atmospheric environment such as wind, radiation, and shading.

[0063] The average temperature at all times of the day is averaged to eliminate the time fluctuation term and obtain the daily average temperature. The calculation formula is as follows:

[0064]

[0065] Among them, T mis the daily average temperature, T m_j is the average temperature of the cross section at time j, and n is the number of times the temperature is collected on that day.

[0066] Subtract the daily average temperature from the average temperature at each time of the day to eliminate the annual trend term in the time series and obtain the momentary fluctuation temperature. The calculation formula is as follows:

[0067] T f =T m_j -T m

[0068] Among them, T f The temperature is fluctuating.

[0069] As an implementation of the embodiment of the present invention, step S3 is specifically:

[0070] The trigonometric function is used to fit the annual variation model of the daily average temperature:

[0071]

[0072] Among them, T m_F is the time domain function of the daily average temperature; t is the date ordinal number, such as January 1st corresponds to t=1, January 2nd corresponds to t=2, and so on; T av is the annual average of daily average temperature; T am is the annual fluctuation range of daily average temperature; T is the fluctuation period of daily average temperature, in days; is the phase, indicating the date when the daily average temperature reaches the annual minimum.

[0073] After fitting the daily average temperature of the structure using a quasi-trigonometric function, the corresponding fitting residuals can be further separated. Based on the historical correlation between the fitting residual time series data, the ARMA (p, q) model is used to model the daily average temperature fitting residuals. The expression of the ARMA (p, q) model is:

[0074] y t =c+φ1y t-1 +φ2y t-2 +...+φ p y t-p +ε t +θ1ε t-1 +θ2ε t-2 +...+θ q ε t-q

[0075] Among them, y t is the predicted value of the daily average temperature fitting residual; t-i (i=1,2,...,p) is the observed data; p is the order of AR; q is the order of MA; c is the mean of the observed data; φi ,θ i is the model parameter; ε t is white noise, ε t ~N(0,σ 2 ). The modeling steps of the ARMA model are as follows:

[0076] ①: Use autocorrelation function (ACF) and partial autocorrelation function (PACF) to perform correlation analysis on the data. The expression of PACF is:

[0077]

[0078] Where, h is the time interval; For x t In a given {x t-1 ,x t-2 ,...,x t-h+1 The optimal linear fit of} is expressed as:

[0079]

[0080] Sample-based partial autocorrelation function The calculation formula is:

[0081]

[0082] Among them, φ hh is the partial autocorrelation function; is the sample autocorrelation function; h is the time lag.

[0083] ②: According to the correlation analysis, the model order is initially estimated;

[0084] ③: Use maximum likelihood estimation to fit the model parameters;

[0085] ④: Use the error ACF diagram and Ljung-box test to test the goodness of fit of the model. If the model error is white noise, it means that the model fits well and can pass the test; if the model error is not white noise, it means that the model fits poorly and cannot pass the test, and the order needs to be re-determined.

[0086] The expression of Ljung-box statistic is:

[0087]

[0088] Among them, Q is the Ljung-box statistic, which follows a chi-square distribution with hq degrees of freedom. Therefore, if the p-value corresponding to Q is less than the significance level α, the error is not white noise. If the p-value is greater than the significance level α, the error can be considered to be white noise. q is the number of model parameters. ; n is the sample size; is the autocorrelation coefficient between the residuals with time interval k; h is the total number of test time intervals, and its value is [1, +∞].

[0089] ⑤: Use the Akaike Information Criterion (AIC) to select the final model. The Akaike Information Criterion shows that the goodness of fit of the model can be measured by the sum of the model's fitting accuracy and the model's complexity. The AIC expression is:

[0090]

[0091] Where k is the number of model parameters; n is the number of training samples; is the mean square estimation error when the number of model parameters is k, and the expression is;

[0092]

[0093] Among them, SSE(k) is the residual sum of squares when the number of model parameters is k, and the expression of the residual sum of squares is:

[0094] Final daily average temperature forecast value T m_t =T m_t =T m_F +ARMA(p,q).

[0095] As an implementation method of the present invention, in step S4, the fluctuating temperature is counted to obtain a daily fluctuation range T with a probability guarantee. range Specifically: According to the fluctuating temperature data, the daily extreme value empirical probability distribution function is calculated. The daily minimum value of the fluctuating temperature with a probability of exceeding less than 5% is determined to be T f_0.05 The maximum value of the fluctuating temperature day with a probability of exceeding less than 5% is T f_0.95 Then the daily maximum and minimum values ​​covering the 95% guarantee rate of historical data can form a fluctuation range [T f_0.05 ,T f_0.95 ]. Similarly, we can take the daily maximum value of the fluctuation temperature with a probability of exceeding less than 1% on the empirical distribution function, and obtain the fluctuation range [T f_0.01 ,T f_0.99 ] can cover the daily maximum and daily minimum values ​​of historical data with a 99% guaranteed probability.

[0096] As an implementation of the present invention, in step S5, according to the daily average temperature forecast value T m_t and the daily fluctuation range of the fluctuating temperature Trange The maximum average temperature T in the next day can be predicted by the following formula: u ;

[0097] T u =T m_t +T range

[0098] Taking the main beam of a long-span bridge as an example, the technical solution of the embodiment of the present invention is described in detail in conjunction with the accompanying drawings:

[0099] (1) Structural measurement point arrangement and real-time signal acquisition

[0100] like Figure 2 As shown in the figure, temperature measuring points E1-E12, M1-M12, W1-W12, and T1-T7 are arranged horizontally and vertically in the cross section of the steel box girder of a large-span bridge to monitor the temperature field data at different positions of the steel box girder cross section; a data acquisition instrument is used to collect and store the temperature sensor data of each measuring point at a certain frequency, and a wireless transmission module is used to realize long-distance real-time transmission of multi-source temperature data, and the collected multi-source sensor data is transmitted to the cloud platform.

[0101] (2) Data update and calculation prediction

[0102] The temperature data of all measuring points of the structure in the past year stored on the cloud platform are called in real time by the calculation and analysis module. The overall variation law model of the daily average temperature of the steel box girder is fitted based on the embodiment of the present invention as follows:

[0103]

[0104] The ARMA prediction model of the daily average temperature fitting residual of the steel box girder is obtained as follows:

[0105] y t =0.63y t-1 +0.38ε t-1 +ε t

[0106] Steel box girder daily average temperature prediction model:

[0107]

[0108] According to the historical fluctuation temperature data, the daily extreme value empirical probability distribution function is statistically analyzed to determine the fluctuation temperature extreme value distribution interval with a certain exceedance probability. Combined with the daily average temperature forecast value, the final uniform temperature daily maximum value forecast result is obtained, such as Figure 3 shown.

[0109] Comparison of the prediction results of the structural uniform temperature rolling prediction method and the deep learning method (DNN) in the embodiment of the present invention Figure 3 , Figure 4As shown in Table 1; it can be seen that the prediction error of the method of the present invention and the prediction envelope rate of the daily average temperature extreme value are basically consistent with the deep learning method, but the calculation stability and efficiency of the method of the embodiment of the present invention are significantly higher than the DNN method. Core TM On a laptop computer with i7-7700HR CPU @ 2.80GHz and 16G memory, it takes only 0.3s to calculate the prediction result using the method of the present invention, while it takes 18 minutes to obtain the final prediction result by training the DNN model once. Therefore, the method of the embodiment of the present invention better predicts the uniform temperature extreme value of the structure and has higher calculation efficiency.

[0110] Table 1

[0111]

[0112] The above results show that the invented prediction method can be well used for real-time monitoring and dynamic prediction of uniform temperature of structures.

[0113] Embodiment 2:

[0114] The embodiment of the present invention further provides a structural uniform temperature rolling prediction device, comprising:

[0115] An acquisition module is used to obtain historical structural temperature data;

[0116] Decomposition module, used to obtain daily average temperature and fluctuating temperature based on historical structural temperature data;

[0117] The first processing module is used to model the daily average temperature using a quasi-trigonometric function and an autoregressive moving average model to obtain a predicted value of the daily average temperature;

[0118] The second processing module is used to obtain the daily fluctuation range of the fluctuating temperature according to the fluctuating temperature;

[0119] The prediction module is used to predict the maximum value of the average temperature in the next day based on the predicted value of the daily average temperature and the daily fluctuation range of the fluctuating temperature.

[0120] As an implementation method of the embodiment of the present invention, the acquisition module integrates the temperature data of each measuring point of the structure at the same time, calculates the uniform temperature of the structure cross section at that time,

[0121]

[0122] Among them, T ij is the uniform temperature of the structural section at measuring point i at time j, A i is the area represented by measuring point i.

[0123] The uniform temperature of the structural section at different times is calculated cyclically to obtain the historical structural temperature data.

[0124] As an implementation method of the embodiment of the present invention, the decomposition module averages the uniform temperature at all times of the day to obtain the daily average temperature.

[0125]

[0126] Among them, T m is the daily average temperature, T m_j is the average temperature of the cross section at time j, and n is the number of times the temperature is collected on that day;

[0127] Subtract the daily average temperature from the average temperature at each time of the day to get the fluctuating temperature.

[0128] T f =T m_j -T m

[0129] Among them, T f The temperature is fluctuating.

[0130] Embodiment 3:

[0131] An embodiment of the present invention further provides a structural uniform temperature rolling prediction system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a structural uniform temperature rolling prediction method when executed by the processor.

[0132] Embodiment 4:

[0133] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and the computer program executes a rolling prediction method for uniform temperature of a structure when running.

[0134] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for predicting uniform temperature of a structure, characterized in that: include: Step S1, obtaining historical structural temperature data; Step S2, obtaining the daily average temperature and the fluctuating temperature according to the historical structural temperature data; Step S3, using a quasi-trigonometric function and an autoregressive moving average model to model the daily average temperature, and obtaining a predicted value of the daily average temperature; Step S4, according to the fluctuating temperature, statistically obtain the daily fluctuation range of the fluctuating temperature; Step S5: predict the maximum value of the average temperature in the next day based on the predicted value of the daily average temperature and the daily fluctuation range of the fluctuating temperature.

2. The method for predicting the uniform temperature of a structure according to claim 1, characterized in that: In step S1, the temperature data of each measuring point of the structure at the same time are integrated to calculate the uniform temperature of the structure cross section at that time. Among them, T ij is the uniform temperature of the structural section at measuring point i at time j, A i is the area represented by measuring point i. The uniform temperature of the structural section at different times is calculated cyclically to obtain the historical structural temperature data.

3. The method for predicting uniform temperature of a structure according to claim 2, characterized in that: In step S3, the average temperature at all times of the day is averaged to obtain the daily average temperature. Among them, T m is the daily average temperature, T m_j is the average temperature of the cross section at time j, and n is the number of times the temperature is collected on that day; Subtract the daily average temperature from the average temperature at each time of the day to get the fluctuating temperature. T f =T m_j -T m Among them, T f The temperature is fluctuating.

4. A structural uniform temperature rolling prediction device, characterized in that: include: An acquisition module is used to obtain historical structural temperature data; Decomposition module, used to obtain daily average temperature and fluctuating temperature based on historical structural temperature data; The first processing module is used to model the daily average temperature using a quasi-trigonometric function and an autoregressive moving average model to obtain a predicted value of the daily average temperature; The second processing module is used to obtain the daily fluctuation range of the fluctuating temperature according to the fluctuating temperature; The prediction module is used to predict the maximum value of the average temperature in the next day based on the predicted value of the daily average temperature and the daily fluctuation range of the fluctuating temperature.

5. The structural uniform temperature rolling prediction device according to claim 4, characterized in that: The acquisition module integrates the temperature data of each measuring point of the structure at the same time and calculates the uniform temperature of the structure section at that time. Among them, T ij is the uniform temperature of the structural section at measuring point i at time j, A i is the area represented by measuring point i. The uniform temperature of the structural section at different times is calculated cyclically to obtain the historical structural temperature data.

6. The structural uniform temperature rolling prediction device according to claim 5, characterized in that: The decomposition module averages the average temperature at all times of the day to obtain the daily average temperature. Among them, T m is the daily average temperature, T m_j is the average temperature of the cross section at time j, and n is the number of times the temperature is collected on that day; Subtract the daily average temperature from the average temperature at each time of the day to get the fluctuating temperature. T f =T m_j -T m Among them, T f The temperature is fluctuating.

7. A structural uniform temperature rolling prediction system, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the method for rolling prediction of uniform temperature of a structure as described in any one of claims 1 to 3 is executed.

8. A storage medium, characterized in that: The storage medium stores a computer program, which, when running, executes the method for rolling prediction of uniform temperature of a structure according to any one of claims 1 to 3.

Images