Prediction method of stress monitoring data and related device

Abstract

The invention discloses a stress monitoring data prediction method and a related device, and relates to the field of stress monitoring, and the method comprises the steps: obtaining n pieces of stress monitoring actual data, and taking each randomly generated smoothing coefficient as a current smoothing coefficient; for each current smoothing coefficient, performing secondary exponential smoothing prediction on the first n-m pieces of stress monitoring actual data based on the current smoothing coefficient to obtain the last m pieces of stress monitoring prediction data; determining a target smoothing coefficient based on the last m pieces of stress monitoring prediction data and the last m pieces of stress monitoring actual data; for each current smoothing coefficient, determining an optimized smoothing coefficient based on the current smoothing coefficient, the target smoothing coefficient and a preset inertia weight; by continuously optimizing the smoothing coefficient until the preset number of iterations is reached, the selection of the smoothing coefficient is more objective and scientific, and secondary exponential smoothing prediction is performed based on the optimized target smoothing coefficient, so that the accuracy of predicting stress monitoring data is improved.

Description

Technical Field

[0001] The present application relates to the field of stress monitoring, and in particular to a method for predicting stress monitoring data and related devices. Background Art

[0002] In industrial equipment and mechanical systems, stress monitoring data plays a vital role. These stress data are collected in real time through high-precision sensors, covering the stress conditions of the equipment under various working conditions. They are not only the basis for evaluating the health status of the equipment and optimizing maintenance strategies, but also the key to ensuring the long-term stable operation of the system. Therefore, it is particularly important to predict stress monitoring data.

[0003] At present, stress monitoring data is mainly predicted by the quadratic exponential smoothing method, which is the most commonly used method in time series prediction.

[0004] However, the smoothing coefficients in the quadratic exponential smoothing method are usually randomly generated or manually adjusted based on experience, which may lead to inaccurate prediction results. Summary of the invention

[0005] In view of the above problems, the present application provides a stress monitoring data prediction method and related devices to improve the accuracy of predicted stress monitoring data. The specific scheme is as follows:

[0006] The first aspect of the present application provides a method for predicting stress monitoring data, comprising:

[0007] Obtaining n pieces of actual stress monitoring data, and dividing the n pieces of actual stress monitoring data into first nm pieces of actual stress monitoring data and last m pieces of actual stress monitoring data, where m is a preset number of avoidance time nodes, m≥1 and n>m;

[0008] Randomly generate a plurality of smoothing coefficients based on a preset range, and use each randomly generated smoothing coefficient as a current smoothing coefficient;

[0009] For each of the current smoothing coefficients, a secondary exponential smoothing prediction is performed on the first nm stress monitoring actual data based on the current smoothing coefficient to obtain the next m stress monitoring prediction data;

[0010] Determining a target smoothing coefficient based on the next m pieces of stress monitoring prediction data and the next m pieces of stress monitoring actual data;

[0011] For each of the current smoothing coefficients, determining an optimized smoothing coefficient based on the current smoothing coefficient, the target smoothing coefficient and a preset inertia weight;

[0012] The optimized smoothing coefficients are respectively used as the current smoothing coefficients, and the step of performing a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficients for each of the current smoothing coefficients to obtain the next m stress monitoring prediction data is returned until the number of returns reaches a preset number of iterations to obtain the optimized target smoothing coefficients;

[0013] Based on the optimized target smoothing coefficient, a secondary exponential smoothing prediction is performed on the n stress monitoring actual data to obtain the next k stress monitoring prediction data of the n stress monitoring actual data, where k≥1.

[0014] In a possible implementation, for each of the current smoothing coefficients, performing a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficient to obtain the next m stress monitoring prediction data includes:

[0015] For each of the current smoothing coefficients, taking the first nm pieces of stress monitoring actual data as first target stress monitoring actual data;

[0016] Performing a secondary exponential smoothing prediction on the first target stress monitoring actual data based on the current smoothing coefficient to obtain a subsequent stress monitoring prediction data of the first target stress monitoring actual data;

[0017] The first target stress monitoring actual data and the stress monitoring prediction data after the first target stress monitoring actual data are used as the first target stress monitoring actual data, and the step of performing a secondary exponential smoothing prediction on the first target stress monitoring actual data based on the current smoothing coefficient to obtain the stress monitoring prediction data after the first target stress monitoring actual data is returned to be executed, until the number of returns reaches (m-1) times, and the last m stress monitoring prediction data are obtained.

[0018] In a possible implementation, determining a target smoothing coefficient based on the next m pieces of stress monitoring prediction data and the next m pieces of stress monitoring actual data includes:

[0019] For each of the last m stress monitoring prediction data, using a Hube loss function to compare the last m stress monitoring prediction data with the last m stress monitoring actual data to obtain a set of Hube losses, wherein the set of Hube losses includes m Hube losses;

[0020] A target group of Hube losses is determined from each group of the Hube losses, and the current smoothing coefficient corresponding to the target group of Hube losses is determined as the target smoothing coefficient, wherein the sum of the m Hube losses in the target group of Hube losses is the smallest.

[0021] In a possible implementation, the calculation process of the sum of the m Hube losses includes:

[0022] Obtaining m date information of m pieces of actual stress monitoring data corresponding to the m Hubbe losses;

[0023] Sorting the m date information from early to late, and allocating m preset loss weights to the m Hube losses corresponding to the sorted m date information in ascending order;

[0024] The products of each of the Hubbe losses and the loss weights corresponding to the Hubbe losses are added together to obtain the sum of the m Hubbe losses.

[0025] In a possible implementation, performing secondary exponential smoothing prediction on the n pieces of actual stress monitoring data based on the optimized target smoothing coefficient to obtain subsequent k pieces of stress monitoring prediction data of the n pieces of actual stress monitoring data includes:

[0026] Using the n pieces of stress monitoring actual data as second target stress monitoring actual data;

[0027] Performing a secondary exponential smoothing prediction on the second target stress monitoring actual data based on the optimized target smoothing coefficient to obtain the next stress monitoring prediction data of the second target stress monitoring actual data;

[0028] The second target stress monitoring actual data and the stress monitoring prediction data after the second target stress monitoring actual data are used as the second target stress monitoring actual data, and the step of performing a quadratic exponential smoothing prediction on the second target stress monitoring actual data based on the optimized target smoothing coefficient to obtain the stress monitoring prediction data after the second target stress monitoring actual data is returned to, until the number of returns reaches (k-1) times, and the k stress monitoring prediction data after the n stress monitoring actual data are obtained.

[0029] In a possible implementation, after performing secondary exponential smoothing prediction on the n pieces of actual stress monitoring data based on the optimized target smoothing coefficient to obtain the next k pieces of predicted stress monitoring data of the n pieces of actual stress monitoring data, the method further includes:

[0030] Obtaining the next k pieces of stress monitoring actual data of the n pieces of stress monitoring actual data;

[0031] Calculate the difference between the next k pieces of actual stress monitoring data of the n pieces of actual stress monitoring data and the next k pieces of predicted stress monitoring data of the n pieces of actual stress monitoring data;

[0032] Based on the difference and a preset difference, it is determined whether a predicted abnormality occurs.

[0033] A second aspect of the present application provides a stress monitoring data prediction method and device, comprising:

[0034] An acquisition unit, used for acquiring n pieces of actual stress monitoring data, and dividing the n pieces of actual stress monitoring data into first nm pieces of actual stress monitoring data and last m pieces of actual stress monitoring data, where m is a preset number of avoidance time nodes, m≥1 and n>m;

[0035] A first determining unit, configured to randomly generate a plurality of smoothing coefficients based on a preset range, and use each randomly generated smoothing coefficient as a current smoothing coefficient;

[0036] A first prediction unit is used for performing a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficient for each of the current smoothing coefficients to obtain the next m stress monitoring prediction data;

[0037] A second determination unit is used to determine a target smoothing coefficient based on the next m pieces of stress monitoring prediction data and the next m pieces of stress monitoring actual data;

[0038] A third determining unit, configured to determine, for each of the current smoothing coefficients, an optimized smoothing coefficient based on the current smoothing coefficient, the target smoothing coefficient and a preset inertia weight;

[0039] a return unit, configured to respectively use each of the optimized smoothing coefficients as the current smoothing coefficient, and return to execute the step of performing a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficient for each of the current smoothing coefficients to obtain the next m stress monitoring prediction data, until the number of returns reaches a preset number of iterations, and the optimized target smoothing coefficient is obtained;

[0040] The second prediction unit is used to perform secondary exponential smoothing prediction on the n stress monitoring actual data based on the optimized target smoothing coefficient to obtain subsequent k stress monitoring prediction data of the n stress monitoring actual data, where k≥1.

[0041] A third aspect of the present application provides a computer program product, comprising computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements the stress monitoring data prediction method of the first aspect or any implementation of the first aspect.

[0042] A fourth aspect of the present application provides an electronic device, comprising at least one processor and a memory connected to the processor, wherein:

[0043] The memory is used to store computer programs;

[0044] The processor is used to execute the computer program so that the electronic device can implement the stress monitoring data prediction method of the first aspect or any implementation manner of the first aspect.

[0045] A fifth aspect of the present application provides a computer storage medium, which carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement the stress monitoring data prediction method of the above-mentioned first aspect or any implementation method of the first aspect.

[0046] By means of the above technical scheme, the present application provides a prediction method and related device for stress monitoring data, the method comprising: obtaining n actual stress monitoring data, randomly generating multiple smoothing coefficients based on a preset range, and taking each randomly generated smoothing coefficient as a current smoothing coefficient; for each current smoothing coefficient, performing a secondary exponential smoothing prediction on the first nm actual stress monitoring data based on the current smoothing coefficient to obtain the next m stress monitoring prediction data, where m is the preset number of avoidance time nodes, m≥1 and n>m; determining a target smoothing coefficient based on the next m stress monitoring prediction data and the next m actual stress monitoring data; for each current smoothing coefficient, determining an optimized smoothing coefficient based on the current smoothing coefficient, the target smoothing coefficient and a preset inertia weight; taking each optimized smoothing coefficient as the current smoothing coefficient, and returning to perform a secondary exponential smoothing prediction on the first nm actual stress monitoring data based on the current smoothing coefficient to obtain the next m stress monitoring prediction data, until the number of returns reaches the preset number of iterations to obtain the optimized target smoothing coefficient; performing a secondary exponential smoothing prediction on the n actual stress monitoring data based on the optimized target smoothing coefficient. This scheme continuously optimizes the smoothing coefficient until the preset number of iterations is reached, and introduces inertia weight to make the selection of smoothing coefficient more objective and scientific, avoiding the limitations of directly using randomly generated smoothing coefficients and manually adjusting smoothing coefficients, which is conducive to improving the accuracy of predicted stress monitoring data. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0048] Figure 1 A schematic diagram of a flow chart of a method for predicting stress monitoring data provided in an embodiment of the present application;

[0049] Figure 2A schematic diagram of comparison between monitoring prediction data corresponding to a smoothing coefficient and actual stress monitoring data provided in an embodiment of the present application;

[0050] Figure 3 A schematic diagram of comparison between monitoring prediction data and actual stress monitoring data corresponding to another smoothing coefficient provided in an embodiment of the present application;

[0051] Figure 4 A schematic diagram of the change of the Hube loss corresponding to a smoothing coefficient provided in an embodiment of the present application;

[0052] Figure 5 A schematic diagram of the change of the Hube loss corresponding to another smoothing coefficient provided in an embodiment of the present application;

[0053] Figure 6 A schematic diagram of a change in an optimized smoothing coefficient corresponding to a smoothing coefficient provided in an embodiment of the present application;

[0054] Figure 7 A schematic diagram of changes in an optimized smoothing coefficient corresponding to another smoothing coefficient provided in an embodiment of the present application;

[0055] Figure 8 A schematic diagram of the structure of a stress monitoring data prediction device provided in an embodiment of the present application;

[0056] Fig. 9 A schematic diagram of the hardware structure of a stress monitoring data prediction device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0057] The following describes the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. The terms used in the implementation method section of the present application are only used to explain the specific embodiments of the present application, and are not intended to limit the present application.

[0058] The embodiments of the present application are described below in conjunction with the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0059] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and need not be used to describe a specific order or sequential order. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances, which is only to describe the distinction mode adopted by the objects of the same attributes when describing in the embodiments of the present application. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, so that the process, method, system, product or equipment comprising a series of units need not be limited to those units, but may include other units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0060] In order to improve the accuracy of predicted stress monitoring data, the present application provides a method for predicting stress monitoring data. The following further describes the method for predicting stress monitoring data provided by the present application in detail in conjunction with the accompanying drawings and specific implementation methods.

[0061] Please refer to the attached Figure 1 , Figure 1 A flow chart of a method for predicting stress monitoring data provided in an embodiment of the present application. The method may include the following steps:

[0062] Step S101: obtaining n pieces of actual stress monitoring data, and dividing the n pieces of actual stress monitoring data into first nm pieces of actual stress monitoring data and last m pieces of actual stress monitoring data, where m is a preset number of avoidance time nodes, m≥1 and n>m.

[0063] In this application, a connection can be established with a database storing actual stress monitoring data, which usually involves configuration of database connection parameters, such as database type, address, port, user name and password, etc. Once the connection is successful, the required n (such as 35) actual stress monitoring data are extracted from the database according to the query conditions (such as time range, monitoring point number, etc.).

[0064] Furthermore, due to sensor failure, data transmission errors, etc., there may be missing values ​​in the extracted stress monitoring data. In order to maintain the integrity of the data, the mean of the data before and after the missing value can be used for filling. In order to remove the noise in the extracted stress monitoring data and improve the accuracy of the data, the normal distribution 3σ principle can be used for denoising. In order to compare and calculate data of different dimensions on the same scale, the data can be scaled to a specific range (such as between 0 and 1) according to the normalization formula.

[0065] In the preprocessed data set, when the actual stress monitoring data is relatively stable or has an obvious trend, m can be taken to be larger, otherwise, m can be taken to be smaller. This is to avoid inaccurate prediction results due to large data fluctuations or the influence of recent events. Therefore, the first nm (such as 33) actual stress monitoring data can be used as training data, and the last m (such as 2) data can be used as verification data to evaluate the prediction accuracy of the model. It should be noted that the value of m cannot be too large or too small, and needs to be reasonable. If m is too small, the verification data may not be sufficient to accurately evaluate the prediction ability of the model. If m is too large, it may lead to insufficient training data and affect the training effect of the model. m can also be understood as the number of time nodes to be avoided.

[0066] Step S102: randomly generating a plurality of smoothing coefficients based on a preset range, and taking each randomly generated smoothing coefficient as a current smoothing coefficient.

[0067] It should be noted that this process utilizes the Great Cane Rat Algorithm, which does not require a large amount of historical data, and can still effectively perform long-term predictions when data is scarce, and adapt to various drastic environmental conditions. The Great Cane Rat Algorithm is a swarm intelligence optimization algorithm that can optimize search and optimization problems by simulating the intelligent behavior of great cane rats in the process of foraging. The initialization process of the Great Cane Rat Algorithm includes setting parameters such as the range of population individuals, the number of populations, and the number of iterations. Among them, the range of population individuals is equivalent to the preset range of the smoothing coefficient in this embodiment, and the preset range can be determined based on experience, data characteristics or algorithm requirements, for example. The number of populations is equivalent to the number of smoothing coefficients in this embodiment. The number of iterations refers to the total number of times or generations that the algorithm runs, which determines the depth and breadth of the algorithm's exploration in the search space, and the possibility of finding the optimal solution.

[0068] In the present application, multiple (eg, 200) smoothing coefficients are randomly generated based on a preset range (eg, [0.0001, 0.9999]). The random generation process ensures a variety of smoothing coefficients, and each randomly generated smoothing coefficient is used as the current smoothing coefficient for analysis and prediction.

[0069] Among them, the smoothing coefficient is an important parameter in the fields of time series analysis, signal processing or data smoothing, and is usually used to control the degree of the smoothing process, that is, how to strike a balance between retaining the original characteristics of the data and removing noise. When using exponential smoothing methods (such as quadratic exponential smoothing) for time series forecasting, the smoothing coefficient determines the degree of influence of historical data on the current forecast value. A higher smoothing coefficient means more emphasis on recent data, while a lower smoothing coefficient means more emphasis on the overall trend of historical data.

[0070] Step S103: for each current smoothing coefficient, perform a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficient to obtain the next m stress monitoring prediction data.

[0071] In the present application, firstly, for each current smoothing coefficient, the first nm stress monitoring actual data are used as the first target stress monitoring actual data. Then, based on the current smoothing coefficient, a second exponential smoothing prediction is performed on the first target stress monitoring actual data to obtain the last stress monitoring prediction data of the first target stress monitoring actual data. Finally, the first target stress monitoring actual data and the last stress monitoring prediction data of the first target stress monitoring actual data are used as the first target stress monitoring actual data, and the step of performing a second exponential smoothing prediction on the first target stress monitoring actual data based on the current smoothing coefficient to obtain the last stress monitoring prediction data of the first target stress monitoring actual data is returned until the number of returns reaches (m-1) times to obtain the last m stress monitoring prediction data.

[0072] Specifically, using the currently selected smoothing coefficient and the first target stress monitoring actual data, a secondary exponential smoothing prediction is performed to obtain the stress monitoring prediction data after the first target stress monitoring actual data. The newly obtained stress monitoring prediction data is updated to the existing data set (i.e., the first target stress monitoring actual data). Using the updated data set (i.e., the first target stress monitoring actual data and the stress monitoring prediction data after the first target stress monitoring actual data) and the current smoothing coefficient, a secondary exponential smoothing prediction is performed again to obtain the next stress monitoring prediction data. Repeat the above steps until the number of returns reaches (m-1) times, then stop the prediction, and obtain the next m stress monitoring prediction data. The advantage of this method is that it can adapt to data changes more flexibly, and each prediction is based on the latest data set, thereby improving the accuracy of the prediction. At the same time, this method can also better reflect the trend and periodic changes of the data. However, its disadvantage is that the amount of calculation is large, and multiple predictions and data updates are required.

[0073] As another possible implementation method, there is no need to generate prediction data one by one, but directly use the quadratic exponential smoothing method based on all available historical data (the first nm stress monitoring actual data) to predict the next m stress monitoring prediction data at one time. The advantage of this method is that it is highly efficient and can obtain all the required prediction data at one time. However, it may also lead to a lower accuracy of the prediction results due to factors such as the complexity of the model and the nonlinearity of the data.

[0074] Among them, the first exponential smoothing method can refer to the following formula:

[0075] S t (1) =a*yt +(1-a)*S t-1 (1)

[0076] In the formula, S t (1) represents the exponential smoothing value of period t; y t represents the actual value of period t; S t-1 (1) represents the first exponential smoothing value of period t-1; a represents the smoothing coefficient.

[0077] If only from y 1 The starting data, then the method to determine the initial value is

[0078] 1) Take S 0 Equal to y 1 ;

[0079] 2) After accumulating some data, take S 1 Equal to the simple arithmetic mean of the previous data.

[0080] When there is no obvious trend change in the time series, the smoothing coefficient can be used for prediction. The prediction formula is:

[0081] y t+1 '=a*y t +(1-a)*y t '

[0082] In the formula, y t+1 ' represents the forecast value of period t+1, that is, the smoothed value S of period t t (1) ;y t ' represents the forecast value of period t, that is, the smoothed value S of period t-1 t-1 (1) .

[0083] The degree of lag deviation of the exponential smoothing value sequence decreases with the increase of the smoothing coefficient, but when the change of the time series shows a linear trend, there will still be obvious lag deviation when using the first exponential smoothing method for prediction. Therefore, it is also necessary to make corrections. The calculation formula of the second exponential smoothing method based on the first exponential smoothing is:

[0084] S t (2) =a*S t (1) +(1-a)*S t-1 (2)

[0085] In the formula, S t (2) represents the quadratic exponential smoothing value of period t; S t(1) represents the exponential smoothing value of period t; S t-1 (2) Represents the quadratic exponential smoothing value of period t-1.

[0086] The prediction formula of the quadratic exponential smoothing method is:

[0087] y t+T '=a t +b t *T

[0088] S t (2) =a*S t (1) +(1-a)*S t-1 (2)

[0089] a t =2S t (1) -S t (2)

[0090] b t =a / (1-a)(S t (1) S t (2) )

[0091] In the formula, y t+T ' represents the predicted value of period t+T; T represents the backward prediction time.

[0092] Step S104: Determine a target smoothing coefficient based on the next m pieces of stress monitoring prediction data and the next m pieces of stress monitoring actual data.

[0093] In the present application, firstly, for each of the last m stress monitoring prediction data, the last m stress monitoring prediction data are compared with the last m stress monitoring actual data using the Hube loss function to obtain a group of Hube losses, wherein a group of Hube losses includes m Hube losses. Then, a target group of Hube losses is determined from each group of Hube losses, and the current smoothing coefficient corresponding to the target group of Hube losses is determined as the target smoothing coefficient, and the sum of the m Hube losses in the target group of Hube losses is the smallest.

[0094] For easier understanding, please refer to Figure 2 and Figure 3 , Figure 2 A schematic diagram of comparison between monitoring prediction data corresponding to a smoothing coefficient and actual stress monitoring data provided in an embodiment of the present application; Figure 3 A schematic diagram for comparing monitoring prediction data and actual stress monitoring data corresponding to another smoothing coefficient provided in an embodiment of the present application.

[0095] The calculation process of the sum of m Hube losses may specifically include: first, obtaining m date information of m stress monitoring actual data corresponding to the m Hube losses. Then, sorting the m date information from early to late, and assigning m preset loss weights to the m Hube losses corresponding to the sorted m date information in order from small to large. Finally, summing up the product of each Hube loss and the loss weight corresponding to each Hube loss to obtain the sum of the m Hube losses.

[0096] It should be noted that in order to optimize the performance of the prediction model, it is necessary to find a suitable smoothing coefficient that balances the smoothness of the model and the fit to the data. The Hube loss function is used as a tool to quantify the prediction error and adjust the smoothing coefficient accordingly. The Hube loss function combines the advantages of MSE (mean-square error) and MAE (Mean Absolute Error). When the prediction error is small (less than or equal to a certain threshold δ), MSE is used to penalize the error, which helps to maintain the model's sensitivity to the data. When the prediction error is large (greater than the threshold δ), MAE is used, which helps to reduce the model's sensitivity to outliers. The specific formula can be expressed as:

[0097]

[0098] In the formula, y true Represents the true value (which can be substituted into the actual stress monitoring data); pred represents the predicted value (which can be substituted into the stress monitoring prediction data); δ represents the Huber loss parameter; Huber Loss represents the Huber loss.

[0099] In the quadratic exponential smoothing method, since recent data is more important for predicting future short-term data, in order to ensure that the subsequent weight distribution can reflect the influence of time sequence, the m date information is sorted from early to late, that is, the more recent data should be assigned a higher weight. The m preset loss weights are assigned to the m Hube losses corresponding to the sorted m date information in order from small to large. The weight distribution should match the order of the date information, that is, the Hube losses corresponding to the data with later dates should be assigned a higher weight. The preset loss weights can be set in advance based on experience or actual needs. They reflect the importance of data at different time points to the prediction results. The product of each Hube loss and its corresponding loss weight is added to obtain the sum of the m Hube losses.

[0100] In order to find the optimal smoothing coefficient, the Hube losses corresponding to multiple different smoothing coefficients are calculated. The combination of the m Hube losses in each group of Hube losses reflects the overall prediction performance of the smoothing coefficient. Generally speaking, the smaller the sum of the Hube losses, the better the performance of the smoothing coefficient in the prediction. Therefore, after comparing the sum of the Hube losses corresponding to all smoothing coefficients, the group of Hube losses with the smallest sum is selected as the target group of Hube losses. The smoothing coefficient corresponding to this group of Hube losses is the target smoothing coefficient found on the current data set.

[0101] For easier understanding, please refer to Figure 4 and Figure 5 , Figure 4 A schematic diagram of the change of the Hube loss corresponding to a smoothing coefficient provided in an embodiment of the present application; Figure 5 A schematic diagram of the change of the Hube loss corresponding to another smoothing coefficient provided in an embodiment of the present application.

[0102] Step S105: for each current smoothing coefficient, determine an optimized smoothing coefficient based on the current smoothing coefficient, the target smoothing coefficient and the preset inertia weight.

[0103] In this application, in order to prevent the occurrence of local optimal solutions, the inertia weight is introduced to simulate the movement mechanism of the cane rat, and a new population of individuals is regenerated, that is, each current smoothing coefficient is optimized to obtain an optimized smoothing coefficient. The specific optimization process can refer to the following formula:

[0104] Optimized smoothing coefficient = current smoothing coefficient + inertia weight × (target smoothing coefficient - current smoothing coefficient)

[0105] The inertia weight (usually expressed as ω, omega) is a number between 0 and 1, which determines the degree of dependence of the new solution on the current solution. When the inertia weight is large, the new solution is more likely to search locally near the current solution; when the inertia weight is small, the new solution is more likely to jump out of the current area and search globally.

[0106] In order to understand the role of inertia weight more intuitively, the following example is given:

[0107] Assume that in the process of optimizing the smoothing coefficient, the current smoothing coefficient is 0.6 and the target smoothing coefficient is 0.8. The optimized smoothing coefficient in the next iteration is determined based on these two values ​​and the preset inertia weight.

[0108] If the inertia weight is set to 0.9, then according to the above formula, the optimized smoothing coefficient will be calculated as: optimized smoothing coefficient = 0.6 + 0.9 × (0.8-0.6) = 0.6 + 0.18 = 0.78. In this case, the new solution is very close to the current solution 0.6, and is more inclined to perform local search within the current area.

[0109] If the inertia weight is set to 0.1, then according to the above formula, the optimized smoothing coefficient will be calculated as: optimized smoothing coefficient = 0.6 + 0.1 × (0.8-0.6) = 0.6 + 0.02 = 0.62. In this case, although the new solution is still affected by the current solution, it is already further away from the current solution and is more likely to jump out of the current area and conduct a global search.

[0110] Step S106: taking each optimized smoothing coefficient as the current smoothing coefficient, and determining whether the number of returns reaches a preset number of iterations.

[0111] If the number of returns does not reach the preset number of iterations, the process returns to step S103; if the number of returns reaches the preset number of iterations, the optimized target smoothing coefficient is obtained.

[0112] In the present application, this step uses the smoothing coefficient optimized in each iteration as the new current smoothing coefficient, and iteratively executes steps S103, S104 and S105. Each iteration generates a set of prediction data based on the current smoothing coefficient, and calculates the corresponding Hube loss to determine a new target smoothing index, and then optimizes the current smoothing index based on the new target smoothing index until the number of returns reaches the preset number of iterations. The algorithm will stop iterating. The last iteration does not need to execute step S105, that is, it is no longer necessary to optimize the current smoothing index based on the new target smoothing index, and the calculation is sufficient to determine the latest target smoothing index (that is, the optimized target smoothing coefficient).

[0113] For easier understanding, please refer to Figure 6 and Figure 7 , Figure 6 A schematic diagram of a change in an optimized smoothing coefficient corresponding to a smoothing coefficient provided in an embodiment of the present application; Figure 7 A schematic diagram of the change of an optimized smoothing coefficient corresponding to another smoothing coefficient provided in an embodiment of the present application.

[0114] Step S107: performing secondary exponential smoothing prediction on the n pieces of actual stress monitoring data based on the optimized target smoothing coefficient to obtain the next k pieces of stress monitoring prediction data of the n pieces of actual stress monitoring data, where k≥1.

[0115] In the present application, firstly, n stress monitoring actual data are taken as the second target stress monitoring actual data. Then, based on the optimized target smoothing coefficient, the second target stress monitoring actual data is subjected to a secondary exponential smoothing prediction to obtain the last stress monitoring prediction data of the second target stress monitoring actual data. Finally, the second target stress monitoring actual data and the last stress monitoring prediction data of the second target stress monitoring actual data are taken as the second target stress monitoring actual data, and the step of performing a secondary exponential smoothing prediction on the second target stress monitoring actual data based on the optimized target smoothing coefficient to obtain the last stress monitoring prediction data of the second target stress monitoring actual data is returned until the number of returns reaches (k-1) times, and the last k stress monitoring prediction data of the n stress monitoring actual data are obtained.

[0116] Specifically, the optimized target smoothing coefficient and the second target stress monitoring actual data are used to perform a secondary exponential smoothing prediction to obtain the last stress monitoring prediction data of the second target stress monitoring actual data. The newly obtained stress monitoring prediction data is updated to the existing data set (i.e., the second target stress monitoring actual data). The updated data set (i.e., the second target stress monitoring actual data and the last stress monitoring prediction data of the second target stress monitoring actual data) and the optimized target smoothing coefficient are used to perform a secondary exponential smoothing prediction again to obtain the next stress monitoring prediction data. Repeat the above steps until the number of returns reaches (k-1) times and stop the prediction to obtain the last k stress monitoring prediction data of the n stress monitoring actual data. The advantage of this method is that it can adapt to data changes more flexibly, and each prediction is based on the latest data set, thereby improving the accuracy of the prediction. At the same time, this method can also better reflect the trend and periodic changes of the data. However, its disadvantage is that the amount of calculation is large, and multiple predictions and data updates are required.

[0117] As another possible implementation method, there is no need to generate prediction data one by one, but directly use the quadratic exponential smoothing method based on all available historical data (n stress monitoring actual data) to predict the next k stress monitoring prediction data of n stress monitoring actual data at one time. The advantage of this method is that it is highly efficient and can obtain all the required prediction data at one time. However, it may also lead to a lower accuracy of the prediction results due to factors such as the complexity of the model and the nonlinearity of the data.

[0118] In summary, the embodiment of the present application provides a method for predicting stress monitoring data, the method comprising: obtaining n actual stress monitoring data, randomly generating multiple smoothing coefficients based on a preset range, and taking each randomly generated smoothing coefficient as the current smoothing coefficient. For each current smoothing coefficient, a secondary exponential smoothing prediction is performed on the first nm actual stress monitoring data based on the current smoothing coefficient to obtain the next m stress monitoring prediction data, where m is the preset number of avoidance time nodes, m≥1 and n>m. Based on each of the next m stress monitoring prediction data and the next m stress monitoring actual data, a target smoothing coefficient is determined. For each current smoothing coefficient, an optimized smoothing coefficient is determined based on the current smoothing coefficient, the target smoothing coefficient and the preset inertia weight. Taking each optimized smoothing coefficient as the current smoothing coefficient, returning to perform a secondary exponential smoothing prediction on the first nm actual stress monitoring data based on the current smoothing coefficient for each current smoothing coefficient to obtain the next m stress monitoring prediction data, until the number of returns reaches the preset number of iterations, and the optimized target smoothing coefficient is obtained. A secondary exponential smoothing prediction is performed on the n actual stress monitoring data based on the optimized target smoothing coefficient. This scheme continuously optimizes the smoothing coefficient until the preset number of iterations is reached, and introduces inertia weight to make the selection of smoothing coefficient more objective and scientific, avoiding the limitations of directly using randomly generated smoothing coefficients and manually adjusting smoothing coefficients, which is conducive to improving the accuracy of predicted stress monitoring data.

[0119] Based on the above embodiment, the method may further include: first obtaining the last k pieces of stress monitoring actual data of the n pieces of stress monitoring actual data, then calculating the difference between the last k pieces of stress monitoring actual data of the n pieces of stress monitoring actual data and the last k pieces of stress monitoring predicted data of the n pieces of stress monitoring actual data, and finally determining whether a prediction abnormality occurs based on the difference and a preset difference.

[0120] The difference between the last k data of the n stress monitoring actual data and the corresponding last k predicted data reflects the degree of difference between the predicted data and the actual data. Based on the calculated difference and the preset difference threshold (i.e., the preset difference), it can be determined whether the prediction is abnormal. If the difference exceeds the preset difference threshold, it is considered that the prediction is abnormal.

[0121] For ease of understanding, an example is given below: When the test load is 30.8t, the actual stress monitoring data of a certain point on a ship unloader within 40 (n+k) days is measured. The actual stress monitoring data of the first 35 (n) days is used to optimize the global optimal smoothing coefficient (the optimized target smoothing coefficient) by the large cane rat algorithm, and then the prediction is made by the quadratic exponential smoothing method to obtain the stress monitoring prediction data of the next 5 (k) days. The difference between the last k stress monitoring actual data and the last k stress monitoring prediction data is the prediction error. For details, please refer to Table 2:

[0122] Table 2

[0123] Period (days) Measured value Prediction value 1 error Prediction value 2 error 36 47.04 46.34 -0.70 46.94 -0.10 37 46.60 46.48 -0.12 46.77 -0.17 38 47.09 46.68 -0.39 46.32 -0.75 39 46.23 46.51 +0.28 46.38 +0.15 40 46.77 46.62 -0.15 46.63 -0.14

[0124] It can be seen from the table that after using the above method for prediction many times, the prediction error is small and is below the preset difference (such as 0.80), indicating that the method has good reliability.

[0125] Furthermore, if there is an abnormal prediction, the weights of different points can be set. For details, please refer to the following formula:

[0126]

[0127] In the formula, p is the predicted abnormality, n is the number of predefined consecutive exceeding thresholds, and y i is the predicted value or true value, t is the predefined threshold value; g is the range that can exceed the threshold value under theoretical conditions. If there are multiple threshold values ​​with different standards, g can be the difference between two different standards. If there is only one standard, g can be selected according to the actual situation.

[0128] In summary, this embodiment can timely detect situations where stress exceeds the normal range by comparing the error with the predefined threshold, thereby enhancing the early warning capability and helping to prevent structural damage or safety accidents caused by abnormal stress. By setting weights for different points, resource allocation can be optimized based on the importance of each point to the overall structure. For points with higher weights, more attention and monitoring resources can be given to ensure the safety of the overall structure.

[0129] A method for predicting stress monitoring data provided by an embodiment of the present application is introduced above. A device for executing the method for predicting stress monitoring data will be introduced below.

[0130] See also Figure 8 , Figure 8 A schematic diagram of the structure of a stress monitoring data prediction device provided in an embodiment of the present application.

[0131] like Figure 8 As shown, the prediction device for stress monitoring data includes:

[0132] The acquisition unit 11 is used to acquire n pieces of actual stress monitoring data, and divide the n pieces of actual stress monitoring data into first nm pieces of actual stress monitoring data and last m pieces of actual stress monitoring data, where m is a preset number of avoidance time nodes, m≥1 and n>m.

[0133] The first determining unit 12 is used to randomly generate a plurality of smoothing coefficients based on a preset range, and use each randomly generated smoothing coefficient as a current smoothing coefficient.

[0134] The first prediction unit 13 is used for performing a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficient for each of the current smoothing coefficients to obtain the next m stress monitoring prediction data.

[0135] The second determination unit 14 is used to determine a target smoothing coefficient based on the next m pieces of stress monitoring prediction data and the next m pieces of stress monitoring actual data.

[0136] The third determining unit 15 is configured to determine, for each of the current smoothing coefficients, an optimized smoothing coefficient based on the current smoothing coefficient, the target smoothing coefficient and a preset inertia weight.

[0137] The return unit 16 is used to use each of the optimized smoothing coefficients as the current smoothing coefficient, and return to execute the step of performing a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficient for each of the current smoothing coefficients to obtain the next m stress monitoring prediction data, until the number of returns reaches a preset number of iterations to obtain the optimized target smoothing coefficient.

[0138] The second prediction unit 17 is used to perform a secondary exponential smoothing prediction on the n stress monitoring actual data based on the optimized target smoothing coefficient to obtain the next k stress monitoring prediction data of the n stress monitoring actual data, where k≥1.

[0139] In a possible implementation, the first prediction unit 13 includes:

[0140] The first determination subunit is used to use the first nm pieces of stress monitoring actual data as first target stress monitoring actual data for each of the current smoothing coefficients.

[0141] The first prediction subunit is used to perform a secondary exponential smoothing prediction on the first target stress monitoring actual data based on the current smoothing coefficient to obtain a subsequent stress monitoring prediction data of the first target stress monitoring actual data.

[0142] The first return subunit is used to take the first target stress monitoring actual data and the stress monitoring prediction data after the first target stress monitoring actual data as the first target stress monitoring actual data, and return to execute the step of performing a secondary exponential smoothing prediction on the first target stress monitoring actual data based on the current smoothing coefficient to obtain the stress monitoring prediction data after the first target stress monitoring actual data, until the number of returns reaches (m-1) times, and the last m stress monitoring prediction data are obtained.

[0143] In a possible implementation, the second determining unit 14 includes:

[0144] The comparison subunit is used to compare the last m stress monitoring prediction data with the last m stress monitoring actual data using the Hube loss function for each of the last m stress monitoring prediction data to obtain a group of Hube losses, wherein the group of Hube losses includes m Hube losses.

[0145] The second determination subunit is used to determine a target group of Hube losses from each group of Hube losses, and determine the current smoothing coefficient corresponding to the target group of Hube losses as the target smoothing coefficient, wherein the sum of m Hube losses in the target group of Hube losses is the smallest.

[0146] In a possible implementation, the device further includes:

[0147] The information acquisition unit is used to acquire m date information of m pieces of actual stress monitoring data corresponding to the m Hubbe losses.

[0148] An allocating unit is used to sort the m date information from early to late, and allocate the m preset loss weights to the m Hube losses corresponding to the sorted m date information in order from small to large.

[0149] The adding unit is used to add the products of each of the Hubbe losses and the loss weights corresponding to the Hubbe losses to obtain the sum of the m Hubbe losses.

[0150] In a possible implementation, the second prediction unit 17 includes:

[0151] The third determining subunit is configured to use the n pieces of stress monitoring actual data as second target stress monitoring actual data.

[0152] The second prediction subunit is used to perform a secondary exponential smoothing prediction on the second target stress monitoring actual data based on the optimized target smoothing coefficient to obtain the next stress monitoring prediction data of the second target stress monitoring actual data.

[0153] The second return subunit is used to take the second target stress monitoring actual data and the stress monitoring prediction data after the second target stress monitoring actual data as the second target stress monitoring actual data, and return to execute the step of performing a quadratic exponential smoothing prediction on the second target stress monitoring actual data based on the optimized target smoothing coefficient to obtain the stress monitoring prediction data after the second target stress monitoring actual data, until the number of returns reaches (k-1) times, and the k stress monitoring prediction data after the n stress monitoring actual data are obtained.

[0154] In a possible implementation, the device further includes:

[0155] The data acquisition unit is used to acquire the last k pieces of stress monitoring actual data of the n pieces of stress monitoring actual data.

[0156] The calculation unit is used to calculate the difference between the last k pieces of stress monitoring actual data of the n pieces of stress monitoring actual data and the last k pieces of stress monitoring predicted data of the n pieces of stress monitoring actual data.

[0157] The fourth determining unit is used to determine whether a predicted abnormality occurs based on the difference and a preset difference.

[0158] The present application also provides an electronic device in an embodiment. Fig. 9 As shown, it shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present application. The electronic device in the embodiments of the present application may include but is not limited to fixed terminals such as mobile phones, notebook computers, PDAs (personal digital assistants), PADs (tablet computers), desktop computers, etc. Fig. 9 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0159] like Fig. 9 As shown, the electronic device may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage device 908 into a random access memory (RAM) 903. When the electronic device is powered on, various programs and data required for the operation of the electronic device are also stored in the RAM 903. The processing device 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0160] Typically, the following devices may be connected to the I / O interface 905: an input device 906 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 907 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 908 including, for example, a memory card, a hard disk, etc.; and a communication device 909. The communication device 909 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although Fig. 9 An electronic device having various devices is shown, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0161] An embodiment of the present application also provides a computer program product including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements any stress monitoring data prediction method provided in the embodiment of the present application.

[0162] A computer-readable storage medium is also provided in an embodiment of the present application. The storage medium carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement any stress monitoring data prediction method provided in the embodiment of the present application.

[0163] It should also be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the device embodiments provided by the present application, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines.

[0164] Through the description of the above implementation mode, the technicians in the field can clearly understand that the present application can be implemented by means of software plus necessary general hardware, and of course, it can also be implemented by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. In general, all functions completed by computer programs can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be various, such as analog circuits, digital circuits or special circuits. However, for the present application, software program implementation is a better implementation mode in more cases. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a readable storage medium, such as a computer floppy disk, a U disk, a mobile hard disk, a ROM, a RAM, a disk or an optical disk, etc., including a number of instructions to enable a computer device (which can be a personal computer, a training device, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0165] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0166] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a training device, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, training device, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device, a data center, etc. that includes one or more available media integrations. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

Claims

1. A method for predicting stress monitoring data, characterized in that: include: Obtaining n pieces of actual stress monitoring data, and dividing the n pieces of actual stress monitoring data into first nm pieces of actual stress monitoring data and last m pieces of actual stress monitoring data, where m is a preset number of avoidance time nodes, m≥1 and n>m; Randomly generate a plurality of smoothing coefficients based on a preset range, and use each randomly generated smoothing coefficient as a current smoothing coefficient; For each of the current smoothing coefficients, a secondary exponential smoothing prediction is performed on the first nm stress monitoring actual data based on the current smoothing coefficient to obtain the next m stress monitoring prediction data; Determining a target smoothing coefficient based on the next m pieces of stress monitoring prediction data and the next m pieces of stress monitoring actual data; For each of the current smoothing coefficients, determining an optimized smoothing coefficient based on the current smoothing coefficient, the target smoothing coefficient and a preset inertia weight; The optimized smoothing coefficients are respectively used as the current smoothing coefficients, and the step of performing a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficients for each of the current smoothing coefficients to obtain the next m stress monitoring prediction data is returned until the number of returns reaches a preset number of iterations to obtain the optimized target smoothing coefficients; Based on the optimized target smoothing coefficient, a secondary exponential smoothing prediction is performed on the n stress monitoring actual data to obtain the next k stress monitoring prediction data of the n stress monitoring actual data, where k≥1.

2. The stress monitoring data prediction method according to claim 1, characterized in that: For each of the current smoothing coefficients, performing secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficient to obtain the next m stress monitoring prediction data includes: For each of the current smoothing coefficients, taking the first nm pieces of stress monitoring actual data as first target stress monitoring actual data; Performing a secondary exponential smoothing prediction on the first target stress monitoring actual data based on the current smoothing coefficient to obtain a subsequent stress monitoring prediction data of the first target stress monitoring actual data; The first target stress monitoring actual data and the stress monitoring prediction data after the first target stress monitoring actual data are used as the first target stress monitoring actual data, and the step of performing a secondary exponential smoothing prediction on the first target stress monitoring actual data based on the current smoothing coefficient to obtain the stress monitoring prediction data after the first target stress monitoring actual data is returned to be executed, until the number of returns reaches (m-1) times, and the last m stress monitoring prediction data are obtained.

3. The stress monitoring data prediction method according to claim 1, characterized in that: The determining of the target smoothing coefficient based on the next m pieces of stress monitoring prediction data and the next m pieces of stress monitoring actual data includes: For each of the last m stress monitoring prediction data, using a Hube loss function to compare the last m stress monitoring prediction data with the last m stress monitoring actual data to obtain a set of Hube losses, wherein the set of Hube losses includes m Hube losses; A target group of Hube losses is determined from each group of the Hube losses, and the current smoothing coefficient corresponding to the target group of Hube losses is determined as the target smoothing coefficient, wherein the sum of the m Hube losses in the target group of Hube losses is the smallest.

4. The method for predicting stress monitoring data according to claim 3, characterized in that: The calculation process of the sum of the m Hube losses includes: Obtaining m date information of m pieces of actual stress monitoring data corresponding to the m Hubbe losses; Sorting the m date information from early to late, and allocating m preset loss weights to the m Hube losses corresponding to the sorted m date information in ascending order; The products of each of the Hubbe losses and the loss weights corresponding to the Hubbe losses are added together to obtain the sum of the m Hubbe losses.

5. The method for predicting stress monitoring data according to claim 1, characterized in that: The performing of secondary exponential smoothing prediction on the n pieces of actual stress monitoring data based on the optimized target smoothing coefficient to obtain subsequent k pieces of stress monitoring prediction data of the n pieces of actual stress monitoring data comprises: Using the n pieces of stress monitoring actual data as second target stress monitoring actual data; Performing a secondary exponential smoothing prediction on the second target stress monitoring actual data based on the optimized target smoothing coefficient to obtain the next stress monitoring prediction data of the second target stress monitoring actual data; The second target stress monitoring actual data and the stress monitoring prediction data after the second target stress monitoring actual data are used as the second target stress monitoring actual data, and the step of performing a quadratic exponential smoothing prediction on the second target stress monitoring actual data based on the optimized target smoothing coefficient to obtain the stress monitoring prediction data after the second target stress monitoring actual data is returned to, until the number of returns reaches (k-1) times, and the k stress monitoring prediction data after the n stress monitoring actual data are obtained.

6. The method for predicting stress monitoring data according to claim 1, characterized in that: After performing secondary exponential smoothing prediction on the n pieces of actual stress monitoring data based on the optimized target smoothing coefficient to obtain the next k pieces of predicted stress monitoring data of the n pieces of actual stress monitoring data, the method further includes: Obtaining the next k pieces of stress monitoring actual data of the n pieces of stress monitoring actual data; Calculate the difference between the next k pieces of actual stress monitoring data of the n pieces of actual stress monitoring data and the next k pieces of predicted stress monitoring data of the n pieces of actual stress monitoring data; Based on the difference and a preset difference, it is determined whether a predicted abnormality occurs.

7. A method and device for predicting stress monitoring data, characterized in that: include: An acquisition unit, used for acquiring n pieces of actual stress monitoring data, and dividing the n pieces of actual stress monitoring data into first nm pieces of actual stress monitoring data and last m pieces of actual stress monitoring data, where m is a preset number of avoidance time nodes, m≥1 and n>m; A first determining unit, configured to randomly generate a plurality of smoothing coefficients based on a preset range, and use each randomly generated smoothing coefficient as a current smoothing coefficient; A first prediction unit is used for performing a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficient for each of the current smoothing coefficients to obtain the next m stress monitoring prediction data; A second determination unit is used to determine a target smoothing coefficient based on the next m pieces of stress monitoring prediction data and the next m pieces of stress monitoring actual data; A third determining unit, configured to determine, for each of the current smoothing coefficients, an optimized smoothing coefficient based on the current smoothing coefficient, the target smoothing coefficient and a preset inertia weight; a return unit, configured to respectively use each of the optimized smoothing coefficients as the current smoothing coefficient, and return to execute the step of performing a secondary exponential smoothing prediction on the first nm stress monitoring actual data based on the current smoothing coefficient for each of the current smoothing coefficients to obtain the next m stress monitoring prediction data, until the number of returns reaches a preset number of iterations, and the optimized target smoothing coefficient is obtained; The second prediction unit is used to perform secondary exponential smoothing prediction on the n stress monitoring actual data based on the optimized target smoothing coefficient to obtain subsequent k stress monitoring prediction data of the n stress monitoring actual data, where k≥1.

8. A computer program product, characterized in that It comprises computer-readable instructions, and when the computer-readable instructions are executed on an electronic device, the electronic device implements the stress monitoring data prediction method according to any one of claims 1 to 6.

9. An electronic device, characterized in that: The method comprises at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer program so that the electronic device can implement the stress monitoring data prediction method according to any one of claims 1 to 6.

10. A computer storage medium, characterized in that: The storage medium carries one or more computer programs, and when the one or more computer programs are executed by an electronic device, the electronic device can implement the stress monitoring data prediction method as described in any one of claims 1 to 6.

Images