Method, device and equipment for predicting change trend of ruins and diseases and storage medium
By obtaining and preprocessing the disease data of urban wall sites and using the timing prediction model to predict future disease changes, the problems of inefficient prediction efficiency and poor accuracy in the existing technology are solved, and the protection efficiency of urban wall sites is improved.
Patent Information
- Application Number
- CN202411252948.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-09
- Publication Date
- 2024-12-24
AI Technical Summary
In the prior art, the disease prediction efficiency of urban wall sites is low, and the prediction results are poor, resulting in poor protection effect.
By obtaining disease-related data of the target site in the preset historical period, data preprocessing is performed, feature collections are extracted, and the timing prediction model is used to predict future disease change trends, including crack expansion rate, temperature and humidity changes.
It improves the accuracy of predicting the change trend of the site disease and improves the protection efficiency of urban wall sites.
Smart Images
Figure CN119180206A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of pre-inspection of heritage site diseases, and in particular to a method, device, equipment and storage medium for predicting the changing trend of heritage site diseases. Background Art
[0002] As an important part of historical and cultural heritage, city wall sites have extremely high historical, artistic and scientific values. However, due to the long-term exposure of city wall sites to the natural environment, they are affected by factors such as climate change, geological activities, biological activity trajectories and human destruction, which have caused the city wall sites to gradually develop cracks, erosion, collapse and other diseases, which seriously affect the preservation of the city wall sites. Therefore, there is an urgent need for a device that can accurately predict the trend of changes in the diseases of city wall sites to improve the protection efficiency of city wall sites.
[0003] In the relevant technology, relevant staff usually use relevant measuring tools to periodically measure the relics, traces, remains, cracks, humidity, temperature and other site data of city wall sites. Site protection experts predict the trend of disease changes of city wall sites based on work experience and real-time measured site data, and take corresponding protection measures to protect city wall sites based on manual prediction results.
[0004] However, when protecting city wall sites based on relevant technologies, there are problems such as low efficiency of manual prediction and poor accuracy of manual prediction results, resulting in poor protection effects of city wall sites. Summary of the invention
[0005] The purpose of this application is to provide a method, device, equipment and storage medium for predicting the trend of changes in site damage, which can improve the accuracy of the prediction results of the trend of changes in site damage, so as to improve the protection efficiency of city wall sites.
[0006] The embodiment of the present application is implemented as follows:
[0007] In a first aspect of an embodiment of the present application, a method for predicting a trend of changes in site damage is provided, the method comprising:
[0008] Obtain disease-related data of the target site within a preset historical period, including: crack data, temperature data, and humidity data;
[0009] Performing data preprocessing on disease-related data within a preset historical period, extracting a first target feature set within the preset historical period, and obtaining a first target time series data sequence within the preset historical period, wherein the first target feature set includes: at least one first target feature;
[0010] According to the first target time series data sequence, the first target feature set, the preset future time period and the target time series prediction model, the prediction results of the disease change trend of the target site in the future time period are determined. The disease change trend prediction results include: crack propagation rate prediction results, temperature rise and fall amplitude prediction results and humidity increase and decrease amplitude prediction results.
[0011] As a possible implementation method, according to the first target time series data sequence, the first target feature set, the preset future period and the target time series prediction model, the prediction result of the disease change trend of the target site in the future period is determined, including:
[0012] Determine the historical change trend of each first target feature within a preset historical period according to the first target time series data sequence and the first target feature set;
[0013] Determine a weight value corresponding to each first target feature according to the target time series prediction model, the first target feature set, and the historical change trend of each first target feature;
[0014] The weight values corresponding to the first target features, the historical change trends of the first target features, and the future time periods are input into the target time series prediction model, and the target time series prediction model determines the prediction results of the disease change trends in the future time period.
[0015] As a possible implementation method, the target time series prediction model determines the prediction results of disease change trends in the future period, including:
[0016] The target time series prediction model predicts the future change trend of each first target feature in the future time period based on the weight value corresponding to each first target feature, the historical change trend of each first target feature and the future time period;
[0017] The target time series prediction model is based on the future change trend of each first target feature in the future period to obtain the prediction result of the disease change trend in the future period.
[0018] As a possible implementation method, data preprocessing is performed on disease-related data within a preset historical period, a first target feature set within the preset historical period is extracted, and a first target time series data sequence within the preset historical period is obtained, including:
[0019] Performing data cleaning on disease-related data within a preset historical period to obtain a first target disease data set within the preset historical period;
[0020] Performing feature extraction on the first target disease data set to obtain a first target feature set within a preset historical period;
[0021] The first target disease data set is time-aligned based on a preset standard time axis to obtain a first target time series data sequence within a preset historical period.
[0022] As a possible implementation method, data cleaning is performed on disease-related data within a preset historical period to obtain a first target disease data set within the preset historical period, including:
[0023] Determine missing values and invalid values in the disease-related data within the preset historical period based on the disease-related data within the preset historical period and the preset standard disease-related data;
[0024] According to the preset standard disease-related data, invalid values in the disease-related data within the preset historical period are deleted, and missing values in the disease-related data within the preset historical period are filled to obtain a first target disease data set within the preset historical period.
[0025] As a possible implementation method, after determining the disease change trend prediction result according to the first target time series data sequence, the first target feature set, the preset future period, and the target time series prediction model, it also includes:
[0026] According to the disease change trend prediction results and the preset disease weight value, determine the alarm level corresponding to the disease change trend prediction results, and issue corresponding alarm information according to the alarm level;
[0027] According to the warning level and warning information corresponding to the prediction results of disease change trend, a disease prediction report is generated, and a sequence of target protection measures is pushed in descending order according to the warning level.
[0028] As a possible implementation method, the above method for predicting the trend of site damage changes also includes:
[0029] Obtain historical disease-related data sets;
[0030] Performing data preprocessing on the historical disease-related data set to obtain a second target disease data set corresponding to the historical disease-related data set, extracting a second target feature set corresponding to the second target disease data set, and obtaining a second target time series data sequence corresponding to the second target disease data set, wherein the second target feature set includes the first target feature set;
[0031] Performing data segmentation on the second target disease data set to obtain a training data set, a verification data set, and a test data set;
[0032] Constructing an initial time series prediction model according to the training data set, the third target feature set corresponding to the training data set, the third target time series data sequence corresponding to the training data set, and the preset deep learning model;
[0033] Iteratively verify the initial time series prediction model according to the verification data set, the fourth target feature set corresponding to the verification data set, and the fourth target time series data sequence corresponding to the verification data set, and determine the target weight value corresponding to each second target feature in the second target feature set;
[0034] According to the target weight values corresponding to each second target feature, the test data set, the fifth target feature set corresponding to the test data set, and the fifth target time series data sequence corresponding to the test data set, the initial time series prediction model is iteratively tested, and the target time series prediction model is determined according to the iterative test results.
[0035] According to a second aspect of the embodiment of the present application, there is provided a device for predicting the trend of changes in site damage, the device comprising:
[0036] An acquisition module is used to obtain disease-related data of the target site within a preset historical period, and the disease-related data includes: crack data, temperature data, and humidity data;
[0037] A data processing module is used to perform data preprocessing on disease-related data within a preset historical period, extract a first target feature set within the preset historical period, and obtain a first target time series data sequence within the preset historical period, wherein the first target feature set includes: at least one first target feature;
[0038] The prediction module is used to determine the prediction results of the disease change trend of the target site in the future time period based on the first target time series data sequence, the first target feature set, the preset future time period and the target time series prediction model. The disease change trend prediction results include: crack propagation rate prediction results, temperature rise and fall amplitude prediction results and humidity increase and decrease amplitude prediction results.
[0039] According to a third aspect of an embodiment of the present application, a computer device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the method for predicting the trend of changes in site diseases as described in the first aspect above is implemented.
[0040] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the method for predicting the trend of changes in site diseases as described in the first aspect is implemented.
[0041] The beneficial effects of the embodiments of the present application include:
[0042] The embodiment of the present application provides a method for predicting the trend of changes in the disease of a site, by obtaining the disease-related data of the target site in a preset historical period, the disease-related data includes: crack data, temperature data and humidity data; preprocessing the disease-related data in the preset historical period, extracting the first target feature set corresponding to the disease-related data in the preset historical period, and obtaining the first target time series data sequence corresponding to the disease-related data in the preset historical period, the first target feature set includes multiple first target features; according to the first target time series data sequence, the first target feature set, the preset future period and the target time series prediction model, predicting the disease change trend of the target site in the future period. Among them, the disease change trend prediction results include: crack expansion rate prediction results, temperature rise and fall amplitude prediction results and temperature increase and decrease amplitude prediction results. In this way, the accuracy of the prediction results of the change trend of the site disease can be improved, so as to improve the protection efficiency of the city wall type site. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0044] Figure 1 A flow chart of the first method for predicting the trend of changes in site diseases provided in the embodiment of the present application;
[0045] Figure 2 A flow chart of a second method for predicting the trend of changes in site diseases provided in an embodiment of the present application;
[0046] Figure 3 A flow chart of a third method for predicting the trend of changes in site diseases provided in an embodiment of the present application;
[0047] Figure 4 A flow chart of a fourth method for predicting the trend of changes in site diseases provided in an embodiment of the present application;
[0048] Figure 5 A flow chart of a fifth method for predicting the trend of changes in site diseases provided in an embodiment of the present application;
[0049] Figure 6 A flow chart of a sixth method for predicting the trend of changes in site diseases provided in an embodiment of the present application;
[0050] Figure 7 A flow chart of a seventh method for predicting the trend of changes in site diseases provided in an embodiment of the present application;
[0051] Figure 8 A visualization interface for disease change trend prediction results provided in an embodiment of the present application;
[0052] Fig. 9 A schematic diagram of the structure of a device for predicting the changing trend of site diseases provided in an embodiment of the present application;
[0053] Fig.10 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0054] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.
[0055] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.
[0056] At present, patrol personnel usually use relevant measurement tools to periodically measure the relics, traces, remains, cracks, humidity, temperature and other site data of city wall sites. Site protection experts predict the trend of disease changes in city wall sites based on work experience and real-time measured site data, and take corresponding protection measures to protect city wall sites based on manual prediction results. However, this solution relies too much on manual measurement and professional experience, which leads to low efficiency and poor accuracy of manual prediction results, which in turn leads to poor protection of city wall sites.
[0057] To this end, the embodiment of the present application provides a method for predicting the trend of changes in site damage, by collecting the target site's damage-related data within a preset historical period; performing data preprocessing on the damage-related data within the preset historical period, extracting the first target feature set within the preset historical period, and obtaining the first target time series data sequence within the preset historical period; and determining the prediction result of the target site's damage change trend in the future period based on the first target time series data sequence, the first target feature set, and the preset future period and target time series prediction model. In this way, the accuracy of the prediction result of the site's damage change trend can be improved to improve the protection efficiency of city wall sites.
[0058] A method for predicting the changing trend of site disease damage provided in an embodiment of the present application is applied to a computer device, a terminal device or a third-party server, etc. The present application takes the application to a computer device as an example, but it does not mean that the method for predicting the changing trend of site disease damage can only be executed on a computer device, and the present application does not make any specific limitation on this.
[0059] Optionally, inspection personnel of city wall sites can select multiple site monitoring points in the target site in advance, and deploy site data monitoring equipment at the site monitoring points. The site data monitoring equipment automatically collects disease-related data at each site monitoring point in the target site in real time, and transmits the real-time collected disease-related data to a computer device. The computer device stores a pre-trained target time series prediction model. The target time series prediction model predicts the disease change trend of the target site in the future time period based on the disease-related data in the historical time period collected by the site data monitoring equipment and the future time period input by the user.
[0060] Optionally, the site data monitoring equipment can be implemented by a variety of data sensors or other data acquisition equipment, and this application does not make specific limitations on this.
[0061] The following is a detailed explanation of the method for predicting the changing trend of site disease damage provided in the embodiments of the present application.
[0062] Figure 1 A flow chart of a method for predicting the changing trend of heritage site diseases provided in this application. The method can be applied to computer equipment, user terminal equipment or third-party servers.
[0063] See also Figure 1 The present application embodiment provides a method for predicting the trend of changes in site diseases, including:
[0064] S101. Obtain disease-related data of a target site within a preset historical period, where the disease-related data include: crack data, temperature data, and humidity data.
[0065] Optionally, the disease-related data is used to indicate the relevant data causing the disease at each monitoring point in the city wall ruins, such as the temperature data, humidity data, crack data, etc. that cause the crack expansion disease at point A of the wall of the city wall ruins. It is worth noting that the disease-related data of each monitoring point in the city wall ruins are isolated and do not interfere with each other, and the disease-related data of each monitoring point are stored separately.
[0066] Optionally, the target site is used to indicate the monitoring point selected by the user in the human-computer interaction interface of the computer device. The target site can be any monitoring point in the city wall type site. The user can select any monitoring point in the city wall type site in the human-computer interaction interface of the computer device to predict the trend of disease changes.
[0067] Optionally, the disease-related data include: crack data, temperature data and humidity data. Among them, the crack data is used to indicate the width and depth of cracks when cracks occur at the target site in the city wall type site; the temperature data is used to indicate the surface temperature at the target site in the city wall type site; the humidity data is used to indicate the water content and moisture content inside the target site in the city wall type site.
[0068] Optionally, the preset historical period refers to the time course of disease-related data extracted for fitting the disease change trend of the target site in the future period. The preset historical period is associated with the future period, the preset historical period and the future period are adjacent to each other on the time axis, and the preset historical period is much larger than the future period.
[0069] For example, if the disease change trend of point A in the city wall ruins in the next week (September 1st-September 7th) is predicted, it is necessary to call the disease-related data of the preset historical period (August 16th-August 31st) close to the next week; if the disease change trend of point A in the city wall ruins in the next 24 hours (September 1st) is predicted, it is necessary to call the disease-related data of 48 hours in the preset historical period (August 30th-August 31st) close to the next 24 hours.
[0070] It is worth noting that the selection range of the preset historical period is determined by the length of the preset future period selected by the user. The preset historical period can be twice, three times, or other multiples of the future period, and this application does not make any specific restrictions on this.
[0071] S102, preprocessing the disease-related data within a preset historical period, extracting a first target feature set within the preset historical period, and obtaining a first target time series data sequence within the preset historical period, wherein the first target feature set includes: at least one first target feature.
[0072] Optionally, data preprocessing refers to data processing such as noise reduction and extraction of the disease-related data collected directly from the target site by the site data monitoring equipment. The first target feature refers to the characteristic factor that directly affects the development trend of the site disease. For example, the first target feature can be season, climate, weather, rainfall, etc. The first target feature set is used to indicate the set of first target features corresponding to the disease-related data within a preset historical period. It is worth noting that the factors affecting the development trend of a certain disease in the target site can be a single factor or multiple factors. For example, the crack expansion trend is affected by multiple factors such as climate, weather, rainfall, etc. Each factor that affects the development trend of the site disease is a first target feature.
[0073] Optionally, the first target time series data sequence is used to indicate the time series data sequence obtained by arranging the disease-related data within a preset historical period according to a preset standard time axis. For example, if the disease-related data are arranged according to a 24-hour time axis every day, the data sequence obtained after arrangement according to the time axis is the target time series data sequence.
[0074] S103. Determine the prediction results of the disease change trend of the target site in the future period according to the first target time series data sequence, the first target feature set, the preset future period and the target time series prediction model. The disease change trend prediction results include: crack propagation rate prediction results, temperature rise and fall amplitude prediction results and humidity increase and decrease amplitude prediction results.
[0075] Optionally, the preset future time period refers to the predicted time period selected by the user in the human-computer interaction interface of the computer device. The preset future time period can be 1 day, 5 days, a week, 15 days, a month, etc. The preset future time period is randomly set by the user, and this application does not make any specific limitations on this.
[0076] Optionally, the target time series prediction model is obtained through training and learning based on a large amount of historical disease-related data. The target time series prediction model can actively read the disease-related data of the historical time period that is adapted to it based on the future time period selected by the user, and automatically fit the disease development trend of the target site in the future time period according to the first target feature combination corresponding to the disease-related data in the historical time period, the first target time series data sequence and the preset future time period; the target time series prediction model can also passively receive the disease-related data of the historical time period sent by the computer device, and automatically fit the disease development trend of the target site in the future time period based on the disease data in the historical time period and the future time period input by the user.
[0077] It is worth noting that the target time series prediction model can simultaneously fit the changing trends of multiple diseases at the same target site monitoring point, or it can only predict the changing trend of one disease. This application does not make specific limitations on this.
[0078] Optionally, the damage change trend prediction result is used to indicate the target time series prediction model's prediction of the damage change trend of the target site in the future period input by the user, such as predicting the crack expansion rate in the future period, predicting the cumulative change in humidity in the future period, etc. Among them, the crack expansion rate prediction result is used to indicate the cumulative change in the crack expansion rate of the target site in the future period; the temperature rise and fall amplitude prediction result is used to indicate the cumulative change amplitude of the temperature rise or fall of the target site in the future period; the humidity increase and decrease amplitude prediction result is used to indicate the cumulative increase or decrease in humidity or water content of the target site in the future period.
[0079] In an embodiment of the present application, by obtaining the disease-related data of the target site within a preset historical period, the disease-related data includes: crack data, temperature data and humidity data; preprocessing the disease-related data within the preset historical period, extracting the first target feature set corresponding to the disease-related data within the preset historical period, and obtaining the first target time series data sequence corresponding to the disease-related data within the preset historical period, the first target feature set includes multiple first target features; according to the first target time series data sequence, the first target feature set, the preset future period and the target time series prediction model, predict the disease change trend of the target site in the future period. Among them, the disease change trend prediction results include: crack expansion rate prediction results, temperature rise and fall amplitude prediction results and temperature increase and decrease amplitude prediction results. In this way, the accuracy of the prediction results of the site disease change trend can be improved to improve the protection efficiency of city wall sites.
[0080] For a possible implementation, see Figure 2 The operation of step S103 may specifically be:
[0081] S201. Determine the historical change trend of each first target feature within a preset historical period according to the first target time series data sequence and the first target feature set.
[0082] Optionally, the change trend of the first target feature is used to indicate the cumulative change of each first feature in the first target feature set within a preset historical period, such as the temperature cumulatively rising by 6° C. in the past day.
[0083] Optionally, based on the first time series data sequence, the disease-related data at each time point on the preset standard time axis can be determined, and based on the correlation between the disease-related data and each first target feature in the first target feature set, the time series corresponding to the first target feature can be determined, and based on the time series corresponding to the first target feature and the value of the first target feature at each time point in the time series, the cumulative change of the first target feature in the historical period can be determined, that is, the change trend of the first target feature can be determined.
[0084] S202: Determine a weight value corresponding to each first target feature according to the target time series prediction model, the first target feature set, and the historical change trend of each first target feature.
[0085] Optionally, the target time series prediction model can determine the degree of influence of each first target feature on the disease-related data based on all first target features contained in the disease-related data within a preset historical period and the historical change trends of each first target feature, and determine the weight value corresponding to each first target feature according to the degree of influence of each first target feature on the disease-related data.
[0086] Optionally, the weight value of the first target feature is used to characterize the degree of correlation between the first target feature and the development trend of the site disease. The larger the weight value, the greater the influence of the first target feature on the development trend of the site disease; conversely, the smaller the weight value, the smaller the influence of the first target feature on the development trend of the site disease.
[0087] S203, inputting the weight value corresponding to each first target feature, the historical change trend of each first target feature and the future time period into the target time series prediction model, and using the target time series prediction model to determine the prediction result of the disease change trend in the future time period.
[0088] Optionally, the weight values corresponding to each first target feature, the historical change trends of each first target feature and the future time periods are input into the target time series prediction model, and the target time series prediction model is fitted to predict the disease change trends of the target site in the future time period.
[0089] For a possible implementation, see Figure 3 The operation of step S203 may specifically be:
[0090] S301. The target time series prediction model predicts the future change trend of each first target feature in a future time period based on the weight value corresponding to each first target feature, the historical change trend of each first target feature and the future time period.
[0091] Optionally, the target time series prediction model determines the influence of each first target feature on the disease development trend of the target site according to the weight value corresponding to each first target feature; the target time series prediction model can determine the cumulative change of each first target feature in the historical period, and the correlation between the cumulative change of each first target feature and the disease change trend in the historical period according to the historical change trend of each first target feature; the target time series prediction model predicts the change trend of each first target feature in the first target feature set in the future period according to the influence of each first target feature on the disease development trend of the target site, the influence of each first target feature on the disease development trend of the target site, and the future period input by the user. Among them, the future change trend is used to indicate the prediction result of the change trend of each first target feature in the future period.
[0092] S302, obtaining a prediction result of the disease change trend in the future time period by the target time series prediction model based on the future change trend of each first target feature in the future time period.
[0093] Optionally, the target time series prediction model predicts future disease-related data of the target site in a future time period according to the future change trends of each first target feature, and obtains a prediction result of the disease change trend of the target site in the future time period based on the fitting of the future disease-related data.
[0094] For a possible implementation, see Figure 4 The operation of step S102 may specifically be:
[0095] S401. Clean the disease-related data within a preset historical period to obtain a first target disease data set within the preset historical period.
[0096] Optionally, data cleaning is performed on the disease-related data within a preset historical period, which can effectively reduce the noise of the disease-related data within the historical period. After data cleaning and noise reduction, a first target disease data set within the preset historical period is obtained.
[0097] Optionally, the first target disease data set is used to indicate the disease-related data after deduplication and noise reduction, that is, the first target disease data set is a data set consisting of valid data of the disease-related data.
[0098] S402: Extract features from the first target disease data set to obtain a first target feature set within a preset historical period.
[0099] Optionally, feature extraction is performed on the first target disease data set to obtain all first target features that affect the development trend of the disease within a preset historical period.
[0100] S403: Time-align the first target disease data set based on a preset standard time axis to obtain a first target time series data sequence within a preset historical period.
[0101] Optionally, the preset standard timeline is a standard timeline pre-set by the user. The preset standard timeline may be a 12-hour, 24-hour, or other timeline. This application takes the preset standard timeline of 24 hours as an example, but does not mean that the preset standard timeline can only select a 24-hour timeline. This application does not make any specific limitations on this.
[0102] Optionally, the first target disease data set is time-aligned based on a preset standard timeline. Time alignment refers to aligning the target disease data contained in the first target disease data set to a preset standard timeline, i.e., comparing the crack data at 13:00 on Monday with the crack data at 13:00 on Tuesday, etc.
[0103] Optionally, all disease-related data in the first target disease data set are aligned to a preset standard time axis to obtain a first target time series data sequence within a preset historical period, and the first target time series data sequence is a time-aligned disease-related data sequence.
[0104] For a possible implementation, see Figure 5 The operation of step S401 may specifically be:
[0105] S501. Determine missing values and invalid values in the disease-related data within a preset historical period according to the disease-related data within a preset historical period and the preset standard disease-related data.
[0106] Optionally, the preset standard disease-related data refers to a valid value database of disease-related data obtained by training based on historical disease-related data. By comparing the disease-related data in a preset historical period with the valid value database, missing values and invalid values in the disease-related data in the preset historical period can be determined. Among them, missing values refer to disease-related data lost at a time node on a preset standard time axis, and invalid values refer to data with a huge deviation between a certain disease-related data in a preset historical period and the preset standard disease-related data.
[0107] S502. According to preset standard disease-related data, invalid values in the disease-related data within a preset historical period are deleted, and missing values in the disease-related data within the preset historical period are filled to obtain a first target disease data set within the preset historical period.
[0108] Optionally, based on the valid values of disease-related data contained in preset standard disease-related data, invalid values in the disease-related data within the preset historical period are deleted, and the disease-related data within the preset historical period are filled with the valid values of the disease-related data indicated by the preset standard disease-related data, so as to obtain the first target disease data set with a complete time series within the preset historical period.
[0109] For a possible implementation, see Figure 6 After step S103, the above method for predicting the trend of site damage changes also includes:
[0110] S601. Determine the alarm level corresponding to the disease change trend prediction result according to the disease change trend prediction result and the preset disease weight value, and issue corresponding alarm information according to the alarm level.
[0111] Optionally, the preset disease weight value is the degree of impact of the disease set by the user based on the correlation between various first target features and the disease. As the climate and seasons change, the preset disease weight value also changes dynamically, and the user can adjust the preset disease weight value at any time.
[0112] Optionally, the alarm level is used to indicate the severity of the disease change trend. The alarm levels include level 1 alarm, level 2 alarm, level 3 alarm and level 4 alarm. Level 1 alarm is used to indicate the most serious disease alarm information that requires the most urgent treatment. The alarm level of level 1 alarm is higher than the alarm level of level 2 alarm. The alarm level of level 2 alarm is higher than the alarm level of level 3 alarm. The alarm level of level 3 alarm is higher than the alarm level of level 4 alarm. This application takes the alarm level including four levels as an example, but it does not mean that the alarm level can only be level 4. This application does not make specific limitations on this. Among them, the alarm information is used to indicate the warning content corresponding to each alarm level.
[0113] S602: Generate a disease prediction report based on the warning level and warning information corresponding to the disease change trend prediction result, and push a target protection measure sequence in descending order of the warning level.
[0114] Optionally, a disease prediction report is generated based on the alarm level and alarm content of the predicted results of the disease change trend of the target site in the future time period, and the disease prediction report includes the predicted results of the disease change trend of the target site in the future time period, the alarm level corresponding to the predicted results of the disease change trend, the alarm content corresponding to the alarm level, etc.
[0115] Optionally, the target protection measure sequence refers to the order of managing the disease change trend of the target site in the future period, and the target protection measure sequence is determined based on the alarm level corresponding to the prediction result of the disease change trend of the target site in the future period.
[0116] For a possible implementation, see Figure 7 The above-mentioned method for predicting the trend of site damage changes also includes:
[0117] S701. Obtain a historical disease-related data set.
[0118] Optionally, the historical disease-related data set is used to indicate disease-related data in a past time period. The historical disease-related data set may be a disease-related data set accumulated in the past one, two, or three years, and this application does not make any specific limitation on this.
[0119] S702. Preprocess the historical disease-related data set to obtain a second target disease data set corresponding to the historical disease-related data set, extract a second target feature set corresponding to the second target disease data set, and obtain a second target time series data sequence corresponding to the second target disease data set, wherein the second target feature set includes the first target feature set.
[0120] Optionally, the second target disease data set is used to indicate the disease-related data set obtained after deduplication and noise reduction of historical disease-related data; the second target feature set is used to indicate the target feature set of disease development trends that affect historical disease-related data; and the second target disease data set is time-aligned based on a preset standard time axis to obtain a second target time series data sequence.
[0121] Optionally, the second target feature set includes all the first target features included in the first target feature set, that is, the types of second target features included in the second target feature set are greater than or equal to the types of first target features included in the first target feature set.
[0122] S703: Segment the second target disease data set to obtain a training data set, a verification data set, and a test data set.
[0123] Optionally, the second target disease data set is divided into a training data set, a verification data set and a test data set according to a data segmentation rule. For example, the data segmentation rule may be to divide 40% of the second target disease data set into a training data set, 30% of the second target disease data set into a test data set, and 30% of the second target disease data set into a verification data set. This application does not make any specific limitations on this.
[0124] Optionally, the training data set is used to train and optimize the target time series prediction model, the test data set is used to test the performance of the target time series prediction model; and the verification data set is used to verify whether the target time series prediction model is qualified.
[0125] S704: Construct an initial time series prediction model based on the training data set, the third target feature set corresponding to the training data set, the third target time series data sequence corresponding to the training data set, and a preset deep learning model.
[0126] Optionally, the third target feature set refers to a set of influencing factors of the disease change trend corresponding to the disease-related data in the training data set, and the third target time series data sequence is used to indicate the target time series data sequence obtained after the valid values of the disease-related data in the training data set are aligned with the preset standard time axis.
[0127] Optionally, the preset deep learning model can be implemented by a neural network model, and the initial time series prediction model refers to a time series prediction model obtained through training and learning of a training data set.
[0128] S705. Iteratively verify the initial time series prediction model based on the verification data set, the fourth target feature set corresponding to the verification data set, and the fourth target time series data sequence corresponding to the verification data set, and determine the target weight value corresponding to each second target feature in the second target feature set.
[0129] Optionally, the fourth target feature set refers to a set of influencing factors of the disease change trend corresponding to the disease-related data in the verification data set, and the fourth target time series data sequence is used to indicate the target time series data sequence obtained after the valid values of the disease-related data in the verification data set are aligned with the preset standard time axis time.
[0130] Optionally, the validation data set can further optimize the parameters of the initial time series prediction model, and re-determine the target weight value corresponding to each second target feature in the second target feature set according to the optimal parameters of the initial time series prediction model.
[0131] S706. Iteratively test the initial time series prediction model according to the target weight values corresponding to each second target feature, the test data set, the fifth target feature set corresponding to the test data set, and the fifth target time series data sequence corresponding to the test data set, and determine the target time series prediction model according to the iterative test results.
[0132] Optionally, the fifth target feature set refers to a set of influencing factors of the disease change trend corresponding to the disease-related data in the test data set, and the fifth target time series data sequence is used to indicate the target time series data sequence obtained after the valid values of the disease-related data in the test data set are aligned with the preset standard time axis time.
[0133] Optionally, based on a first-level test data set of target weight values corresponding to each second target feature, the performance of the initial time series prediction model is iteratively tested, and the initial time series prediction model with the best performance is used as the target time series prediction model.
[0134] As an optional implementation, Figure 8 A visualization interface for disease change trend prediction results provided in this application, see Figure 8 , predict the disease change trend prediction results and alarm levels of each monitoring point in the target site in the next day, and the specific prediction process will not be described in detail in this application.
[0135] The following is a description of the apparatus, device, and computer-readable storage medium provided in the present application for predicting the trend of changes in site diseases. The specific implementation process and technical effects are described above and will not be repeated below.
[0136] Fig. 9 This is a schematic diagram of the structure of a device for predicting the trend of changes in site diseases provided by the present application, see Fig. 9 , the device comprises:
[0137] The acquisition module 901 is used to acquire disease-related data of the target site within a preset historical period, and the disease-related data includes: crack data, temperature data, and humidity data;
[0138] The data processing module 902 is used to perform data preprocessing on the disease-related data within the preset historical period, extract the first target feature set within the preset historical period, and obtain the first target time series data sequence within the preset historical period, wherein the first target feature set includes: at least one first target feature;
[0139] The prediction module 903 is used to determine the prediction results of the disease change trend of the target site in the future time period based on the first target time series data sequence, the first target feature set, the preset future time period and the target time series prediction model. The disease change trend prediction results include: crack propagation rate prediction results, temperature rise and fall amplitude prediction results and humidity increase and decrease amplitude prediction results.
[0140] As an optional implementation, the prediction module 903 is specifically used for:
[0141] Determine the historical change trend of each first target feature within a preset historical period according to the first target time series data sequence and the first target feature set;
[0142] Determine a weight value corresponding to each first target feature according to the target time series prediction model, the first target feature set, and the historical change trend of each first target feature;
[0143] The weight values corresponding to the first target features, the historical change trends of the first target features, and the future time periods are input into the target time series prediction model, and the target time series prediction model determines the prediction results of the disease change trends in the future time period.
[0144] As an optional implementation, the prediction module 903 is further configured to:
[0145] The target time series prediction model predicts the future change trend of each first target feature in the future time period based on the weight value corresponding to each first target feature, the historical change trend of each first target feature and the future time period;
[0146] The target time series prediction model is based on the future change trend of each first target feature in the future period to obtain the prediction result of the disease change trend in the future period.
[0147] As an optional implementation, the data processing module 902 is specifically used for:
[0148] Performing data cleaning on disease-related data within a preset historical period to obtain a first target disease data set within the preset historical period;
[0149] Performing feature extraction on the first target disease data set to obtain a first target feature set within a preset historical period;
[0150] The first target disease data set is time-aligned based on a preset standard time axis to obtain a first target time series data sequence within a preset historical period.
[0151] As an optional implementation, the data processing module 902 is further configured to:
[0152] Determine missing values and invalid values in the disease-related data within the preset historical period based on the disease-related data within the preset historical period and the preset standard disease-related data;
[0153] According to the preset standard disease-related data, invalid values in the disease-related data within the preset historical period are deleted, and missing values in the disease-related data within the preset historical period are filled to obtain a first target disease data set within the preset historical period.
[0154] As an optional implementation, the device for predicting the trend of site damage changes further includes an alarm push module 904, which is specifically used to:
[0155] According to the disease change trend prediction results and the preset disease weight value, determine the alarm level corresponding to the disease change trend prediction results, and issue corresponding alarm information according to the alarm level;
[0156] According to the warning level and warning information corresponding to the prediction results of disease change trend, a disease prediction report is generated, and a sequence of target protection measures is pushed in descending order according to the warning level.
[0157] As an optional implementation, the construction process of the prediction module 903 is specifically as follows:
[0158] Performing data preprocessing on the historical disease-related data set to obtain a second target disease data set corresponding to the historical disease-related data set, extracting a second target feature set corresponding to the second target disease data set, and obtaining a second target time series data sequence corresponding to the second target disease data set, wherein the second target feature set includes the first target feature set;
[0159] Performing data segmentation on the second target disease data set to obtain a training data set, a verification data set, and a test data set;
[0160] Constructing an initial time series prediction model according to the training data set, the third target feature set corresponding to the training data set, the third target time series data sequence corresponding to the training data set, and the preset deep learning model;
[0161] Iteratively verify the initial time series prediction model according to the verification data set, the fourth target feature set corresponding to the verification data set, and the fourth target time series data sequence corresponding to the verification data set, and determine the target weight value corresponding to each second target feature in the second target feature set;
[0162] According to the target weight values corresponding to each second target feature, the test data set, the fifth target feature set corresponding to the test data set, and the fifth target time series data sequence corresponding to the test data set, the initial time series prediction model is iteratively tested, and the target time series prediction model is determined according to the iterative test results.
[0163] The above-mentioned device is used to execute the method provided by the aforementioned embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.
[0164] The above modules may be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASICs), or one or more microprocessors, or one or more field programmable gate arrays (FPGAs). For another example, when a module is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).
[0165] Fig.10 is a schematic diagram of the structure of a computer device provided by this application. Fig.10 The computer device includes: a memory 1001 and a processor 1002. The memory 1001 stores a computer program that can be run on the processor 1002. When the processor 1002 executes the computer program, the steps in any of the above method embodiments are implemented.
[0166] The present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.
[0167] Optionally, the present application also provides a program product, such as a computer-readable storage medium, including a program, which, when executed by a processor, is used to execute any of the above-mentioned method embodiments for predicting the changing trend of site disease damage.
[0168] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0169] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0170] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional units.
[0171] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor (English: processor) to perform some steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (English: Read-Only Memory, abbreviated: ROM), random access memory (English: Random Access Memory, abbreviated: RAM), disk or optical disk and other media that can store program codes.
[0172] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
[0173] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims
1. A method for predicting the trend of changes in site damage, characterized in that: The method comprises: Obtaining disease-related data of the target site within a preset historical period, wherein the disease-related data includes: crack data, temperature data, and humidity data; Performing data preprocessing on disease-related data within a preset historical period, extracting a first target feature set within the preset historical period, and obtaining a first target time series data sequence within the preset historical period, wherein the first target feature set includes: at least one first target feature; According to the first target time series data sequence, the first target feature set, the preset future time period and the target time series prediction model, the prediction results of the disease change trend of the target site in the future time period are determined, and the prediction results of the disease change trend include: the prediction results of the crack propagation rate, the prediction results of the temperature rise and fall amplitude, and the prediction results of the humidity increase and decrease amplitude.
2. The method for predicting the trend of changes in site diseases according to claim 1, characterized in that: The step of determining the prediction result of the disease change trend of the target site in the future time period according to the first target time series data sequence, the first target feature set, the preset future time period, and the target time series prediction model includes: Determine, according to the first target time series data sequence and the first target feature set, a historical change trend of each first target feature within a preset historical period; Determine a weight value corresponding to each first target feature according to the target time series prediction model, the first target feature set, and the historical change trend of each first target feature; The weight values corresponding to the first target features, the historical change trends of the first target features and the future time period are input into the target time series prediction model, and the target time series prediction model determines the prediction result of the disease change trend in the future time period.
3. The method for predicting the changing trend of site damage according to claim 2, characterized in that: The prediction result of the disease change trend in the future period determined by the target time series prediction model includes: The target time series prediction model predicts the future change trend of each first target feature in the future time period based on the weight value corresponding to each first target feature, the historical change trend of each first target feature and the future time period; The target time series prediction model obtains a prediction result of the disease change trend in the future time period based on the future change trend of each first target feature in the future time period.
4. The method for predicting the changing trend of site damage according to claim 1, characterized in that: The data preprocessing is performed on the disease-related data within the preset historical period, the first target feature set within the preset historical period is extracted, and the first target time series data sequence within the preset historical period is obtained, including: Performing data cleaning on disease-related data within a preset historical period to obtain a first target disease data set within the preset historical period; Performing feature extraction on the first target disease data set to obtain a first target feature set within the preset historical period; The first target disease data set is time-aligned based on a preset standard time axis to obtain a first target time series data sequence within the preset historical period.
5. The method for predicting the trend of changes in site damage according to claim 4, characterized in that: The data cleaning of the disease-related data within the preset historical period is performed to obtain a first target disease data set within the preset historical period, including: Determine missing values and invalid values in the disease-related data within the preset historical period based on the disease-related data within the preset historical period and the preset standard disease-related data; According to the preset standard disease-related data, invalid values in the disease-related data within the preset historical period are deleted, and missing values in the disease-related data within the preset historical period are filled to obtain a first target disease data set within the preset historical period.
6. The method for predicting the changing trend of site damage according to claim 1, characterized in that: After determining the disease change trend prediction result according to the first target time series data sequence, the first target feature set, the preset future time period and the target time series prediction model, the method further includes: According to the disease change trend prediction result and the preset disease weight value, determine the alarm level corresponding to the disease change trend prediction result, and issue corresponding alarm information according to the alarm level; According to the alarm level and alarm information corresponding to the disease change trend prediction result, a disease prediction report is generated, and a target protection measure sequence is pushed in descending order according to the alarm level.
7. The method for predicting the changing trend of site damage according to any one of claims 1 to 6, characterized in that: The method further comprises: Obtain historical disease-related data sets; Performing data preprocessing on the historical disease-related data set to obtain a second target disease data set corresponding to the historical disease-related data set, extracting a second target feature set corresponding to the second target disease data set, and obtaining a second target time series data sequence corresponding to the second target disease data set, wherein the second target feature set includes the first target feature set; Performing data segmentation on the second target disease data set to obtain a training data set, a verification data set, and a test data set; Constructing an initial time series prediction model according to the training data set, the third target feature set corresponding to the training data set, the third target time series data sequence corresponding to the training data set, and a preset deep learning model; Iteratively verify the initial time series prediction model according to the verification data set, the fourth target feature set corresponding to the verification data set, and the fourth target time series data sequence corresponding to the verification data set, and determine the target weight value corresponding to each second target feature in the second target feature set; According to the target weight values corresponding to each second target feature, the test data set, the fifth target feature set corresponding to the test data set, and the fifth target time series data sequence corresponding to the test data set, the initial time series prediction model is iteratively tested, and the target time series prediction model is determined according to the test results.
8. A device for predicting the changing trend of site damage, characterized in that: The device comprises: An acquisition module is used to acquire disease-related data of a target site within a preset historical period, wherein the disease-related data includes: crack data, temperature data, and humidity data; A data processing module is used to perform data preprocessing on disease-related data within a preset historical period, extract a first target feature set within the preset historical period, and obtain a first target time series data sequence within the preset historical period, wherein the first target feature set includes: at least one first target feature; A prediction module is used to determine the prediction results of the disease change trend of the target site in the future time period based on the first target time series data sequence, the first target feature set, the preset future time period and the target time series prediction model, and the disease change trend prediction results include: crack propagation rate prediction results, temperature rise and fall amplitude prediction results and humidity increase and decrease amplitude prediction results.
9. A computer device, characterized in that: include: A memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, the steps of the method described in any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 7.