Building risk processing method and device based on building hidden danger treatment and meteorological risk prediction and early warning management, and computer equipment

Abstract

The invention relates to a building risk processing method and device based on building hidden danger treatment and meteorological risk prediction and early warning management, and computer equipment. The method comprises the steps of determining current risk severity, current risk occurrence probability and current risk exposure frequency of a to-be-analyzed building according to building structure data, equipment data and environment data of the to-be-analyzed building, and obtaining a first risk value through a trained first risk prediction model; inputting the current meteorological data of the to-be-analyzed building into the trained second risk prediction model to obtain a second risk value; performing fusion processing on the first risk value and the second risk value to obtain a target risk value of the to-be-analyzed building; and generating a risk processing instruction for the to-be-analyzed building according to the target risk value, and performing risk processing on the to-be-analyzed building according to the risk processing instruction. By adopting the method, the risk processing accuracy of the building can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a building risk processing method and device based on building hidden danger treatment and meteorological risk prediction and early warning management, computer equipment, computer readable storage medium and computer program product. BACKGROUND

[0002] At present, in order to prolong the service life of the building, how to accurately process the risk of the building is very important.

[0003] In the traditional technology, when processing the risk of the building, the artificial decision-making method is generally used. However, this artificial decision-making method has subjective factors and is prone to errors, resulting in low accuracy of risk processing of the building. SUMMARY

[0004] Therefore, it is necessary to provide a building risk processing method and device based on building hidden danger treatment and meteorological risk prediction and early warning management, computer equipment, computer readable storage medium and computer program product, which can improve the accuracy of risk processing of the building.

[0005] In a first aspect, the present application provides a building risk processing method based on building hidden danger treatment and meteorological risk prediction and early warning management, comprising:

[0006] In response to a risk processing request for a building to be analyzed, obtaining building structure data, equipment data and environment data of the building to be analyzed;

[0007] According to the building structure data, the equipment data and the environment data, determining the current risk type and the current risk position of the building to be analyzed;

[0008] According to the current risk type and the current risk position, determining the current risk severity, the current risk occurrence probability and the current risk exposure frequency of the building to be analyzed;

[0009] Inputting the current risk severity, the current risk occurrence probability and the current risk exposure frequency into a trained first risk prediction model to obtain a first risk value of the building to be analyzed;

[0010] Obtaining current meteorological data of the building to be analyzed, inputting the current meteorological data into a trained second risk prediction model to obtain a second risk value of the building to be analyzed;

[0011] In a case where the first risk value and the second risk value are both less than a preset risk value, the first risk value and the second risk value are fused according to a first model weight of the first risk prediction model and a second model weight of the second risk prediction model, to obtain a target risk value of the building to be analyzed;

[0012] According to the target risk value, a risk processing instruction for the building to be analyzed is generated, and the building to be analyzed is processed according to the risk processing instruction.

[0013] In one of the embodiments, the current risk type and the current risk position of the building to be analyzed are determined according to the building structure data, the equipment data and the environment data, comprising:

[0014] The building structure data, the equipment data and the environment data are input into a trained risk type prediction model to obtain a first prediction probability of the building to be analyzed under each preset risk type, and the building structure data, the equipment data and the environment data are input into a trained risk position prediction model to obtain a second prediction probability of the building to be analyzed under each preset risk position;

[0015] From the preset risk types, a preset risk type with the maximum first prediction probability is selected as the current risk type, and from the preset risk positions, a preset risk position with the maximum second prediction probability is selected as the current risk position.

[0016] In one of the embodiments, the first risk prediction model comprises a first risk prediction network corresponding to the current risk severity, a second risk prediction network corresponding to the current risk occurrence probability, a third risk prediction network corresponding to the current risk exposure frequency, and an attention mechanism network.

[0017] The current risk severity, the current risk occurrence probability and the current risk exposure frequency are input into the trained first risk prediction model to obtain the first risk value of the building to be analyzed, comprising:

[0018] The current risk severity, the current risk occurrence probability and the current risk exposure frequency are respectively subjected to feature extraction processing to obtain a first feature vector of the current risk severity, a second feature vector of the current risk occurrence probability and a third feature vector of the current risk exposure frequency.

[0019] inputting the first feature vector into the first risk prediction network to obtain a third risk value of the building to be analyzed, inputting the second feature vector into the second risk prediction network to obtain a fourth risk value of the building to be analyzed, and inputting the third feature vector into the third risk prediction network to obtain a fifth risk value of the building to be analyzed;

[0020] inputting the third risk value, the fourth risk value and the fifth risk value into the attention mechanism network respectively to obtain a first weight corresponding to the third risk value, a second weight corresponding to the fourth risk value and a third weight corresponding to the fifth risk value;

[0021] performing summation processing on the third risk value, the fourth risk value and the fifth risk value according to the first weight, the second weight and the third weight to obtain a first risk value of the building to be analyzed.

[0022] In one of the embodiments, the inputting the current meteorological data into the trained second risk prediction model to obtain the second risk value of the building to be analyzed comprises:

[0023] performing combination processing on the current meteorological data to obtain combination data corresponding to the current meteorological data;

[0024] performing feature extraction processing on the current meteorological data and the combination data respectively to obtain a fourth feature vector of the current meteorological data and a fifth feature vector of the combination data;

[0025] performing splicing processing on the fourth feature vector and the fifth feature vector to obtain a spliced feature vector;

[0026] inputting the spliced feature vector into the trained second risk prediction model to obtain the second risk value of the building to be analyzed.

[0027] In one of the embodiments, before the fusion processing on the first risk value and the second risk value according to the first model weight of the first risk prediction model and the second model weight of the second risk prediction model to obtain the target risk value of the building to be analyzed, the method further comprises:

[0028] obtaining a first prediction accuracy of the first risk prediction model and a second prediction accuracy of the second risk prediction model;

[0029] corresponding relationship between the prediction accuracy and the model weight, obtain the model weight corresponding to the first prediction accuracy as a first initial model weight of the first risk prediction model, and query the corresponding relationship to obtain the model weight corresponding to the second prediction accuracy as a second initial model weight of the second risk prediction model;

[0030] normalize the first initial model weight and the second initial model weight to obtain a first model weight of the first risk prediction model and a second model weight of the second risk prediction model.

[0031] In one of the embodiments, the generating, according to the target risk value, of the risk processing instruction for the building to be analyzed comprises:

[0032] determining, according to the target risk value, a current risk level of the building to be analyzed;

[0033] obtaining a historical risk level of the building to be analyzed, and performing fusion processing on the current risk level and the historical risk level to obtain a target risk level of the building to be analyzed;

[0034] generating, according to the target risk level, a risk processing instruction for the building to be analyzed.

[0035] In a second aspect, the present application further provides a building risk processing device based on building hidden danger governance and meteorological risk prediction and early warning management, comprising:

[0036] a data acquisition module configured to, in response to a risk processing request for a building to be analyzed, acquire building structure data, equipment data and environment data of the building to be analyzed;

[0037] a first determination module configured to determine, according to the building structure data, the equipment data and the environment data, a current risk type and a current risk location of the building to be analyzed;

[0038] a second determination module configured to determine, according to the current risk type and the current risk location, a current risk severity, a current risk occurrence probability and a current risk exposure frequency of the building to be analyzed;

[0039] a first prediction module configured to input the current risk severity, the current risk occurrence probability and the current risk exposure frequency into a trained first risk prediction model to obtain a first risk value of the building to be analyzed;

[0040] a second prediction module, configured to acquire current meteorological data of the building to be analyzed, and input the current meteorological data into the trained second risk prediction model to obtain a second risk value of the building to be analyzed;

[0041] a risk fusion module, configured to, in a case where both the first risk value and the second risk value are less than a preset risk value, perform fusion processing on the first risk value and the second risk value according to a first model weight of the first risk prediction model and a second model weight of the second risk prediction model to obtain a target risk value of the building to be analyzed;

[0042] a risk processing module, configured to generate a risk processing instruction for the building to be analyzed according to the target risk value, and perform corresponding risk processing on the building to be analyzed according to the risk processing instruction.

[0043] In a third aspect, the present application further provides a computer device, comprising a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0044] in response to a risk processing request for a building to be analyzed, acquiring building structure data, equipment data and environment data of the building to be analyzed;

[0045] determining a current risk type and a current risk location of the building to be analyzed according to the building structure data, the equipment data and the environment data;

[0046] determining a current risk severity, a current risk occurrence probability and a current risk exposure frequency of the building to be analyzed according to the current risk type and the current risk location;

[0047] inputting the current risk severity, the current risk occurrence probability and the current risk exposure frequency into a trained first risk prediction model to obtain a first risk value of the building to be analyzed;

[0048] acquiring current meteorological data of the building to be analyzed, and inputting the current meteorological data into a trained second risk prediction model to obtain a second risk value of the building to be analyzed;

[0049] in a case where both the first risk value and the second risk value are less than a preset risk value, performing fusion processing on the first risk value and the second risk value according to a first model weight of the first risk prediction model and a second model weight of the second risk prediction model to obtain a target risk value of the building to be analyzed;

[0050] According to the target risk value, a risk processing instruction for the building to be analyzed is generated, and corresponding risk processing is performed on the building to be analyzed according to the risk processing instruction.

[0051] In a fourth aspect, the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the following steps:

[0052] In response to a risk processing request for a building to be analyzed, building structure data, equipment data and environment data of the building to be analyzed are acquired;

[0053] According to the building structure data, the equipment data and the environment data, a current risk type and a current risk location of the building to be analyzed are determined;

[0054] According to the current risk type and the current risk location, a current risk severity, a current risk occurrence probability and a current risk exposure frequency of the building to be analyzed are determined;

[0055] The current risk severity, the current risk occurrence probability and the current risk exposure frequency are input into a trained first risk prediction model to obtain a first risk value of the building to be analyzed;

[0056] Current meteorological data of the building to be analyzed are acquired, and the current meteorological data are input into a trained second risk prediction model to obtain a second risk value of the building to be analyzed;

[0057] In a case where the first risk value and the second risk value are both less than a preset risk value, the first risk value and the second risk value are fused according to a first model weight of the first risk prediction model and a second model weight of the second risk prediction model to obtain a target risk value of the building to be analyzed;

[0058] According to the target risk value, a risk processing instruction for the building to be analyzed is generated, and corresponding risk processing is performed on the building to be analyzed according to the risk processing instruction.

[0059] In a fifth aspect, the present application further provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement the following steps:

[0060] In response to a risk processing request for a building to be analyzed, building structure data, equipment data and environment data of the building to be analyzed are acquired;

[0061] determine a current risk type and a current risk location of the building to be analyzed according to the building structure data, the equipment data and the environment data;

[0062] determine a current risk severity, a current risk occurrence probability and a current risk exposure frequency of the building to be analyzed according to the current risk type and the current risk location;

[0063] input the current risk severity, the current risk occurrence probability and the current risk exposure frequency into a trained first risk prediction model to obtain a first risk value of the building to be analyzed;

[0064] obtain current weather data of the building to be analyzed, input the current weather data into a trained second risk prediction model to obtain a second risk value of the building to be analyzed;

[0065] in a case where both the first risk value and the second risk value are less than a preset risk value, perform fusion processing on the first risk value and the second risk value according to a first model weight of the first risk prediction model and a second model weight of the second risk prediction model to obtain a target risk value of the building to be analyzed;

[0066] generate a risk processing instruction for the building to be analyzed according to the target risk value, and perform corresponding risk processing on the building to be analyzed according to the risk processing instruction.

[0067] The aforementioned building risk management method, device, computer equipment, storage medium, and computer program product based on building hazard management and meteorological risk prediction and early warning management first responds to a risk management request for the building to be analyzed by acquiring the building structure data, equipment data, and environmental data of the building to be analyzed. Based on the building structure data, equipment data, and environmental data, it determines the current risk type and current risk location of the building to be analyzed. Then, based on the current risk type and current risk location, it determines the current risk severity, current risk occurrence probability, and current risk exposure frequency of the building to be analyzed. Finally, it inputs the current risk severity, current risk occurrence probability, and current risk exposure frequency into the trained... The first risk prediction model obtains the first risk value of the building to be analyzed. Then, the current meteorological data of the building to be analyzed is obtained and input into the trained second risk prediction model to obtain the second risk value of the building to be analyzed. Then, if both the first risk value and the second risk value are less than the preset risk value, the first risk value and the second risk value are fused according to the first model weight of the first risk prediction model and the second model weight of the second risk prediction model to obtain the target risk value of the building to be analyzed. Finally, based on the target risk value, a risk handling instruction for the building to be analyzed is generated, and the corresponding risk handling is performed on the building to be analyzed according to the risk handling instruction. In this way, when performing risk management on buildings, by analyzing the building's structural data, equipment data, and environmental data, the current risk type and location of the building under analysis can be accurately determined. This allows for the accurate determination of the current risk severity, probability of occurrence, and frequency of exposure. The trained first risk prediction model can accurately obtain the first risk value of the building. Combining this with the second risk value obtained from the trained second risk prediction model based on the building's current meteorological data, the target risk value can be determined more accurately. This, in turn, allows for the more precise generation of risk management instructions for the building, facilitating more accurate risk management and improving the accuracy of risk management. Furthermore, the entire process requires no human intervention, avoiding the subjective factors and errors inherent in manually generated decisions that can lead to lower accuracy in risk management, further enhancing the overall accuracy of risk management. Attached Figure Description

[0068] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the accompanying drawings needed to be used in the description of the embodiments of the present application or the related art will be briefly introduced. Obviously, the accompanying drawings in the following description only only some embodiments of the present application, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0069] Figure 1 The flowchart of a building risk processing method based on building hidden danger management and meteorological risk prediction and early warning management in an embodiment is shown in the figure.

[0070] Figure 2 The flowchart of a step of determining the current risk type and the current risk position of the building to be analyzed in an embodiment is shown in the figure.

[0071] Figure 3 The flowchart of a building risk processing method based on building hidden danger management and meteorological risk prediction and early warning management in another embodiment is shown in the figure.

[0072] Figure 4 The structural block diagram of a building risk processing device based on building hidden danger management and meteorological risk prediction and early warning management in an embodiment is shown in the figure.

[0073] Figure 5 The internal structure diagram of a computer device in an embodiment is shown in the figure. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below in combination with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0075] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0076] In an exemplary embodiment, as Figure 1As shown, a building risk processing method based on building hidden danger governance and meteorological risk prediction and early warning management is provided, and the embodiment is exemplified by applying the method to a server; it can be understood that the method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and is realized through interaction between the terminal and the server. The terminal can be, but is not limited to, various personal computers, notebook computers, smart phones and tablet computers; the server can be realized by an independent server or a server cluster composed of multiple servers. In the embodiment, the method includes the following steps:

[0077] In step S101, in response to a risk processing request for a building to be analyzed, building structure data, equipment data and environment data of the building to be analyzed are acquired.

[0078] The building to be analyzed refers to a building that needs to be risk evaluated and risk processed.

[0079] The risk processing request refers to a request for risk processing of the building to be analyzed.

[0080] The building structure data is used to represent data of a structure state of the building to be analyzed, including structure types (such as frame structure, brick-concrete structure), component parameters (such as beam / column size, wall thickness, foundation depth), structure damage information (such as crack position and length, deformation degree, steel corrosion condition).

[0081] The equipment data is used to represent data related to the running state of facility equipment in the building to be analyzed, including electrical equipment (such as line load, switch state, aging degree), water supply and drainage equipment (such as pipeline pressure, leakage monitoring, water pump operation parameter), fire-fighting equipment (such as smoke alarm state, fire extinguisher expiration date).

[0082] The environment data includes data of an internal environment (such as indoor temperature, humidity, harmful gas concentration) and an external environment (such as surrounding topography, drainage condition, surrounding building construction vibration, underground water level change, etc.) of the building to be analyzed.

[0083] Exemplarily, the server receives a risk processing request for a building to be analyzed sent by a terminal through a network channel between the server and the terminal, and performs rationality verification on the risk processing request to obtain a rationality verification result of the risk processing request; in a case where the rationality verification result of the risk processing request indicates that the risk processing request passes, the server acquires a building identifier corresponding to the building to be analyzed in response to the risk processing request; then, the server determines a building information model corresponding to the building identifier, and extracts building structure data of the building to be analyzed from the building information model; then, the server determines a device identifier matching the building identifier, and acquires device data corresponding to the device identifier from a database as device data of the building to be analyzed; then, the server collects internal environment data of the building to be analyzed through an internal environment sensor (such as an indoor temperature and humidity sensor) associated with the building to be analyzed, and acquires external environment data of the building to be analyzed through a geographic information system associated with the building to be analyzed, and takes the internal environment data and the external environment data of the building to be analyzed as environment data of the building to be analyzed.

[0084] In step S102, a current risk type and a current risk position of the building to be analyzed are determined according to the building structure data, the device data and the environment data.

[0085] The current risk type refers to a risk type to which the building to be analyzed belongs at a current time.

[0086] The current risk position refers to a specific spatial position of the building to be analyzed at the current time where the risk occurs.

[0087] Exemplarily, the server performs risk identification processing on the building structure data, the equipment data and the environment data to obtain a risk identification result; in a case where the risk identification result indicates that there is abnormal building structure data in the building structure data, the server takes a risk type (also referred to as a structure type risk, such as “damage to load-bearing member”) corresponding to the abnormal building structure data as a current risk type of the building to be analyzed, and determines a spatial position corresponding to the abnormal building structure data from a building information model of the building to be analyzed as a current risk position of the building to be analyzed; in a case where the risk identification result indicates that there is no abnormal building structure data in the building structure data, and there is abnormal equipment data in the equipment data, the server takes a risk type (also referred to as an equipment type risk, such as “electrical system failure”) corresponding to the abnormal equipment data as the current risk type of the building to be analyzed, and determines a spatial position corresponding to the abnormal equipment data from the building information model of the building to be analyzed as the current risk position of the building to be analyzed; in a case where the risk identification result indicates that there is no abnormal building structure data in the building structure data, there is no abnormal equipment data in the equipment data, and there is abnormal environment data in the environment data, the server takes a risk type (also referred to as an environment type risk, such as “damage caused by high humidity”) corresponding to the abnormal environment data as the current risk type of the building to be analyzed, and determines a spatial position corresponding to the abnormal environment data from the building information model of the building to be analyzed as the current risk position of the building to be analyzed; in a case where the risk identification result indicates that there is no abnormal building structure data in the building structure data, there is no abnormal equipment data in the equipment data, and there is no abnormal environment data in the environment data, the server determines that the building to be analyzed has no risk.

[0088] In step S103, the current risk severity, the current risk occurrence probability and the current risk exposure frequency of the building to be analyzed are determined according to the current risk type and the current risk position.

[0089] The current risk severity is used to represent the severity level of damage or consequences caused by the current risk type at the current risk position to the building to be analyzed, which is usually measured from the dimensions of personnel safety, property loss, function influence, etc.

[0090] The current risk occurrence probability is used to represent the possibility of the current risk type occurring at the current risk position, which is usually represented by percentage or level (such as “extremely high, high, medium, low”).

[0091] The current risk exposure frequency is used to represent the frequency of personnel exposure to the risk at the current risk position when the current risk type occurs at the current risk position.

[0092] Exemplarily, the server inputs the current risk type into a plurality of risk severity prediction models to obtain a risk severity output by each risk severity prediction model, and fuses the risk severity output by each risk severity prediction model to obtain a current risk severity of the building to be analyzed; then, the server determines a building having a similarity greater than a preset similarity with the building to be analyzed as a relevant building corresponding to the building to be analyzed, and takes a historical risk occurrence probability corresponding to the current risk type in the relevant building as a current risk occurrence probability of the building to be analyzed (for example, buildings of the same structure type and service life have consistent occurrence probabilities of the same type of risk).

[0093] Step S104, inputting the current risk severity, the current risk occurrence probability and the current risk exposure frequency into the trained first risk prediction model to obtain a first risk value of the building to be analyzed.

[0094] The first risk prediction model refers to a network model capable of obtaining a risk value of the building to be analyzed by using the current risk severity, the current risk occurrence probability and the current risk exposure frequency of the building to be analyzed, such as a convolutional neural network model.

[0095] The first risk value refers to a risk value of the building to be analyzed obtained based on the current risk severity, the current risk occurrence probability and the current risk exposure frequency of the building to be analyzed.

[0096] Exemplarily, the server respectively performs feature extraction processing on the current risk severity, the current risk occurrence probability and the current risk exposure frequency to obtain a first feature vector of the current risk severity, a second feature vector of the current risk occurrence probability and a third feature vector of the current risk exposure frequency; then, the server performs fusion processing on the first feature vector, the second feature vector and the third feature vector to obtain a fusion feature vector of the building to be analyzed; and then, the server inputs the fusion feature vector into the trained first risk prediction model to obtain the first risk value of the building to be analyzed.

[0097] Step S105, obtaining current meteorological data of the building to be analyzed, inputting the current meteorological data into the trained second risk prediction model to obtain a second risk value of the building to be analyzed.

[0098] The current meteorological data refers to real-time meteorological data related to the area where the building to be analyzed is located, including temperature (extreme high or low temperature), precipitation (heavy rain, heavy rain), wind speed and direction (typhoon, strong wind), humidity (high humidity causing mold or structural dampness), lightning intensity, snow depth, frost warning and other data.

[0099] The second risk prediction model refers to a network model capable of obtaining a risk value of the building to be analyzed using the current meteorological data of the building to be analyzed, such as a random forest model.

[0100] The second risk value refers to the risk value of the building to be analyzed based on the current meteorological data of the building to be analyzed.

[0101] Exemplarily, the server obtains the building identifier corresponding to the building to be analyzed, and obtains the real-time meteorological data corresponding to the building identifier from the meteorological database as the current meteorological data of the building to be analyzed; then, the server performs preprocessing on the current meteorological data, such as removing outliers, unifying data formats, and aligning space-time, to obtain preprocessed meteorological data; then, the server identifies the current data type of the preprocessed meteorological data, queries the corresponding relationship between the data type and the feature extraction model, obtains the feature extraction model corresponding to the current data type of the preprocessed meteorological data as the target feature extraction model corresponding to the preprocessed meteorological data; then, the server inputs the preprocessed meteorological data into the target feature extraction model for feature extraction processing to obtain the feature vector of the preprocessed meteorological data; then, the server inputs the feature vector of the preprocessed meteorological data into the trained second risk prediction model, and obtains the second risk value of the building to be analyzed through the second risk prediction model.

[0102] In step S106, in the case that the first risk value and the second risk value are both less than the preset risk value, the first risk value and the second risk value are fused according to the first model weight of the first risk prediction model and the second model weight of the second risk prediction model to obtain the target risk value of the building to be analyzed.

[0103] The preset risk value is used to represent a threshold value for judging whether the risk value is in a fusion range. It should be noted that only when the first risk value and the second risk value are both lower than the threshold value, it is considered that both types of risks are in a controllable state, and the premise of fusion calculation is met (if any risk value exceeds the preset value, it means that there is a high risk hidden danger, and the single high risk needs to be disposed of preferentially, without fusion).

[0104] The first model weight refers to the model weight of the first risk prediction model.

[0105] The second model weight refers to the model weight of the second risk prediction model.

[0106] wherein the target risk value is a risk value obtained by fusing the first risk value and the second risk value.

[0107] Exemplarily, the server determines the first model weight of the first risk prediction model and the second model weight of the second risk prediction model when both the first risk value and the second risk value are less than the preset risk value, and performs summation processing on the first risk value and the second risk value according to the first model weight and the second model weight to obtain the target risk value of the building to be analyzed.

[0108] Further, the server takes the first risk value as the target risk value of the building to be analyzed when the first risk value is greater than the preset risk value and the second risk value is less than the preset risk value, takes the second risk value as the target risk value of the building to be analyzed when the first risk value is less than the preset risk value and the second risk value is greater than the preset risk value, and takes the maximum risk value between the first risk value and the second risk value as the target risk value of the building to be analyzed when both the first risk value and the second risk value are greater than the preset risk value.

[0109] Step S107, generating a risk processing instruction for the building to be analyzed according to the target risk value, and performing corresponding risk processing on the building to be analyzed according to the risk processing instruction.

[0110] wherein the risk processing instruction is an instruction for performing risk processing on the building to be analyzed.

[0111] Exemplarily, the server generates a risk processing instruction for the building to be analyzed according to the target risk value, then performs integrity verification on the risk processing instruction to obtain an integrity verification result of the risk processing request, and performs corresponding risk processing on the building to be analyzed according to the risk processing instruction when the integrity verification result of the risk processing request indicates that the risk processing request passes.

[0112] ​In the building risk processing method based on building hidden danger management and meteorological risk prediction and early warning management, in response to a risk processing request for a building to be analyzed, building structure data, equipment data and environment data of the building to be analyzed are acquired, and current risk type and current risk position of the building to be analyzed are determined according to the building structure data, the equipment data and the environment data. Then, current risk severity, current risk occurrence probability and current risk exposure frequency of the building to be analyzed are determined according to the current risk type and the current risk position, and the current risk severity, the current risk occurrence probability and the current risk exposure frequency are input into the first risk prediction model to obtain a first risk value of the building to be analyzed. Next, current meteorological data of the building to be analyzed are acquired, and the current meteorological data are input into the second risk prediction model to obtain a second risk value of the building to be analyzed. Then, in a case where the first risk value and the second risk value are less than a preset risk value, the first risk value and the second risk value are fused according to a first model weight of the first risk prediction model and a second model weight of the second risk prediction model to obtain a target risk value of the building to be analyzed. Finally, a risk processing instruction for the building to be analyzed is generated according to the target risk value, and the building to be analyzed is processed according to the risk processing instruction. In this way, when processing the building, the current risk type and the current risk position of the building to be analyzed can be accurately determined by analyzing the building structure data, the equipment data and the environment data of the building, so that the current risk severity, the current risk occurrence probability and the current risk exposure frequency of the building to be analyzed can be accurately determined. The first risk value of the building to be analyzed can be accurately obtained through the trained first risk prediction model, and the target risk value of the building to be analyzed can be more accurately determined by combining the second risk value of the building to be analyzed obtained based on the current meteorological data of the building to be analyzed and the trained second risk prediction model, so that the risk processing instruction for the building to be analyzed can be more accurately generated, which is beneficial to more accurately processing the building to be analyzed, thereby improving the risk processing accuracy of the building. Moreover, the whole process does not need manual intervention, which avoids the subjective factors in the decision-making method by using manual generation, and is less likely to cause errors, thereby further improving the risk processing accuracy of the building.

[0113] In one example embodiment, as shown in FIG. 1, the step S102 of determining the current risk type and the current risk position of the building to be analyzed according to the building structure data, the equipment data and the environment data includes the following steps. Figure 2

[0114] ​Step S201, input the building structure data, equipment data and environment data into the trained risk type prediction model to obtain the first prediction probability of the to-be-analyzed building under each preset risk type, and input the building structure data, equipment data and environment data into the trained risk position prediction model to obtain the second prediction probability of the to-be-analyzed building under each preset risk position.

[0115] Step S202, from each preset risk type, the preset risk type with the maximum first prediction probability is selected as the current risk type, and from each preset risk position, the preset risk position with the maximum second prediction probability is selected as the current risk position.

[0116] Among them, the risk type prediction model refers to a network model capable of predicting the risk type of the to-be-analyzed building by using the building structure data, equipment data and environment data of the to-be-analyzed building, such as a support vector machine model.

[0117] Among them, the preset risk type refers to a pre-set risk type, including a structure type risk, an equipment type risk and an environment type risk. It should be noted that the preset risk type is determined according to the situation.

[0118] Among them, the first prediction probability refers to the possibility that the risk type prediction model determines the preset risk type to be correct.

[0119] Among them, the risk position prediction model refers to a network model capable of predicting the risk position of the to-be-analyzed building by using the building structure data, equipment data and environment data of the to-be-analyzed building, such as a clustering model.

[0120] Among them, the preset risk position refers to a pre-defined set of specific spatial positions in the to-be-analyzed building where risks may occur, which can be determined based on the building structure and functional zoning of the to-be-analyzed building, such as "3-layer east facade load-bearing wall", "1-layer power distribution room A area power distribution box", etc.

[0121] Among them, the second prediction probability refers to the possibility that the risk position prediction model determines the preset risk position to be correct.

[0122] Exemplarily, the server denoises the building structure data, the equipment data and the environment data to obtain denoised building structure data, denoised equipment data and denoised environment data; then, the server extracts a feature vector of the denoised building structure data, a feature vector of the denoised equipment data and a feature vector of the denoised environment data, and performs projection fusion processing on the feature vector of the denoised building structure data, the feature vector of the denoised equipment data and the feature vector of the denoised environment data to obtain a corresponding one-dimensional feature value (it should be noted that core information is retained in the projection fusion process, such as selecting principal components with higher cumulative contribution rate through principal component analysis); then, the server inputs the one-dimensional feature value into the trained risk type prediction model to obtain a first prediction probability of the building under analysis under each preset risk type, and inputs the one-dimensional feature value into the trained risk position prediction model to obtain a second prediction probability of the building under analysis under each preset risk position; then, the preset risk type with the largest first prediction probability is selected from each preset risk type, and the preset risk type is taken as the current risk type, and the preset risk position with the largest second prediction probability is selected from each preset risk position, and the preset risk position is taken as the current risk position.

[0123] In this embodiment, by inputting the building structure, equipment and environment data into the trained risk type and position prediction model, the prediction probability of each preset type and position is output, and the item with the largest probability is selected as the current risk type and position, the deep mining ability of the model for complex data is utilized, the accurate identification of the risk type and position is realized, and the subjectivity and limitation of manual judgment are avoided.

[0124] In one exemplary embodiment, the first risk prediction model includes a first risk prediction network corresponding to the current risk severity, a second risk prediction network corresponding to the current risk occurrence probability, a third risk prediction network corresponding to the current risk exposure frequency, and an attention mechanism network.

[0125] Then, the step S104 inputs the current risk severity, the current risk occurrence probability and the current risk exposure frequency into the trained first risk prediction model to obtain the first risk value of the building to be analyzed, and specifically includes the following contents: performing feature extraction processing on the current risk severity, the current risk occurrence probability and the current risk exposure frequency respectively to obtain a first feature vector of the current risk severity, a second feature vector of the current risk occurrence probability and a third feature vector of the current risk exposure frequency; inputting the first feature vector into the first risk prediction network to obtain a third risk value of the building to be analyzed, inputting the second feature vector into the second risk prediction network to obtain a fourth risk value of the building to be analyzed, and inputting the third feature vector into the third risk prediction network to obtain a fifth risk value of the building to be analyzed; inputting the third risk value, the fourth risk value and the fifth risk value into the attention mechanism network respectively to obtain a first weight corresponding to the third risk value, a second weight corresponding to the fourth risk value and a third weight corresponding to the fifth risk value; and performing summation processing on the third risk value, the fourth risk value and the fifth risk value according to the first weight, the second weight and the third weight to obtain the first risk value of the building to be analyzed.

[0126] The first risk prediction network refers to a sub-network in the first risk prediction model for processing the current risk severity.

[0127] The second risk prediction network refers to a sub-network in the first risk prediction model for processing the current risk occurrence probability.

[0128] The third risk prediction network refers to a sub-network in the first risk prediction model for processing the current risk exposure frequency.

[0129] The attention mechanism network refers to a sub-network in the first risk prediction model for dynamically allocating weights.

[0130] The first feature vector is used to represent a feature vector of the current risk severity.

[0131] The second feature vector is used to represent a feature vector of the current risk occurrence probability.

[0132] The third feature vector is used to represent a feature vector of the current risk exposure frequency.

[0133] The third risk value refers to a risk value corresponding to the current risk severity.

[0134] The fourth risk value refers to a risk value corresponding to the current risk occurrence probability.

[0135] The fifth risk value refers to a risk value corresponding to the current risk exposure frequency.

[0136] The first weight corresponds to the third risk value.

[0137] The second weight corresponds to the fourth risk value.

[0138] The third weight corresponds to the fifth risk value.

[0139] Exemplarily, the server inputs the current risk severity as main data and the current risk occurrence probability and the current risk exposure frequency as auxiliary data into the feature extraction model for feature extraction processing to obtain a first feature vector of the current risk severity; then, the server inputs the current risk occurrence probability as main data and the current risk severity and the current risk exposure frequency as auxiliary data into the feature extraction model for feature extraction processing to obtain a second feature vector of the current risk occurrence probability; then, the server inputs the current risk exposure frequency as main data and the current risk severity and the current risk occurrence probability as auxiliary data into the feature extraction model for feature extraction processing to obtain a third feature vector of the current risk exposure frequency; then, the server inputs the first feature vector into the first risk prediction network to obtain the third risk value of the building to be analyzed, inputs the second feature vector into the second risk prediction network to obtain the fourth risk value of the building to be analyzed, and inputs the third feature vector into the third risk prediction network to obtain the fifth risk value of the building to be analyzed; then, the server inputs the third risk value, the fourth risk value and the fifth risk value into the attention mechanism network respectively to obtain the first weight corresponding to the third risk value, the second weight corresponding to the fourth risk value and the third weight corresponding to the fifth risk value; then, the server performs summation processing on the third risk value, the fourth risk value and the fifth risk value according to the first weight, the second weight and the third weight to obtain the first risk value of the building to be analyzed.

[0140] In this embodiment, by performing feature extraction and targeted network prediction on the three core dimensions of the current risk severity, the current risk occurrence probability and the current risk exposure frequency, the unique risk features of each dimension can be accurately captured, feature confusion caused by single network processing can be avoided, and the attention mechanism is introduced to dynamically allocate the weights of the output values of each dimension, thereby ensuring the comprehensiveness of risk assessment.

[0141] In one exemplary embodiment, the step S105 of inputting the current meteorological data into the trained second risk prediction model to obtain the second risk value of the building to be analyzed comprises the following contents: performing combination processing on the current meteorological data to obtain combination data corresponding to the current meteorological data; performing feature extraction processing on the current meteorological data and the combination data respectively to obtain a fourth feature vector of the current meteorological data and a fifth feature vector of the combination data; performing splicing processing on the fourth feature vector and the fifth feature vector to obtain a spliced feature vector; and inputting the spliced feature vector into the trained second risk prediction model to obtain the second risk value of the building to be analyzed.

[0142] The combination data refers to derivative data obtained by performing combination operation on the current meteorological data, and is used to supplement the associated features that cannot be directly reflected by the original data.

[0143] The fourth feature vector is used to represent the representation vector of the current meteorological data.

[0144] The fifth feature vector is used to represent the representation vector of the combination data.

[0145] The spliced feature vector refers to a feature vector obtained by splicing the fourth feature vector and the fifth feature vector.

[0146] For example, the server determines the combination mode corresponding to the current meteorological data; the combination mode is used to represent the rule corresponding to the specific combination processing of the current meteorological data; for example, taking "precipitation amount x continuous rainfall duration" as the combination data reflecting the cumulative rainfall intensity, "wind speed x building wind area coefficient" as the combination data reflecting the actual wind load, or "humidity x temperature" as the combination data reflecting the mold risk, etc.; then, the server determines the to-be-combined meteorological data corresponding to the current meteorological data according to the combination mode; then, the server performs combination processing on the current meteorological data and the to-be-combined meteorological data to obtain the combination data corresponding to the current meteorological data; then, the server inputs the current meteorological data as the main data and the combination data as the auxiliary data into the feature extraction model to perform feature extraction processing to obtain the fourth feature vector of the current meteorological data; then, the server inputs the combination data as the main data and the current meteorological data as the auxiliary data into the feature extraction model to perform feature extraction processing to obtain the fifth feature vector of the combination data; then, the server performs splicing processing on the fourth feature vector and the fifth feature vector according to a preset splicing order to obtain the spliced feature vector; and then, the server inputs the spliced feature vector into the trained second risk prediction model to obtain the second risk value of the building to be analyzed through the second risk prediction model.

[0147] In the embodiment, the combined data is generated by combining the current meteorological data, the correlation between meteorological elements can be mined, the limitation that the original meteorological data can only reflect a single element is made up, the second risk prediction model is trained, the comprehensive risk influence of the meteorological factors on the building can be more accurately captured, and the accuracy and reliability of the risk assessment of the building related to the weather are improved.

[0148] In an exemplary embodiment, before the step S106, the first risk value and the second risk value are fused according to the first model weight of the first risk prediction model and the second model weight of the second risk prediction model to obtain the target risk value of the building to be analyzed, specifically including the following contents: obtaining the first prediction accuracy of the first risk prediction model and the second prediction accuracy of the second risk prediction model; querying the corresponding relationship between the prediction accuracy and the model weight to obtain the model weight corresponding to the first prediction accuracy as the first initial model weight of the first risk prediction model, and querying the corresponding relationship to obtain the model weight corresponding to the second prediction accuracy as the second initial model weight of the second risk prediction model; and normalizing the first initial model weight and the second initial model weight to obtain the first model weight of the first risk prediction model and the second model weight of the second risk prediction model.

[0149] The first prediction accuracy refers to the prediction accuracy of the first risk prediction model.

[0150] The second prediction accuracy refers to the prediction accuracy of the second risk prediction model.

[0151] The corresponding relationship between the prediction accuracy and the model weight is used to represent the mapping relationship between the prediction accuracy and the model weight. For example, the model weight corresponding to the prediction accuracy of 90% is 0.9; the model weight corresponding to the prediction accuracy of 60% is 0.6.

[0152] The first initial model weight refers to the model weight corresponding to the first prediction accuracy.

[0153] The second initial model weight refers to the model weight corresponding to the second prediction accuracy.

[0154] Exemplarily, the server acquires a first prediction accuracy of the first risk prediction model and a second prediction accuracy of the second risk prediction model; then, the server queries a corresponding relationship between the prediction accuracy and the model weight, obtains a model weight corresponding to the first prediction accuracy, and takes the model weight as a first initial model weight of the first risk prediction model, and queries the corresponding relationship between the prediction accuracy and the model weight, obtains a model weight corresponding to the second prediction accuracy, and takes the model weight as a second initial model weight of the second risk prediction model; then, the server normalizes the first initial model weight and the second initial model weight to obtain a normalized first initial model weight and a normalized second initial model weight, which are respectively taken as a first model weight of the first risk prediction model and a second model weight of the second risk prediction model.

[0155] In the embodiment, by directly associating the prediction accuracy of the model with the weight, the model with better prediction performance obtains a higher initial weight, ensuring the objectivity and rationality of the weight distribution, retaining the influence of the accuracy difference of different models on the weight, and ensuring the normativeness of the subsequent risk value fusion calculation, which is conducive to improving the reliability of the model output result in the comprehensive risk assessment.

[0156] In one exemplary embodiment, the step S107 generates a risk processing instruction for the building to be analyzed according to the target risk value, specifically including the following contents: determining the current risk level of the building to be analyzed according to the target risk value; obtaining the historical risk level of the building to be analyzed, and fusing the current risk level and the historical risk level to obtain the target risk level of the building to be analyzed; and generating a risk processing instruction for the building to be analyzed according to the target risk level.

[0157] The current risk level refers to the risk level of the building to be analyzed at the current time.

[0158] The historical risk level refers to the risk level of the building to be analyzed in the past period of time.

[0159] The target risk level refers to the risk level obtained by fusing the current risk level and the historical risk level.

[0160] Exemplarily, the server queries a correspondence relationship between a risk value and a risk level, obtains a risk level corresponding to the target risk value as a current risk level of the building to be analyzed; then, the server obtains a historical risk level of the building to be analyzed, and determines a risk level change trend of the building to be analyzed according to the current risk level and the historical risk level; then, the server fuses the current risk level and the risk level change trend to obtain a target risk level of the building to be analyzed; then, the server generates a risk processing instruction matched with the target risk level according to the target risk level, as a risk processing instruction for the building to be analyzed.

[0161] In this embodiment, the target risk level obtained by fusing the current risk level and the historical risk level avoids one-sidedness that may be caused by only relying on a risk level at a single time point, and the risk processing instruction generated based on the target risk level can be more in line with dynamic characteristics and actual situations of the building risk, thereby ensuring that the instruction is not only directed to the current risk but also takes into account the development trend of the risk, and improving the accuracy and foresight of risk processing.

[0162] In an exemplary embodiment, the trained first risk prediction model can be trained in the following manner: sample building structure data, sample equipment data and sample environment data of a sample building are obtained; a sample risk type and a sample risk position of the sample building are determined according to the sample building structure data, the sample equipment data and the sample environment data; a sample risk severity, a sample risk occurrence probability and a sample risk exposure frequency of the sample building are determined according to the sample risk type and the sample risk position; the sample risk severity, the sample risk occurrence probability and the sample risk exposure frequency are input into the first risk prediction model to be trained, to obtain a predicted risk value of the sample building output by the first risk prediction model to be trained; an actual risk value of the sample building is obtained, and the first risk prediction model to be trained is iteratively trained according to a difference between the predicted risk value and the actual risk value of the sample building output by the first risk prediction model to be trained, to obtain the trained first risk prediction model.

[0163] In an exemplary embodiment, the trained second risk prediction model can be trained in the following manner: sample meteorological data of a sample building are obtained; the sample meteorological data are input into the second risk prediction model to be trained, to obtain a predicted risk value of the sample building output by the second risk prediction model to be trained; an actual risk value of the sample building is obtained, and the second risk prediction model to be trained is iteratively trained according to a difference between the predicted risk value and the actual risk value of the sample building output by the second risk prediction model to be trained, to obtain the trained second risk prediction model.

[0164] In an exemplary embodiment, asFigure 3 As shown, another building risk processing method based on building hidden danger governance and meteorological risk prediction and early warning management is provided. Taking the application of the method to a server as an example, the method specifically comprises the following steps:

[0165] In step S301, in response to a risk processing request for a building to be analyzed, building structure data, equipment data and environment data of the building to be analyzed are acquired.

[0166] In step S302, the building structure data, the equipment data and the environment data are input into a trained risk type prediction model to obtain first prediction probabilities of the building to be analyzed under each preset risk type, and the building structure data, the equipment data and the environment data are input into a trained risk position prediction model to obtain second prediction probabilities of the building to be analyzed under each preset risk position.

[0167] In step S303, a preset risk type with the largest first prediction probability is selected from each preset risk type as a current risk type, and a preset risk position with the largest second prediction probability is selected from each preset risk position as a current risk position.

[0168] In step S304, a current risk severity, a current risk occurrence probability and a current risk exposure frequency of the building to be analyzed are determined according to the current risk type and the current risk position.

[0169] In step S305, the current risk severity, the current risk occurrence probability and the current risk exposure frequency are input into a trained first risk prediction model to obtain a first risk value of the building to be analyzed.

[0170] In step S306, current meteorological data of the building to be analyzed are acquired, and the current meteorological data are input into a trained second risk prediction model to obtain a second risk value of the building to be analyzed.

[0171] In step S307, a first prediction accuracy of the first risk prediction model and a second prediction accuracy of the second risk prediction model are acquired.

[0172] In step S308, a corresponding relationship between prediction accuracies and model weights is queried to obtain a model weight corresponding to the first prediction accuracy as a first initial model weight of the first risk prediction model, and the corresponding relationship is queried to obtain a model weight corresponding to the second prediction accuracy as a second initial model weight of the second risk prediction model.

[0173] In step S309, the first initial model weight and the second initial model weight are normalized to obtain a first model weight of the first risk prediction model and a second model weight of the second risk prediction model.

[0174] Step S310, in the case that the first risk value and the second risk value are both less than the preset risk value, performing fusion processing on the first risk value and the second risk value according to the first model weight of the first risk prediction model and the second model weight of the second risk prediction model, to obtain a target risk value of the building to be analyzed.

[0175] Step S311, determining a current risk level of the building to be analyzed according to the target risk value; obtaining a historical risk level of the building to be analyzed, performing fusion processing on the current risk level and the historical risk level to obtain a target risk level of the building to be analyzed; and generating a risk processing instruction for the building to be analyzed according to the target risk level.

[0176] Step S312, performing corresponding risk processing on the building to be analyzed according to the risk processing instruction.

[0177] In the building risk processing method based on building hidden danger management and meteorological risk prediction and early warning management, when the building is processed, the current risk type and the current risk position of the building to be analyzed can be accurately determined by analyzing the building structure data, the equipment data and the environmental data, so that the current risk severity, the current risk occurrence probability and the current risk exposure frequency of the building to be analyzed can be accurately determined, the first risk value of the building to be analyzed can be accurately obtained by using the trained first risk prediction model, and the target risk value of the building to be analyzed can be more accurately determined by combining the first risk value and the second risk value of the building to be analyzed obtained based on the current meteorological data of the building to be analyzed and the trained second risk prediction model, so that the risk processing instruction for the building to be analyzed can be more accurately generated, the building to be analyzed can be more accurately processed, and the risk processing accuracy of the building is improved. Moreover, the whole process does not need manual intervention, which avoids the subjective factors in the decision-making method by using manual generation, and the error caused by the low risk processing accuracy of the building is avoided, and the risk processing accuracy of the building is further improved.

[0178] In one exemplary embodiment, in order to more clearly illustrate the building risk processing method based on building hidden danger management and meteorological risk prediction and early warning management provided by the embodiments of the present application, the building risk processing method based on building hidden danger management and meteorological risk prediction and early warning management is specifically described in one specific embodiment. In one embodiment, the present application also provides a scheme for hidden danger management and meteorological risk prediction and early warning management of a housing building, which specifically includes the following contents:

[0179] In order to build an office comprehensive business building safety risk management framework, establish a safety risk grading control and hidden danger rectification closed-loop dual prevention mechanism, and build a risk plan management, safety risk assessment foundation, benchmark risk management, dynamic risk management, risk control, emergency response disposal, risk visualization control, risk control analysis and other programs.

[0180] 1. Risk plan management: Develop annual risk control plans by level, issue annual plan formal documents, develop annual safety risk control work plans, and propose overall risk control requirements. The system supports uploading the published annual risk plan documents into the system.

[0181] 2. Safety risk assessment foundation: Includes risk cause maintenance, accident event type maintenance, hazard category maintenance, and other functions to facilitate the development of risk assessment rules and subsequent hazard category statistics.

[0182] 3. Benchmark risk management: Develop assessment work based on annual assessment tasks, use assessment method standards for risk assessment, identify hazard categories, and combine the consequences of accidents caused by various hazards, the frequency of exposure to hazard factors, and the possibility of consequences to obtain relevant risk values.

[0183] 4. Dynamic risk management: Based on problem risk assessment, the units of each building and building conduct supplementary risk assessment on the benchmark risk list in response to risk-prone operations, accident events, typhoons, heavy rain and other severe weather. Regularly invite professional agencies to conduct safety assessment on buildings, especially old buildings, and continuously conduct dynamic risk assessment to dynamically adjust the risk level of buildings. Conduct scene-based continuous operation risk assessment based on specific high-risk operations, accident events, and changes in severe weather during the execution of the plan to determine the final risk level and control level, and dynamically revise the benchmark risk database.

[0184] 5. Risk control: Control risks that have entered the benchmark risk library and dynamic risk library, arrange for the implementer to execute risk response measures. Conduct risk evaluation, risk re-evaluation, and determine whether the risk has been eliminated, downgraded, or upgraded. Strengthen the rectification measures for the upgraded risks.

[0185] 6. Emergency response disposal: Real-time monitoring of natural disasters such as typhoons, thunderstorms, and earthquakes, scientific research and judgment, and establishment of emergency response disposal mechanism. Achieve pre-disaster damage warning and post-disaster damage analysis. When an emergency event occurs, start the warning based on the data monitored by meteorology, earthquake, etc., judge the risk level, issue a warning notice, start emergency response, execute measures, and report response-related information. When the warning is eliminated, release the response according to the conditions. Achieve the functions of warning release, response release, warning elimination, and response elimination.

[0186] 7. Risk visualization management and control: risk visualization management and control, build a risk visualization management and control model, graphically display the risk distribution of buildings of each unit, the system provides risk distribution based on map display function, realizes functions such as layer switching, data filtering, query positioning, linkage analysis, etc.

[0187] 8. Risk management and control analysis: build statistical models of risk level, risk classification, accident event situation, major risk, etc., and graphically display relevant indicators to facilitate risk management personnel to track various risks and form a risk management closed loop.

[0188] The above embodiments, when processing the risk of the building, by analyzing the building structure data, equipment data and environment data of the building, the current risk type and the current risk position of the building to be analyzed can be accurately determined, so that the current risk severity, the current risk occurrence probability and the current risk exposure frequency of the building to be analyzed can be accurately determined, the first risk value of the building to be analyzed can be accurately obtained through the first risk prediction model trained, and the target risk value of the building to be analyzed can be more accurately determined by combining the second risk value of the building to be analyzed obtained based on the current meteorological data of the building to be analyzed and the second risk prediction model trained, and then the risk processing instruction for the building to be analyzed can be more accurately generated, which is beneficial to more accurate risk processing of the building to be analyzed, thereby improving the risk processing accuracy of the building. Moreover, the whole process does not need manual intervention, avoiding the subjective factors in the way of generating decisions by artificial, which is easy to make mistakes, resulting in the defect that the risk processing accuracy of the building is low, further improving the risk processing accuracy of the building. At the same time, the time for statistical analysis of the related business data of the housing building safety hidden danger is saved, the difficulty of understanding the unit housing building safety risk management situation is reduced, the weather condition is dynamically and real-timely monitored, the scientific research and judgment are made, the disaster prediction and early warning, the disaster damage analysis are achieved, the safety risk grading management and control and hidden danger rectification closed loop double prevention mechanism are further realized, the lean level of hidden danger safety management is improved, and the safety management level and ability are ensured from the overall point of view.

[0189] It should be understood that although the steps in the flowcharts involved in the above embodiments are shown in sequence according to the arrows, the steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the execution of the steps is not strictly limited in sequence, and the steps can be executed in other sequences. Moreover, at least some of the steps in the flowcharts involved in the above embodiments can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of the steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least part of other steps or steps or stages in other steps.

[0190] Based on the same inventive concept, the embodiments of the present application also provide a building risk processing device for implementing the building risk processing method based on building hidden danger management and meteorological risk prediction and early warning management. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more building risk processing device embodiments based on building hidden danger management and meteorological risk prediction and early warning management provided below can be referred to the limitations of the building risk processing method based on building hidden danger management and meteorological risk prediction and early warning management described above, which will not be repeated here.

[0191] In one exemplary embodiment, as shown in Figure 4 A building risk processing device based on building hidden danger management and meteorological risk prediction and early warning management is provided, comprising: a data acquisition module 401, a first determination module 402, a second determination module 403, a first prediction module 404, a second prediction module 405, a risk fusion module 406 and a risk processing module 407, wherein:

[0192] The data acquisition module 401 is configured to acquire building structure data, equipment data and environmental data of a building to be analyzed in response to a risk processing request for the building to be analyzed.

[0193] The first determination module 402 is configured to determine a current risk type and a current risk location of the building to be analyzed according to the building structure data, the equipment data and the environmental data.

[0194] The second determination module 403 is configured to determine a current risk severity, a current risk occurrence probability and a current risk exposure frequency of the building to be analyzed according to the current risk type and the current risk location.

[0195] The first prediction module 404 is configured to input the current risk severity, the current risk occurrence probability and the current risk exposure frequency into the trained first risk prediction model to obtain a first risk value of the building to be analyzed.

[0196] The second prediction module 405 is configured to obtain current meteorological data of the building to be analyzed, and input the current meteorological data into the trained second risk prediction model to obtain a second risk value of the building to be analyzed.

[0197] The risk fusion module 406 is configured to, in a case where both the first risk value and the second risk value are less than a preset risk value, perform fusion processing on the first risk value and the second risk value according to a first model weight of the first risk prediction model and a second model weight of the second risk prediction model to obtain a target risk value of the building to be analyzed.

[0198] The risk processing module 407 is configured to generate a risk processing instruction for the building to be analyzed according to the target risk value, and perform corresponding risk processing on the building to be analyzed according to the risk processing instruction.

[0199] In an exemplary embodiment, the first determination module 402 is further configured to input the building structure data, the equipment data and the environmental data into the trained risk type prediction model to obtain a first prediction probability of the building to be analyzed under each preset risk type, and input the building structure data, the equipment data and the environmental data into the trained risk position prediction model to obtain a second prediction probability of the building to be analyzed under each preset risk position; and select, from each preset risk type, a preset risk type with the largest first prediction probability as a current risk type, and select, from each preset risk position, a preset risk position with the largest second prediction probability as a current risk position.

[0200] In an example embodiment, the first prediction module 404 is further configured to perform feature extraction on the current risk severity, the current risk occurrence probability and the current risk exposure frequency respectively to obtain a first feature vector of the current risk severity, a second feature vector of the current risk occurrence probability and a third feature vector of the current risk exposure frequency; input the first feature vector into the first risk prediction network to obtain a third risk value of the building to be analyzed, input the second feature vector into the second risk prediction network to obtain a fourth risk value of the building to be analyzed, and input the third feature vector into the third risk prediction network to obtain a fifth risk value of the building to be analyzed; input the third risk value, the fourth risk value and the fifth risk value into the attention mechanism network respectively to obtain a first weight corresponding to the third risk value, a second weight corresponding to the fourth risk value and a third weight corresponding to the fifth risk value; and perform summation processing on the third risk value, the fourth risk value and the fifth risk value according to the first weight, the second weight and the third weight to obtain a first risk value of the building to be analyzed.

[0201] In an example embodiment, the second prediction module 405 is further configured to perform combination processing on the current weather data to obtain combined data corresponding to the current weather data; perform feature extraction on the current weather data and the combined data respectively to obtain a fourth feature vector of the current weather data and a fifth feature vector of the combined data; perform splicing processing on the fourth feature vector and the fifth feature vector to obtain a spliced feature vector; and input the spliced feature vector into the trained second risk prediction model to obtain a second risk value of the building to be analyzed.

[0202] In an example embodiment, the building risk processing apparatus based on building hidden danger management and meteorological risk prediction and early warning management further comprises a model weight determination module configured to obtain a first prediction accuracy of the first risk prediction model and a second prediction accuracy of the second risk prediction model; query a corresponding relationship between prediction accuracy and model weight to obtain a model weight corresponding to the first prediction accuracy as a first initial model weight of the first risk prediction model, and query the corresponding relationship to obtain a model weight corresponding to the second prediction accuracy as a second initial model weight of the second risk prediction model; and perform normalization processing on the first initial model weight and the second initial model weight to obtain a first model weight of the first risk prediction model and a second model weight of the second risk prediction model.

[0203] In an example embodiment, the risk processing module 407 is further configured to determine a current risk level of the building to be analyzed according to the target risk value; obtain a historical risk level of the building to be analyzed, perform fusion processing on the current risk level and the historical risk level to obtain a target risk level of the building to be analyzed; and generate a risk processing instruction for the building to be analyzed according to the target risk level.

[0204] The various modules in the building risk processing apparatus based on building hidden danger management and meteorological risk prediction and early warning management can be implemented by software, hardware, and combinations thereof, in whole or in part. The various modules can be embedded in or independent of a processor in a computer device in hardware form, or can be stored in a memory in a computer device in software form, so as to be called and executed by a processor to perform the operations corresponding to the various modules.

[0205] In an exemplary embodiment, a computer device, which can be a server, is provided, and an internal structure diagram of the computer device can be as shown in Figure 5 The computer device includes a processor, a memory, an input / output interface (I / O), and a communication interface. The processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The database of the computer device is configured to store building structure data, device data, and environment data. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to communicate with terminals outside through a network connection. The computer program is executed by the processor to implement a building risk processing method based on building hidden danger management and meteorological risk prediction and early warning management.

[0206] Those skilled in the art can understand that Figure 5 The structure shown in the above

[0207] In an exemplary embodiment, a computer device is also provided, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0208] In an exemplary embodiment, a computer readable storage medium is provided, which stores a computer program. The computer program is executed by a processor to implement the steps in the above method embodiments.

[0209] In an example embodiment, a computer program product is provided, including a computer program which, when executed by a processor, implements the steps of any of the above method embodiments.

[0210] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0211] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present disclosure.

[0212] The above-described embodiments are merely illustrative of several embodiments of the present application, which are described in more detail and in a specific manner, but should not be construed as limiting the scope of the patent of the present application. It should be noted that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A building risk management method based on building hazard mitigation and meteorological risk prediction and early warning management, characterized in that, The method includes: In response to a risk management request for the building to be analyzed, the building structure data, equipment data, and environmental data of the building to be analyzed are obtained; Based on the building structure data, the equipment data, and the environmental data, the current risk type and current risk location of the building to be analyzed are determined. Based on the current risk type and the current risk location, the current risk severity, current risk occurrence probability, and current risk exposure frequency of the building to be analyzed are determined. The current risk severity, the current risk occurrence probability, and the current risk exposure frequency are input into the trained first risk prediction model to obtain the first risk value of the building to be analyzed. Obtain the current meteorological data of the building to be analyzed, input the current meteorological data into the trained second risk prediction model, and obtain the second risk value of the building to be analyzed; When both the first risk value and the second risk value are less than the preset risk value, the first risk value and the second risk value are fused according to the first model weight of the first risk prediction model and the second model weight of the second risk prediction model to obtain the target risk value of the building to be analyzed. Based on the target risk value, a risk handling instruction is generated for the building to be analyzed, and the corresponding risk handling is performed on the building to be analyzed in accordance with the risk handling instruction.

2. The method according to claim 1, characterized in that, The step of determining the current risk type and current risk location of the building to be analyzed based on the building structure data, the equipment data, and the environmental data includes: The building structure data, equipment data, and environmental data are input into the trained risk type prediction model to obtain the first predicted probability of the building to be analyzed under each preset risk type; and the building structure data, equipment data, and environmental data are input into the trained risk location prediction model to obtain the second predicted probability of the building to be analyzed under each preset risk location. From the preset risk types, the preset risk type with the highest first predicted probability is selected as the current risk type, and from the preset risk locations, the preset risk location with the highest second predicted probability is selected as the current risk location.

3. The method according to claim 1, characterized in that, The first risk prediction model includes a first risk prediction network corresponding to the current risk severity, a second risk prediction network corresponding to the current risk occurrence probability, a third risk prediction network corresponding to the current risk exposure frequency, and an attention mechanism network; The step of inputting the current risk severity, the current risk occurrence probability, and the current risk exposure frequency into the trained first risk prediction model to obtain the first risk value of the building to be analyzed includes: Feature extraction is performed on the current risk severity, the current risk occurrence probability, and the current risk exposure frequency respectively to obtain a first feature vector of the current risk severity, a second feature vector of the current risk occurrence probability, and a third feature vector of the current risk exposure frequency; The first feature vector is input into the first risk prediction network to obtain the third risk value of the building to be analyzed; the second feature vector is input into the second risk prediction network to obtain the fourth risk value of the building to be analyzed; and the third feature vector is input into the third risk prediction network to obtain the fifth risk value of the building to be analyzed. The third risk value, the fourth risk value, and the fifth risk value are respectively input into the attention mechanism network to obtain the first weight corresponding to the third risk value, the second weight corresponding to the fourth risk value, and the third weight corresponding to the fifth risk value; The third risk value, the fourth risk value, and the fifth risk value are summed according to the first weight, the second weight, and the third weight to obtain the first risk value of the building to be analyzed.

4. The method according to claim 1, characterized in that, The step of inputting the current meteorological data into the trained second risk prediction model to obtain the second risk value of the building to be analyzed includes: The current meteorological data is combined and processed to obtain the combined data corresponding to the current meteorological data; Feature extraction processing is performed on the current meteorological data and the combined data respectively to obtain the fourth feature vector of the current meteorological data and the fifth feature vector of the combined data; The fourth feature vector and the fifth feature vector are concatenated to obtain a concatenated feature vector; The spliced ​​feature vector is input into the trained second risk prediction model to obtain the second risk value of the building to be analyzed.

5. The method according to claim 1, characterized in that, Before fusing the first risk value and the second risk value according to the first model weight of the first risk prediction model and the second model weight of the second risk prediction model to obtain the target risk value of the building to be analyzed, the method further includes: Obtain the first prediction accuracy of the first risk prediction model and the second prediction accuracy of the second risk prediction model; The model weights corresponding to the prediction accuracy are obtained by querying the relationship between prediction accuracy and model weights. The model weights corresponding to the first prediction accuracy are used as the first initial model weights of the first risk prediction model. The model weights corresponding to the second prediction accuracy are obtained by querying the relationship. The model weights corresponding to the second prediction accuracy are used as the second initial model weights of the second risk prediction model. The first initial model weights and the second initial model weights are normalized to obtain the first model weights of the first risk prediction model and the second model weights of the second risk prediction model.

6. The method according to any one of claims 1 to 5, characterized in that, The step of generating risk handling instructions for the building to be analyzed based on the target risk value includes: Based on the target risk value, the current risk level of the building to be analyzed is determined; Obtain the historical risk level of the building to be analyzed, and fuse the current risk level and the historical risk level to obtain the target risk level of the building to be analyzed; Based on the target risk level, generate risk handling instructions for the building to be analyzed.

7. A building risk management device based on building hazard mitigation and meteorological risk prediction and early warning management, characterized in that, The device includes: The data acquisition module is used to acquire the building structure data, equipment data and environmental data of the building to be analyzed in response to a risk handling request for the building to be analyzed. The first determining module is used to determine the current risk type and current risk location of the building to be analyzed based on the building structure data, the equipment data and the environmental data; The second determining module is used to determine the current risk severity, current risk occurrence probability, and current risk exposure frequency of the building to be analyzed based on the current risk type and the current risk location. The first prediction module is used to input the current risk severity, the current risk occurrence probability, and the current risk exposure frequency into the trained first risk prediction model to obtain the first risk value of the building to be analyzed. The second prediction module is used to acquire the current meteorological data of the building to be analyzed, input the current meteorological data into the trained second risk prediction model, and obtain the second risk value of the building to be analyzed. The risk fusion module is used to fuse the first risk value and the second risk value according to the first model weight of the first risk prediction model and the second model weight of the second risk prediction model when both the first risk value and the second risk value are less than a preset risk value, so as to obtain the target risk value of the building to be analyzed. The risk processing module is used to generate risk processing instructions for the building to be analyzed based on the target risk value, and to perform corresponding risk processing on the building to be analyzed according to the risk processing instructions.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images