Building safety evaluation method and device based on vector space model, and storage medium

Through the building safety evaluation method of vector space model, quantitative encoding and similarity calculation are combined with multi-dimensional data, the non-systematic and unscientific problems of steel structure building safety assessment in the existing technology are solved, and a more comprehensive safety assessment is achieved and credibility of the assessment results.

CN119069043AInactive Publication Date: 2024-12-03HUNAN VOCATIONAL INST OF SAFETY TECH

Patent Information

Application Number
CN202411042877.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-12-03
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing steel structure building safety assessment methods lack systematicity and scientificity, the data acquisition is not comprehensive, the evaluation methods are single, it is difficult to dynamically grasp the safety status of the building, and the evaluation results are inconsistent and non-repeatable.

Method used

Using a building safety evaluation method based on vector space model, quantitative encoding and similarity calculation are performed by acquiring and classifying multi-dimensional data (geometric structure, material properties, environmental conditions, usage history and real-time monitoring data), and using Euclidean distance Weighted assessment with cosine similarity to provide a more comprehensive safety assessment.

Benefits of technology

It realizes the comprehensiveness and accuracy of building safety assessment, enhances the credibility and flexibility of assessment, and can adjust it according to specific building characteristics and needs, providing a more comprehensive assessment basis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119069043A_ABST
    Figure CN119069043A_ABST
Patent Text Reader

Abstract

The invention discloses a building safety evaluation method and device based on a vector space model and a storage medium. The method comprises the following steps: acquiring original building data, classifying the original building data, and determining data needing to be quantized and coded in the original building data; according to a pre-configured coding rule, performing quantization coding on the data needing quantization coding; obtaining comparison building data; constructing a multi-dimensional vector data sequence based on the original building data and the comparison building data to obtain a corresponding original data sequence and a comparison vector data sequence; respectively calculating Euclidean distance and cosine similarity of the original data sequence and the contrast vector data sequence to obtain an Euclidean distance sequence and a cosine similarity sequence; respectively carrying out weighted calculation on the Euclidean distance sequence and the cosine similarity sequence to obtain a first similarity value and a second similarity value; and evaluating the building based on the first similarity and the second similarity.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of big data processing technology, and in particular to a building safety evaluation method, device and storage medium based on a vector space model. Background Art

[0002] Steel structure buildings are widely used in high-rise buildings, industrial plants, bridges and other fields due to their high strength, fast construction speed and good earthquake resistance. However, with the increase of service life and changes in environmental conditions, the safety performance of steel structure buildings has gradually become the focus of attention. The existing safety assessment methods for steel structure buildings mainly rely on manual inspection and experience judgment, which lacks systematicity and scientificity. These traditional methods have the following shortcomings:

[0003] Incomplete data acquisition: Existing methods mainly focus on geometric structure and material properties, ignoring the impact of environmental conditions and usage history on building safety.

[0004] The assessment method is single: it mainly relies on on-site inspection, and there is a lack of real-time monitoring methods, making it difficult to dynamically grasp the safety status of the building.

[0005] Inconsistent evaluation standards: The subjective judgments of different evaluators lead to inconsistency and non-repeatability of evaluation results.

[0006] With the development of Internet of Things (IoT), big data and artificial intelligence (AI) technology, intelligent safety assessment methods based on multi-dimensional data have gradually become possible. Using sensors and IoT technology, a variety of building data can be collected in real time, including geometric and structural data, material performance data, environmental condition data, usage history data and real-time monitoring data.

[0007] Deficiencies of existing technology:

[0008] Existing technologies have deficiencies in multi-dimensional data fusion and comprehensive analysis, and it is difficult to fully utilize the advantages of various types of data. Existing similarity assessment methods mainly rely on single-dimensional data, which is difficult to fully reflect the comprehensive status of buildings. Summary of the invention

[0009] In order to solve the above technical problems, the present application provides a building safety evaluation method, device and storage medium based on a vector space model. The technical solution in the present application is described below:

[0010] In a first aspect, the present application provides a building safety evaluation method based on a vector space model, the method comprising:

[0011] Acquire original building data, and classify the original building data, wherein the classification includes: geometric structure data, material performance data, environmental condition data, usage history data, and real-time monitoring data;

[0012] The geometric structure data includes the building age, height-to-width ratio, number of above-ground floors, structural system, structural length-to-width ratio, structural width and structural span;

[0013] The material performance data include material density, material elastic modulus, material chemical resistance, aging degree and structural corrosion ratio;

[0014] Data on environmental conditions include annual precipitation, mean annual temperature, and earthquake frequency;

[0015] The usage history data includes usage load;

[0016] The real-time monitoring data includes deformation vector, column stress, beam stress, vibration frequency and vibration amplitude;

[0017] Determining data that needs to be quantized and encoded in the original building data;

[0018] According to a preconfigured encoding rule, performing quantization encoding on the data that needs quantization encoding;

[0019] Acquire comparison building data, wherein the comparison building data has a consistent data structure with the original building data;

[0020] Constructing a multidimensional vector data sequence based on the original building data and the comparison building data respectively to obtain a corresponding original data sequence and a comparison vector data sequence;

[0021] Calculating the Euclidean distance and cosine similarity of the original data sequence and the comparison vector data sequence respectively to obtain a Euclidean distance sequence and a cosine similarity sequence;

[0022] Performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain a first similarity value and a second similarity value;

[0023] The building is evaluated based on the first similarity and the second similarity.

[0024] Optionally, performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain a first similarity value and a second similarity value comprises:

[0025] Performing Z-score standardization conversion on the first similarity and the second similarity respectively to obtain a first conversion value and a second conversion value;

[0026] Based on a preset weight, performing a weighted calculation on the first conversion value and the second conversion value to obtain a comprehensive similarity;

[0027] The buildings are evaluated according to the comprehensive similarity.

[0028] Optional data that needs to be quantized and encoded include:

[0029] Building age, structural system, chemical resistance of materials and use loads;

[0030] The quantization coding includes:

[0031] Determining the phase code of the building age from a preconfigured table, wherein the preconfigured table divides the year into a plurality of phases of 10 years each, and configures a phase code for each phase;

[0032] Determine the structure code of the structure system from a preconfigured table, the structure system including frame structure, frame central support structure, frame eccentric support structure, simplified structure, giant frame structure, single plant structure and bent structure, wherein each structure is configured with a corresponding structure code;

[0033] Determine the chemical resistance code of the structural system from a preconfigured table, wherein the chemical resistance includes high corrosion resistance, medium corrosion resistance and low corrosion resistance, high fire resistance, medium fire resistance and low fire resistance; wherein each material chemical resistance is configured with a corresponding chemical resistance code and is represented by a binary array;

[0034] The load code of the usage load is determined from a preconfigured table, wherein the usage load includes a high load, a medium load and a low load, wherein each usage load is configured with a corresponding load code.

[0035] Optionally, the cosine similarity is calculated by the following formula:

[0036] D_geometry=∑(i=1to n)w_i*(X_i-Y_i)^2

[0037] The Euclidean distance is calculated by the following formula:

[0038] EuclideanDistance=sqrt(∑(i=1to n)(X_i-Y_i)^2)

[0039] Among them, X_i represents the i-th dimension data in the original building data, Y_i represents the i-th dimension data in the compared building data, D_geometry represents cosine similarity, EuclideanDistance represents Euclidean distance, and sqrt represents square root operation.

[0040] Optionally, performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain a first similarity value and a second similarity value comprises:

[0041] The calculation is done as follows:

[0042] The first similarity value calculation:

[0043] WES=∑(i=1to n)w_i*Euclidean_Distance_i;

[0044] Among them, Euclidean_Distance_i represents the similarity of the i-th dimension data, w_i is the weight of the i-th dimension data, and WES represents the first similarity value;

[0045] Calculation of the second similarity value:

[0046] WCS=∑(i=1ton)w_i*D_geometry_i

[0047] Wherein, D_geometry_i represents the similarity of the i-th dimension data, w_i is the weight of the i-th dimension data, and WCS represents the second similarity value.

[0048] Optionally, based on a preset weight, performing weighted calculation on the first conversion value and the second conversion value to obtain a comprehensive similarity includes:

[0049] The comprehensive similarity is calculated by the following formula:

[0050] S_composite=w_e*WES+w_c*WCS;

[0051] Where S_composite represents the composite similarity, w_e represents the weight of the first similarity value, and w_c represents the weight of the second similarity value.

[0052] A second aspect of the present application provides a building safety evaluation device based on a vector space model, the device comprising:

[0053] A first data acquisition unit, used to acquire original building data and classify the original building data, wherein the classification includes: geometric structure data, material performance data, environmental condition data, usage history data, and real-time monitoring data;

[0054] The geometric structure data includes the building age, height-to-width ratio, number of above-ground floors, structural system, structural length-to-width ratio, structural width and structural span;

[0055] The material performance data include material density, material elastic modulus, material chemical resistance, aging degree and structural corrosion ratio;

[0056] Data on environmental conditions include annual precipitation, mean annual temperature, and earthquake frequency;

[0057] The usage history data includes usage load;

[0058] The real-time monitoring data includes deformation vector, column stress, beam stress, vibration frequency and vibration amplitude;

[0059] A data determination unit, used for determining data that needs to be quantized and encoded in the original building data;

[0060] A data encoding unit, configured to perform quantization encoding on the data that needs quantization encoding according to a preconfigured encoding rule;

[0061] A second data acquisition unit is used to acquire comparative building data, wherein the comparative building data has a consistent data structure with the original building data;

[0062] A sequence construction unit, used to construct a multi-dimensional vector data sequence based on the original building data and the comparison building data, respectively, to obtain a corresponding original data sequence and a comparison vector data sequence;

[0063] A similarity calculation unit, used to calculate the Euclidean distance and cosine similarity of the original data sequence and the comparison vector data sequence respectively, to obtain a Euclidean distance sequence and a cosine similarity sequence;

[0064] A weighted calculation unit, used for performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain a first similarity value and a second similarity value;

[0065] An evaluation unit is configured to evaluate the building based on the first similarity and the second similarity.

[0066] Optionally, the weighted calculation unit is specifically used for:

[0067] Performing Z-score standardization conversion on the first similarity and the second similarity respectively to obtain a first conversion value and a second conversion value;

[0068] Based on a preset weight, performing a weighted calculation on the first conversion value and the second conversion value to obtain a comprehensive similarity;

[0069] The buildings are evaluated according to the comprehensive similarity.

[0070] In a third aspect, the present application provides a building safety evaluation device based on a vector space model, the device comprising:

[0071] Processor, memory, input-output unit, and bus;

[0072] The processor is connected to the memory, the input and output unit, and the bus;

[0073] The memory stores a program, and the processor calls the program to execute the first aspect and any optional method in the first aspect.

[0074] A fourth aspect of the present application provides a computer-readable storage medium, on which a program is stored. When the program is executed on a computer, the program executes the first aspect and any optional method in the first aspect.

[0075] It can be seen from the above technical solutions that this application has the following advantages:

[0076] 1. This method comprehensively considers all aspects of building safety assessment through multi-dimensional building data classification and quantitative coding, including geometric structure, material properties, environmental conditions, usage history and real-time monitoring data, making the assessment more comprehensive and integrated.

[0077] 2. Through pre-configured coding rules and quantitative coding process, the raw data is converted into computable numerical form, making data analysis and comparison more accurate and reliable.

[0078] 3. Representing building data as a multi-dimensional vector sequence helps capture and compare the complex features and relationships between different buildings, improving data expression and analysis capabilities.

[0079] 4. Calculating Euclidean distance and cosine similarity simultaneously can evaluate the similarities and differences between buildings from different perspectives, providing a more comprehensive basis for evaluation.

[0080] 5. By performing weighted calculation on the Euclidean distance and cosine similarity sequences, the first and second similarity values ​​are obtained, which effectively integrates the information of multiple evaluation dimensions and enhances the accuracy and credibility of the evaluation.

[0081] 6. The classification and quantitative coding process of this method can be adjusted and customized according to the characteristics and assessment requirements of specific buildings, and has strong flexibility and applicability. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] In order to more clearly illustrate the technical solution in the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0083] Figure 1A schematic diagram of an embodiment of a building safety assessment method based on a vector space model provided in this application;

[0084] Figure 2 A schematic diagram of an embodiment of a flow chart for calculating comprehensive similarity in a building safety evaluation method based on a vector space model provided in this application;

[0085] Figure 3 A schematic diagram of the structure of an embodiment of a building safety assessment device based on a vector space model provided in this application;

[0086] Figure 4 This is a schematic structural diagram of an embodiment of another building safety assessment device based on a vector space model provided in this application. DETAILED DESCRIPTION

[0087] It should be noted that the method provided in this application can be applied to a terminal, a system, or a server. For example, the terminal can be a smart phone or a computer, a tablet computer, a smart TV, a smart watch, a portable computer terminal, or a fixed terminal such as a desktop computer. For the convenience of explanation, this application uses the terminal as an example for explanation.

[0088] See also Figure 1 , Figure 1 A schematic flow chart of an embodiment of a building safety evaluation method based on a vector space model provided in this application, the building safety evaluation method based on a vector space model includes:

[0089] S101, obtaining original building data, and classifying the original building data;

[0090] The categories include: geometric structure data, material performance data, environmental condition data, usage history data, and real-time monitoring data;

[0091] In this step, the original building data is first obtained, including:

[0092] Geometric data: including building age (year), aspect ratio (ratio of building height to minimum width), number of above-ground floors, structural system (such as frame structure), structural aspect ratio (ratio of building projected length to width), structural width and structural span (the distance between the main vertical structures of the structure).

[0093] Material performance data: including material density, material elastic modulus, material chemical resistance, aging degree and structural corrosion ratio (the proportion of moderately corroded components in the total tested components).

[0094] Environmental conditions data: including annual precipitation, annual average temperature and earthquake frequency.

[0095] Usage history data: including usage load conditions.

[0096] Real-time monitoring data: including real-time data such as deformation vector, column stress, beam stress, vibration frequency and vibration amplitude.

[0097] The collected data is sorted and organized to ensure that each data item corresponds to the geometric structure, material properties, environmental conditions, usage history and real-time monitoring data. Database management systems (such as MySQL, PostgreSQL) or data warehouses (such as Hadoop, AWS S3) can be used to store and manage large amounts of building data.

[0098] The data after sorting and expressing using mathematical symbols are as follows:

[0099] Geometry data:

[0100] Construction Year: Y_{\text{Year}}

[0101] Aspect ratio: Y_{\text{Aspect ratio}}

[0102] Number of above-ground floors: Y_{\text{Number of above-ground floors}}

[0103] Structural system: Y_{\text{Structural system}}

[0104] Structural aspect ratio: Y_{\text{Aspect ratio}}

[0105] Structure width: Y_{\text{Structure width}}

[0106] Structural span: Y_{\text{Structural span}}

[0107] Material performance data:

[0108] Material density: M_{\text{density}}

[0109] Material elastic modulus: M_{\text{Elastic modulus}}Material chemical resistance: M_{\text{Chemical resistance}}Aging degree: M_{\text{Aging degree}}

[0110] Constructing the corrosion ratio: M_{\text{Corrosion ratio}}Environmental condition data:

[0111] Annual precipitation: E_{\text{precipitation}}

[0112] Annual average temperature: E_{\text{average temperature}}Earthquake frequency: E_{\text{Earthquake frequency}}

[0113] Using historical data:

[0114] Usage load condition: H_{\text{Load condition}} Real-time monitoring data:

[0115] Deformation vector: R_{\text{deformation}}

[0116] Column stress: R_{\text{Column stress}}

[0117] Beam stress: R_{\text{Beam stress}}

[0118] Vibration frequency: R_{\text{Vibration frequency}}

[0119] Vibration amplitude: R_{\text{Vibration amplitude}}

[0120] Here is an example of raw building data:

[0121] data = {

[0122] 'Building Year':[1980,1990,2000],

[0123] 'Aspect Ratio':[1.5,2.0,2.5],

[0124] 'Number of floors above ground': [10,20,30],

[0125] 'Structural system': ['frame structure', 'simplified structure', 'huge frame structure'],

[0126] 'Structure aspect ratio': [1.2, 1.3, 1.4],

[0127] 'Structure width': [15, 20, 25],

[0128] 'Structure span':[6,7,8],

[0129] 'Material density':[2400,2500,2600],

[0130] 'Material elastic modulus':[200,210,220],

[0131] 'Corrosion resistance':['High corrosion resistance', 'Medium corrosion resistance', 'Low corrosion resistance'],

[0132] 'Refractory':['High Refractory', 'Medium Refractory', 'Low Refractory'],

[0133] 'Aging degree': ['New', 'Slightly aged', 'Moderately aged'],

[0134] 'Component corrosion ratio':[0.01,0.02,0.03],

[0135] 'Annual precipitation':[1000,1100,1200],

[0136] 'Annual average temperature':[15,16,17],

[0137] 'Earthquake frequency':[0.1,0.2,0.3],

[0138] 'Use load': ['light load', 'medium load', 'high load'],

[0139] 'Deformation vector':[0.05,0.06,0.07],

[0140] 'Column stress':[100,110,120],

[0141] 'Beam stress':[200,210,220],

[0142] 'Vibration frequency':[0.5,0.6,0.7],

[0143] 'Vibration amplitude':[0.01,0.02,0.03]

[0144] }

[0145] S102, determining data that needs to be quantized and encoded in the original building data;

[0146] In this step, it is necessary to determine which data needs to be quantified and encoded based on the specific evaluation objectives and predefined encoding rules. The purpose of quantification encoding is to convert non-numeric or continuous data into a numerical form that can be processed by a computer for subsequent numerical calculation and comparison.

[0147] According to the evaluation objectives, identify and determine the data items that need to be quantified and encoded. These data items may include non-numerical data (such as structural systems), continuous data (such as annual precipitation), categorical data (such as building use), etc. For continuous data, methods such as range scaling (min-max scaling) and normalization (z-score normalization) can be used to map the data to a specific range or standardized distribution. According to the selected quantization method, perform the corresponding encoding processing on each data item. Ensure that the encoded data retains the meaning and important information of the original data, and is suitable for subsequent numerical calculations and comparison operations. For categorical data, encoding conversion can be performed, such as using one-hot encoding or integer encoding.

[0148] S103, performing quantization coding on the data requiring quantization coding according to a preconfigured coding rule;

[0149] According to the preconfigured encoding rules, the selected data is converted into numerical form for subsequent calculation and analysis.

[0150] In this step, you first need to clarify the encoding rules, which include:

[0151] Encoding rules for categorical data: for example, encoding “architecture” as an integer or a one-hot encoding.

[0152] Coding rules for continuous data: for example, scaling or standardizing “building year”.

[0153] Before encoding, make sure your data is clean and consistent. Do the following:

[0154] Data cleaning: remove missing values, outliers and duplicate values.

[0155] Data formatting: Ensure data type and format are consistent.

[0156] For categorical data, such as "architecture", one-hot encoding or integer encoding can be used:

[0157] One-hot encoding:

[0158] Convert each classification label into a binary vector. For example, "frame structure" is encoded as [1,0,0,0,0], "frame center support structure" is encoded as [0,1,0,0,0], and so on.

[0159] Here is a specific example of performing quantization encoding:

[0160] Based on the original building data in step S101, the quantization encoding is performed as follows:

[0161] Structural system code

[0162] structure_system_mapping = {

[0163] 'Frame structure': 1,

[0164]

[0165] Use payload encoding

[0166] usage_load_mapping = {

[0167] 'Low load':1,

[0168] 'Medium load':2,

[0169] 'High load':3

[0170] }

[0171] df['usage load encoding'] = df['usage load'].map(usage_load_mapping)

[0172] Display the encoded data

[0173] S104, obtaining comparison building data, wherein the comparison building data has a data structure consistent with the original building data;

[0174] Obtain comparative building data that is consistent with the original building data structure for comparison and evaluation. Here is a specific example:

[0175] Example raw building data

[0176] original_data=pd.DataFrame({

[0177] 'Building Year':[1990,2000,2010],

[0178] 'Aspect Ratio':[1.5,2.0,2.5],

[0179] 'Number of floors above ground': [10,15,20],

[0180] 'Structural system':['frame structure', 'frame central support structure', 'simplified structure'],

[0181] 'Annual precipitation':[800,1200,1000],

[0182] 'Annual average temperature':[15,18,20],

[0183] 'Earthquake frequency':[2,1,3]

[0184] })

[0185] Example Comparison Building Data

[0186] comparison_data=pd.DataFrame({

[0187] 'Building Year':[1995,2005,2015],

[0188] 'Aspect Ratio':[1.7,2.1,2.3],

[0189] 'Number of floors above ground':[12,14,22],

[0190] 'Structural system':['frame structure','frame eccentric support structure','simplified structure'],

[0191] 'Annual precipitation':[850,1100,950],

[0192] 'Annual average temperature':[16,19,21],

[0193] 'Earthquake frequency':[2,1,2]

[0194] })

[0195] S105, constructing a multidimensional vector data sequence based on the original building data and the comparison building data respectively, to obtain a corresponding original data sequence and a comparison vector data sequence;

[0196] In this step, the original building data and the comparison building data are used to construct multi-dimensional vector data sequences, which will be used for subsequent similarity calculations.

[0197] The standardized and coded original building data and comparison building data can be used. OneHot encoding is performed on the "structural system" to ensure consistent encoding. The continuous data is standardized so that it can be compared on the same scale. The coded categorical data and the standardized continuous data are merged to obtain the final original building data and comparison building data. The processed data is converted into a multi-dimensional vector, which represents the data sequence of the original building and the comparison building respectively.

[0198] S106, respectively calculating the Euclidean distance and cosine similarity of the original data sequence and the comparison vector data sequence to obtain a Euclidean distance sequence and a cosine similarity sequence;

[0199] The Euclidean distance and cosine similarity are calculated for the data in each dimension to obtain the Euclidean distance sequence and cosine similarity sequence respectively.

[0200] Example of original building multi-dimensional vector data sequence (original building data)

[0201] original_vectors=np.array([[0.0,0.25,0.0,0.0,0.0,0.0,0.0,0.66666667,0.0,0.0,0.0,0.0],[0.5,0.75 ,0.5,0.0,1.0,0.0,0.0,0.0,0.5,0.0,0.0,0.0],[1.0,1.0,1.0,0.0,0.0,1.0,0.0,0.0,0.0,1.0,1.0,0.0]]).

[0202] Example Comparison of building multidimensional vector data sequences (Comparison of building data)

[0203] comparison_vectors=np.array([[0.25,0.3,0.16666667,0.0,0.0,0.0,0.0,0.66666667,0.0,0.0,0.0,0.0],[0.75,0 .8,0.41666667,0.0,0.0,0.0,1.0,0.0,0.5,0.0,0.0,0.0],[1.0,0.9,1.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,1.0,1.0]])

[0204] Specifically, first import the required libraries, including numpy, euclidean (from scipy.spatial.distance), and cosine_similarity (from sklearn.metrics.pairwise).

[0205] The original building multi-dimensional vector data sequence and the comparison building multi-dimensional vector data sequence of the example are represented by original_vectors and comparison_vectors respectively.

[0206] Compute the Euclidean distance sequence:

[0207] Use the zip function to iterate over each pair of vectors in original_vectors and comparison_vectors.

[0208] For each pair of vectors, compute the Euclidean distance using the euclidean function and add the result to the euclidean_distances list.

[0209] Convert the euclidean_distances list to a numpy array.

[0210] Again, use the zip function to iterate over each pair of vectors in original_vectors and comparison_vectors.

[0211] For each pair of vectors, compute the cosine similarity using the cosine_similarity function and add the result to the cosine_similarities list.

[0212] Convert the cosine_similarities list to a numpy array.

[0213] Through the above steps, we can get the Euclidean distance sequence and cosine similarity sequence respectively, which will be used for subsequent similarity calculation and evaluation.

[0214] S107, performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain a first similarity value and a second similarity value;

[0215] Based on the preset weights, the Euclidean distance sequence and the cosine similarity sequence are weightedly calculated to obtain a first similarity value and a second similarity value, which comprehensively reflect the similarity between the buildings.

[0216] You can use np.sum(weights*euclidean_distances) to calculate the weighted Euclidean distance and get the first similarity value.

[0217] You can use np.sum(weights*cosine_similarities) to calculate the weighted cosine similarity and get the second similarity value.

[0218] Specifically, it can be calculated as follows:

[0219] The first similarity value calculation:

[0220] WES=∑(i=1to n)w_i*Euclidean_Distance_i;

[0221] Among them, Euclidean_Distance_i represents the similarity of the i-th dimension data, w_i is the weight of the i-th dimension data, and WES represents the first similarity value;

[0222] Calculation of the second similarity value:

[0223] WCS=∑(i=1ton)w_i*D_geometry_i

[0224] Wherein, D_geometry_i represents the similarity of the i-th dimension data, w_i is the weight of the i-th dimension data, and WCS represents the second similarity value.

[0225] S108: Evaluate the building based on the first similarity and the second similarity.

[0226] The building is evaluated based on the first and second similarity values ​​obtained. The evaluation can be qualitative (such as high, medium, and low similarity levels) or quantitative (such as specific similarity scores), depending on actual needs and application scenarios.

[0227] Through these steps, the similarities and differences of buildings in terms of structure, materials, environmental conditions, historical use and real-time monitoring can be systematically evaluated, providing a scientific basis for building safety management and decision-making.

[0228] In the above embodiment, step S107 performs weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively. Figure 2 An optional specific implementation method of obtaining the first similarity value and the second similarity value includes:

[0229] S1071, performing Z-score standardization conversion on the first similarity and the second similarity respectively to obtain a first conversion value and a second conversion value;

[0230] This step includes the following specific operations:

[0231] Calculate the mean (μ1) and standard deviation (σ1) of the Euclidean distance sequence.

[0232] Calculate the mean (μ2) and standard deviation (σ2) of the cosine similarity sequence.

[0233] The Z-score standardization conversion of the Euclidean distance sequence is calculated as follows:

[0234] Z1=(Euclidean_Distance-μ1) / σ1

[0235] The Z-score standardization conversion of the cosine similarity sequence is calculated as follows:

[0236] Z2=(Cosine_Similarity-μ2) / σ2

[0237] S1072: performing weighted calculation on the first conversion value and the second conversion value based on a preset weight to obtain a comprehensive similarity;

[0238] This step includes the following specific operations:

[0239] According to the actual application scenarios and expert experience, the weights of Euclidean distance and cosine similarity are set, denoted as w1 and w2 respectively.

[0240] The first conversion value and the second conversion value are weighted by using the preset weights to obtain the comprehensive similarity S, and the calculation formula is:

[0241] S=w1*Z1+w2*Z2

[0242] S1073, evaluating the building according to the comprehensive similarity;

[0243] This step includes the following specific operations:

[0244] Set the evaluation criteria and thresholds according to the comprehensive similarity S. For example, the following evaluation criteria can be set:

[0245] When S > 1, the building is evaluated as "high security".

[0246] When 0 < S ≤ 1, the building is evaluated as "medium security".

[0247] When S ≤ 0, the building is evaluated as "low security".

[0248] According to the set evaluation criteria, conduct a security evaluation for each building. Through standardization and weighted calculation in the above specific implementation steps, the obtained comprehensive similarity can more comprehensively reflect the similarity between buildings, thereby improving the accuracy and reliability of the security evaluation.

[0249] In an optional embodiment, the data that needs to be quantified and encoded in this application includes:

[0250] Construction age, structural system, chemical resistance of materials, and service load;

[0251] The specific method of the quantification and encoding may include:

[0252] Determine the stage code of the construction age from a pre-configured table. In the pre-configured table, the years are divided into several stages with a 10-year period for each stage, and a stage code is configured for each stage;

[0253] Determine the stage code of the construction age from a pre-configured table. In the pre-configured table, the years are divided into several stages with a 10-year period for each stage, and a stage code is configured for each stage. For example:

[0254] - 2000 - 2009: Encoded as 1

[0255] - 2010 - 2019: Encoded as 2

[0256] - 2020 - 2029: Encoded as 3

[0257] And so on. Through table lookup, determine the stage code to which the construction age belongs.

[0258] Determine the structure code of the structural system from a pre-configured table. The structural system includes frame structure, frame center support structure, frame eccentric support structure, tube structure, mega-frame structure, single-factory structure, and bent structure, and corresponding structure codes are configured for each structure;

[0259] Determine the structure code of the structural system from a pre-configured table. The structural system includes:

[0260] -Frame structure: coded as 1

[0261] -Frame center support structure: coded as 2

[0262] -Frame eccentric support structure: coded as 3

[0263] -Simplified structure: coded as 4

[0264] -Jumbo frame structure: coded as 5

[0265] -Single plant structure: code is 6

[0266] - Shelf structure: coded as 7

[0267] Determine the code of the structural system used in the building by looking up in the table.

[0268] Determine the chemical resistance code of the structural system from a preconfigured table, wherein the chemical resistance includes high corrosion resistance, medium corrosion resistance and low corrosion resistance, high fire resistance, medium fire resistance and low fire resistance; wherein each material chemical resistance is configured with a corresponding chemical resistance code and is represented by a binary array;

[0269] Chemical resistance code that determines the material's chemical resistance from a preconfigured table of:

[0270] -Corrosion resistance:

[0271] -High corrosion resistance: coded as [1,0,0]

[0272] -Medium corrosion resistance: coded as [0,1,0]

[0273] - Low corrosion resistance: coded as [0,0,1]

[0274] - Fire resistance: - High fire resistance: coded as [1,0,0]

[0275] -Medium fire resistance: coded as [0,1,0]

[0276] - Low fire resistance: coded as [0,0,1]

[0277] By looking up the table, determine the binary array code corresponding to the corrosion resistance and fire resistance of the material.

[0278] The load code of the usage load is determined from a preconfigured table, wherein the usage load includes a high load, a medium load and a low load, wherein each usage load is configured with a corresponding load code.

[0279] The load code of the usage load is determined from a preconfigured table, the usage load comprising:

[0280] -High load: coded as 3

[0281] -Medium load: coded as 2

[0282] - Low load: coded as 1

[0283] Determine the load code corresponding to the building's usage load by looking up the table

[0284] In an optional embodiment, the cosine similarity is calculated by the following formula:

[0285] D_geometry=∑(i=1to n)w_i*(X_i-Y_i)^2

[0286] The Euclidean distance is calculated by the following formula:

[0287] EuclideanDistance=sqrt(∑(i=1to n)(X_i-Y_i)^2)

[0288] Among them, X_i represents the i-th dimension data in the original building data, Y_i represents the i-th dimension data in the compared building data, D_geometry represents cosine similarity, EuclideanDistance represents Euclidean distance, and sqrt represents square root operation.

[0289] In an optional embodiment, performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain the first similarity value and the second similarity value comprises:

[0290] The calculation is done as follows:

[0291] The first similarity value calculation:

[0292] WES=∑(i=1to n)w_i*Euclidean_Distance_i;

[0293] Among them, Euclidean_Distance_i represents the similarity of the i-th dimension data, w_i is the weight of the i-th dimension data, and WES represents the first similarity value;

[0294] Calculation of the second similarity value:

[0295] WCS=∑(i=1ton)w_i*D_geometry_i

[0296] Wherein, D_geometry_i represents the similarity of the i-th dimension data, w_i is the weight of the i-th dimension data, and WCS represents the second similarity value.

[0297] Based on the above embodiments, the present application also provides a method for calculating the comprehensive similarity, as follows:

[0298] The comprehensive similarity is calculated by the following formula:

[0299] S_composite=w_e*WES+w_c*WCS;

[0300] Where S_composite represents the composite similarity, w_e represents the weight of the first similarity value, and w_c represents the weight of the second similarity value.

[0301] The above embodiment has described in detail the building safety evaluation method based on the vector space model provided in the present application. The following describes the building safety evaluation device based on the vector space model provided in the present application.

[0302] See also Figure 3 The present application provides a building safety evaluation device based on a vector space model, the device comprising:

[0303] The first data acquisition unit 301 is used to acquire original building data and classify the original building data, wherein the classification includes: geometric structure data, material performance data, environmental condition data, usage history data, and real-time monitoring data;

[0304] The geometric structure data includes the building age, height-to-width ratio, number of above-ground floors, structural system, structural length-to-width ratio, structural width and structural span;

[0305] The material performance data include material density, material elastic modulus, material chemical resistance, aging degree and structural corrosion ratio;

[0306] Data on environmental conditions include annual precipitation, mean annual temperature, and earthquake frequency;

[0307] The usage history data includes usage load;

[0308] The real-time monitoring data includes deformation vector, column stress, beam stress, vibration frequency and vibration amplitude;

[0309] A data determination unit 302, configured to determine data that needs to be quantized and encoded in the original building data;

[0310] The data encoding unit 303 is used to perform quantization encoding on the data that needs quantization encoding according to a preconfigured encoding rule;

[0311] A second data acquisition unit 304 is used to acquire comparative building data, wherein the comparative building data has a consistent data structure with the original building data;

[0312] A sequence construction unit 305 is used to construct a multi-dimensional vector data sequence based on the original building data and the comparison building data to obtain a corresponding original data sequence and a comparison vector data sequence;

[0313] A similarity calculation unit 306 is used to calculate the Euclidean distance and cosine similarity of the original data sequence and the comparison vector data sequence respectively to obtain a Euclidean distance sequence and a cosine similarity sequence;

[0314] A weighted calculation unit 307, configured to perform weighted calculations on the Euclidean distance sequence and the cosine similarity sequence to obtain a first similarity value and a second similarity value;

[0315] The evaluation unit 308 is configured to evaluate the building based on the first similarity and the second similarity.

[0316] Optionally, the weighted calculation unit 307 is specifically used for:

[0317] Performing Z-score standardization conversion on the first similarity and the second similarity respectively to obtain a first conversion value and a second conversion value;

[0318] Based on a preset weight, performing a weighted calculation on the first conversion value and the second conversion value to obtain a comprehensive similarity;

[0319] The buildings are evaluated according to the comprehensive similarity.

[0320] The data that needs to be quantized and encoded include:

[0321] Building age, structural system, chemical resistance of materials and use loads;

[0322] The data encoding unit 303 is specifically used for:

[0323] Determining the phase code of the building age from a preconfigured table, wherein the preconfigured table divides the year into a plurality of phases of 10 years each, and configures a phase code for each phase;

[0324] Determine the structure code of the structure system from a preconfigured table, the structure system including frame structure, frame central support structure, frame eccentric support structure, simplified structure, giant frame structure, single plant structure and bent structure, wherein each structure is configured with a corresponding structure code;

[0325] Determine the chemical resistance code of the structural system from a preconfigured table, wherein the chemical resistance includes high corrosion resistance, medium corrosion resistance and low corrosion resistance, high fire resistance, medium fire resistance and low fire resistance; wherein each material chemical resistance is configured with a corresponding chemical resistance code and is represented by a binary array;

[0326] The load code of the usage load is determined from a preconfigured table, wherein the usage load includes a high load, a medium load and a low load, wherein each usage load is configured with a corresponding load code.

[0327] Optionally, the weighted calculation unit 307 is specifically used for:

[0328] The comprehensive similarity is calculated by the following formula:

[0329] S_composite=w_e*WES+w_c*WCS;

[0330] Where S_composite represents the composite similarity, w_e represents the weight of the first similarity value, and w_c represents the weight of the second similarity value.

[0331] Optionally, the sequence construction unit 305 is specifically used for:

[0332] Import the required libraries into the system, including numpy, euclidean, and cosine_similarity;

[0333] The original building vector data sequence and the comparison building vector data sequence are represented by original_vectors and comparison_vectors respectively;

[0334] Use the zip function to iterate over each pair of vectors in original_vectors and comparison_vectors;

[0335] For each pair of vectors, calculate the Euclidean distance using the euclidean function and add the result to the euclidean_distances list;

[0336] Convert the euclidean_distances list to a numpy array;

[0337] Use the zip function to iterate over each pair of vectors in original_vectors and comparison_vectors;

[0338] For each pair of vectors, calculate the cosine similarity using the cosine_similarity function and add the result to the cosine_similarities list;

[0339] Convert the cosine_similarities list to a numpy array.

[0340] See also Figure 4, the present application also provides a building safety evaluation device based on a vector space model, comprising:

[0341] Processor 401, memory 402, input and output unit 403, bus 404;

[0342] The processor 401 is connected to the memory 402, the input and output unit 403 and the bus 404;

[0343] The memory 402 stores a program, and the processor 401 calls the program to execute any of the above methods.

[0344] The present application also relates to a computer-readable storage medium on which a program is stored, wherein when the program is run on a computer, the computer is caused to execute any of the above methods.

[0345] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0346] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0347] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0348] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0349] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, read-only memory), random access memory (RAM, random access memory), disk or optical disk and other media that can store program code.

Claims

1. A building safety evaluation method based on a vector space model, characterized in that: The method comprises: Acquire original building data, and classify the original building data, wherein the classification includes: geometric structure data, material performance data, environmental condition data, usage history data, and real-time monitoring data; The geometric structure data includes the building age, height-to-width ratio, number of above-ground floors, structural system, structural length-to-width ratio, structural width and structural span; The material performance data include material density, material elastic modulus, material chemical resistance, aging degree and structural corrosion ratio; Data on environmental conditions include annual precipitation, mean annual temperature, and earthquake frequency; The usage history data includes usage load; The real-time monitoring data includes deformation vector, column stress, beam stress, vibration frequency and vibration amplitude; Determining data that needs to be quantized and encoded in the original building data; According to a preconfigured encoding rule, quantization encoding is performed on the data that needs quantization encoding; Acquire comparison building data, wherein the comparison building data has a consistent data structure with the original building data; Constructing a multidimensional vector data sequence based on the original building data and the comparison building data respectively to obtain a corresponding original data sequence and a comparison vector data sequence; Calculating the Euclidean distance and cosine similarity of the original data sequence and the comparison vector data sequence respectively to obtain a Euclidean distance sequence and a cosine similarity sequence; Performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain a first similarity value and a second similarity value; The building is evaluated based on the first similarity and the second similarity.

2. The building safety evaluation method based on vector space model according to claim 1, characterized in that: The performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain a first similarity value and a second similarity value comprises: Performing Z-score standardization conversion on the first similarity and the second similarity respectively to obtain a first conversion value and a second conversion value; Based on a preset weight, performing a weighted calculation on the first conversion value and the second conversion value to obtain a comprehensive similarity; The buildings are evaluated according to the comprehensive similarity.

3. The building safety evaluation method based on vector space model according to claim 2, characterized in that: The data that needs to be quantized and encoded include: Building age, structural system, chemical resistance of materials and use loads; The quantization coding includes: Determining the phase code of the building age from a preconfigured table, wherein the preconfigured table divides the year into a plurality of phases of 10 years each, and configures a phase code for each phase; Determine the structure code of the structure system from a preconfigured table, the structure system including frame structure, frame central support structure, frame eccentric support structure, simplified structure, giant frame structure, single plant structure and bent structure, wherein each structure is configured with a corresponding structure code; Determine the chemical resistance code of the structural system from a preconfigured table, wherein the chemical resistance includes high corrosion resistance, medium corrosion resistance and low corrosion resistance, high fire resistance, medium fire resistance and low fire resistance; wherein each material chemical resistance is configured with a corresponding chemical resistance code and is represented by a binary array; The load code of the usage load is determined from a preconfigured table, wherein the usage load includes a high load, a medium load and a low load, wherein each usage load is configured with a corresponding load code.

4. The building safety evaluation method based on vector space model according to claim 3 is characterized in that: The cosine similarity is calculated by the following formula: D_geometry=∑(i=1to n)w_i*(X_i-Y_i)^2 The Euclidean distance is calculated by the following formula: EuclideanDistance=sqrt(∑(i=1to n)(X_i-Y_i)^2) Among them, X_i represents the i-th dimension data in the original building data, Y_i represents the i-th dimension data in the compared building data, D_geometry represents cosine similarity, EuclideanDistance represents Euclidean distance, and sqrt represents square root operation.

5. The building safety evaluation method based on vector space model according to claim 4, characterized in that: The performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain a first similarity value and a second similarity value comprises: The calculation is done as follows: The first similarity value calculation: WES=∑(i=1to n)w_i*Euclidean_Distance_i; Among them, Euclidean_Distance_i represents the similarity of the i-th dimension data, w_i is the weight of the i-th dimension data, and WES represents the first similarity value; Calculation of the second similarity value: WCS=∑(i=1ton)w_i*D_geometry_i Wherein, D_geometry_i represents the similarity of the i-th dimension data, w_i is the weight of the i-th dimension data, and WCS represents the second similarity value.

6. The building safety evaluation method based on vector space model according to claim 5, characterized in that: Based on a preset weight, performing weighted calculation on the first conversion value and the second conversion value to obtain a comprehensive similarity includes: The comprehensive similarity is calculated by the following formula: S_composite=w_e*WES+w_c*WCS; Where S_composite represents the composite similarity, w_e represents the weight of the first similarity value, and w_c represents the weight of the second similarity value.

7. The building safety evaluation method based on vector space model according to claim 1, characterized in that: The constructing of a multidimensional vector data sequence based on the original building data and the comparison building data to obtain a corresponding original data sequence and comparison vector data sequence comprises: Import the required libraries into the system, including numpy, euclidean, and cosine_similarity; The original building vector data sequence and the comparison building vector data sequence are represented by original_vectors and comparison_vectors respectively; Use the zip function to iterate over each pair of vectors in original_vectors and comparison_vectors; For each pair of vectors, calculate the Euclidean distance using the euclidean function and add the result to the euclidean_distances list; Convert the euclidean_distances list to a numpy array; Use the zip function to iterate over each pair of vectors in original_vectors and comparison_vectors; For each pair of vectors, calculate the cosine similarity using the cosine_similarity function and add the result to the cosine_similarities list; Convert the cosine_similarities list to a numpy array.

8. A building safety evaluation device based on a vector space model, characterized in that: The device comprises: A first data acquisition unit, used to acquire original building data and classify the original building data, wherein the classification includes: geometric structure data, material performance data, environmental condition data, usage history data, and real-time monitoring data; The geometric structure data includes the building age, height-to-width ratio, number of above-ground floors, structural system, structural length-to-width ratio, structural width and structural span; The material performance data include material density, material elastic modulus, material chemical resistance, aging degree and structural corrosion ratio; Data on environmental conditions include annual precipitation, mean annual temperature, and earthquake frequency; The usage history data includes usage load; The real-time monitoring data includes deformation vector, column stress, beam stress, vibration frequency and vibration amplitude; A data determination unit, used for determining data that needs to be quantized and encoded in the original building data; A data encoding unit, configured to perform quantization encoding on the data that needs quantization encoding according to a preconfigured encoding rule; A second data acquisition unit is used to acquire comparative building data, wherein the comparative building data has a consistent data structure with the original building data; A sequence construction unit, used to construct a multi-dimensional vector data sequence based on the original building data and the comparison building data, respectively, to obtain a corresponding original data sequence and a comparison vector data sequence; A similarity calculation unit, used to calculate the Euclidean distance and cosine similarity of the original data sequence and the comparison vector data sequence respectively, to obtain a Euclidean distance sequence and a cosine similarity sequence; A weighted calculation unit, used for performing weighted calculation on the Euclidean distance sequence and the cosine similarity sequence respectively to obtain a first similarity value and a second similarity value; An evaluation unit is configured to evaluate the building based on the first similarity and the second similarity.

9. A building safety evaluation device based on a vector space model, characterized in that: The device comprises: Processor, memory, input-output unit, and bus; The processor is connected to the memory, the input and output unit, and the bus; The memory stores a program, and the processor calls the program to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a program stored thereon, wherein the program, when executed on a computer, performs the method according to any one of claims 1 to 7.

Citation Information

Patent Citations

  • Dam safety monitoring single-node assessment method based on multiple correlation sequences

    CN110188399A

  • Steel structure building group safety assessment method based on geometry and performance similarity

    CN113505954A

  • Comprehensive monitoring, early warning and safety evaluation scheme for tailing pond multi-element disasters

    CN116050837A

Cited By

  • Pest and disease monitoring, analyzing and processing system based on agriculture and forestry economic management

    CN119625542A