METHOD FOR DEVELOPING DATA ON SHAKE INTENSITY ON BUILDING FLOORS WITH VARIATIONS IN SOIL, MATERIALS, AND EARTHQUAKES

Abstract

This invention concerns a method for developing big data on the intensity of earthquake shaking on building floors under various conditions. The objective of this method is to develop more specific big data on the intensity of shaking on each floor of a building compared to previous inventions, specifically the intensity of shaking on each floor of a building, and to accommodate various conditions, namely soil type, building materials, number of floors, earthquake strength, epicenter distance, and hypocenter depth. The results of this method can be directly utilized in an AI-based earthquake disaster mitigation system.The method of developing a data set consists of receiving earthquake recording data with various characteristics (earthquake strength, epicenter distance, and hypocenter depth) from earthquake stations; generating basic parameters of soil type, building material, and number of floors; constructing a building model with a spring-mass-damped system; dynamic analysis of the earthquake time history to evaluate the acceleration of each floor; calculating the shaking intensity of each floor based on the acceleration response from the dynamic analysis; and storing the output in the form of data set to cover all variations of conditions.

Description

BIG DATA DEVELOPMENT METHOD OF SHOCKS INTENSITY ON EACH FLOOR OF THE BUILDING WITH VARIOUS SOIL TYPES, BUILDING MATERIALS, NUMBER OF FLOORS, AND EARTHQUAKE DATA Invention Engineering Field This invention concerns a Big Data Development Method Intensity of Shaking on Each Floor of the Building with Variations Soil Type, Building Material, Number of Floors, and Earthquake Data, more specifically, the present invention relates to the use of artificial intelligence (AI) technology in predicting one of the seismic parameters, namely the value the intensity of the shaking on the building. In its application, especially in the AI-based earthquake disaster mitigation system, a big data of floor shaking intensity values buildings in various building conditions are very varied needed to ensure the level of prediction accuracy that tall. Background of the Invention The intensity of a seismic shock is a numerical value that describes the severity of an earthquake event reviewed from the impact on the earth's surface, humans, and structures (USGS, 2021). In the mitigation system earthquake disaster, the intensity of the shaking of the building is important parameters that must be predicted to minimize victims (Gan, 2023). Analysis of the structure currently used to calculate the value of the shock intensity takes time which is quite long, while the disaster mitigation system it takes a short time to predict the value. The seismic intensity scale used in this invention is the Seismic Intensity Scale (SIS) which was developed in a systematic manner quantitatively by the Japan Meteorological Agency (UMA, 1996). Artificial intelligence (AI) technology, especially Deep Learning (DL) is proposed to solve the problem because its proven accuracy and speed in completing various problems. However, the level of accuracy of this method determined by the amount of data trained in DL System. This invention is intended to answer the problem availability of big data on shock intensity values on each floor of the building. Technological inventions related to big data value the intensity of the shaking on each floor of the building is expressed as stated in national patent number S00202006543 with the title Quantitative Evaluation Methods for Levels Intensity of Building Shaking on Each Floor Due to Earthquake, which proposes the importance of the comfort aspect in general quantitative interpreted with building shocks, in the planning and design phase of a fire-resistant building earthquake. However, the invention cannot yet be applied to disaster mitigation system that requires fast processing speed calculation / prediction. Another patent related to seismic intensity big data was discovered from the World Intellectual Property Organization (WIPO) with title “Seismic Intensity Rapid Evaluation Method Based On Social Media Big Data And Machine Learning” (patent no CN111239812) and “Multifunctional big data seismic intensity instrument system" (patent no. CN113311476). Both patents also related to the mitigation of this earthquake disaster applied in China. The first patent focuses on Rapid justification of post-earthquake seismic areas, while patents The second focuses on the intensity measuring instrumentation system. seismic. The shaking objects analyzed in the patents This has not yet reached the calculation of shocks on each floor. building, but only limited to the ground level (ground) . Furthermore, the proposed invention is intended to overcome the problems that have been described previously, namely related to the development of big data on the intensity of shocks in each building floors. Brief Description of the Invention The main objective of this invention is to develop algorithms for calculating and collecting large amounts of data (big data) data) required in data processing with technology artificial intelligence (AI) for predicting seismic intensity values on each floor of the building. In addition to being intended for the system earthquake early warning, this invention can also be applied in the evaluation of building structures or Structural Health Monitoring System (SHMS) in buildings and in the design phase for considering the comfort aspect. Short Description of Image Figure 1 is a flowchart of big data development. building floor shaking intensity value. Figure 2 is a picture of the SIS value calculation algorithm with applying the Kawasumi filtration formula and Fast transformation Fourier. Figure 3 is an illustration of input variations in building conditions and earthquake and SIS big data output, with explanation of notation as following. Min 1 mass per floor of the superstructure kiimn 1 stiffness of each floor of the superstructure Cin 1 damping per floor of the upper structure i : floor level n : Number of floors H : total height of the building mp : foundation mass kh : horizontal stiffness coefficient ky : rotational stiffness Ch : horizontal damping Cr : rotational damping M : Earthquake magnitude G : Earthquake epicenter Kh: Distance from building to epicenter h : depth of earthquake source from epicenter Figure 4 is an illustration of the big data application process in the process Prediction of building floor shaking intensity values ​​with Deep Learning. Complete Description of the Invention Artificial intelligence (AI) technology, especially Deep Learning methods Learning, has been widely used to predict parameters- certain parameters in earthquake events. Based on previous theories and research, the accuracy of the DL method is determined from the large amount of data collected. This invention is submitted to support the use of AI in predicting value shock intensity (SIS) of each floor of the building due to the earthquake, by developing a method for collecting big data SIS (seismic intensity scale) on each floor of the building. Application of This invention will improve the accuracy of prediction algorithms. the SIS value. This invention will be fully described with reference to to the accompanying images. Referring to the images 1, which shows a flowchart of the Development Method Big Data on Building Floor Shaking Intensity Due to Earthquakes, The stages in this invention begin with data collection earthquakes at various earthquake stations, including magnitude (size) earthquake), distance from the epicenter, and depth of the earthquake source. From the data In this case, dynamic time history analysis was performed on the model spring-mass building to obtain acceleration response on each floor (denoted by a,,,(t) in Figure 2). Response This acceleration is influenced by variations in building materials, which directly affects the mass of each floor of the building (Min) and the stiffness of the building structure (kn) In addition, the variation soil type affects the dynamic parameters on the lower structure of the building, namely horizontal stiffness (k,) and rotational stiffness (k,), as shown in Figure 3. The acceleration response of each floor (denoted by a,,,(t) in Figure 2) which is the result of the previous calculation be input in calculating the SIS value on each floor building by applying the Kawasumi filtration formula and Fast Fourier transform, as shown in Figure 2. These two stages are executed iteratively until they cover all variations of conditions, namely soil type (hard, medium, soft), building materials (concrete, steel, wood), number of floors buildings, and earthquake data (magnitude, epicenter distance, and hypocenter depth), as shown in Figure 3. Next, all the SIS calculation results are collected. in spreadsheet form ready for further use in the data training process in the DL method. From the description above it is clear that the results of this invention can provide benefits for the development of intelligence technology Artificial Intelligence (AI), especially the Deep Learning method in predicting the value of the shaking intensity of each floor of the building (Figure 4). This is demonstrated by the novelty of ideas and methods, as well as coverage of a wide and representative variety of conditions actual conditions in the field. The results of this method can be directly used in the DL method. Based on the benefits mentioned, this invention actually presents a very practical improvements especially in the Method Big Data Development of Building Floor Shaking Intensity Due to the Earthquake.

Claims

1. A method for developing big data on shock intensity on each floor of the building with varying soil types, building materials, number of floors, and earthquake data consists of the following stages: a. calculate the acceleration response of each floor of the building which is modeled with a spring-mass model with variations in soil type, type of building material, quantity floors, and earthquake data, b. operate the Kawasumi filtration formula and Fast Fourier transform (Figure 2) on the output stage a iteratively until it covers the whole data variations mentioned in the first stage to obtain the values ​​of shock intensity on each floor of the building, Cc. recapitulates all intensity value outputs shocks on each floor of the building that has been taking into account variations in soil types, materials buildings, number of floors, and earthquake data in the form of big data spreadsheets.

2. The big data development method of claim 1, wherein: Aa. Variations in soil types include hard soil, loamy soil medium, and soft soil, b. variations in types of building materials include materials concrete, steel, and wood, Cc. Earthquake data that varies from magnitude, distance epicenter, and depth of the earthquake source.

Images